Advanced Remote Sensing for Urban and Landscape Ecology 9819930057, 9789819930050

This book introduces the use of various remote sensing data such as microwave, hyperspectral and very high-resolution (V

265 92 15MB

English Pages 325 [326] Year 2023

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Preface
Contents
Editors and Contributors
1 Data and Urban Poverty: Detecting and Characterising Slums and Deprived Urban Areas in Low- and Middle-Income Countries
1.1 Introduction
1.2 Methodology
1.2.1 Operationalising IDEAMAPS Domains of Deprivation Framework Through Semi-automated SLUMAP Processes
1.2.2 Urban Deprivation Detection Using Open Software and Low-Cost Imagery at City Scale
1.2.3 Urban Deprivation Characterization: At Settlement-Scale
1.2.4 SLUMAP Web-Based Portal: Towards Dissemination That Reaches Out to Local Users
1.3 Results
1.3.1 Detecting Deprived Areas at City Scale
1.3.2 Intra- and Inter-urban Deprivation Characterisation Through Land Use and Land Cover Indicators
1.3.3 Intra-urban Deprivation Characterisation Through Morphology Indicators
1.3.4 SLUMAP Web-Based Data Portal
1.4 Discussion
1.4.1 The Importance of Open-Access Data that Deals with Data Ethics and Privacy
1.4.2 Departing from Binary Slum Versus Non-slum Maps Towards Characterising Living Conditions
1.4.3 Understanding Intra- and Inter-urban Deprivation Diversity in Support of Pro-poor Policies
1.5 Conclusion
References
2 Investigation of Ecological Sustainability Through the Landscape Approach of Geospatial Technology: Study from New Town Project in Eastern India
2.1 Introduction
2.1.1 Background of the Landscape Dynamics in the Study Area
2.2 Database and Methodology
2.2.1 Data Consideration
2.2.2 Extraction of Land Use Land Cover Features in Change Dynamics of Urban Environment
2.2.3 Landscape Metrics to Represent Different Aspects of Landscape Configuration in Ecological Sustainability
2.2.4 Application of Fuzzy Membership Functions for Normalization of Landscape Metric Rasters
2.2.5 Application of Analytical Hierarchy Process (AHP) for Determining Individual Land Use-Based and Composite Land Use-Based Ecological Sustainability
2.2.6 Validation of the Proposed LUCESI Model
2.3 Results and Analysis
2.3.1 Spatio-temporal Dynamics of Land Use and Land Cover in the NTR Region
2.3.2 Spatio-temporal Distribution of Different Facets of Landscape in Ecological Sustainability
2.3.3 Spatio-temporal Distribution of Individual Land Use-Based Ecological Sustainability and Composite Land Use-Based Ecological Sustainability
2.3.4 Measurement of Association Between LUCESI and NDVI
2.4 Discussion
2.5 Conclusion
References
3 Advanced Remote Sensing for Sustainable Decent Housing for the Economically Challenged Urban Households
3.1 Introduction
3.2 Background
3.2.1 Attributes to Adequate Housing
3.2.2 Economically Challenged Urban Households
3.2.3 Percentage of Urban Population Living in the Economically Challenged Urban Areas
3.2.4 Spatial Characteristics for Economically Challenged Urban Areas
3.2.5 Remote Sensing for Economically Challenged Urban Areas
3.2.6 Advanced Remote Sensing Techniques for Mapping Economically Challenged Urban Areas
3.2.7 Approaches to Information Extraction on Economically Challenged Urban Areas
3.3 Study Area
3.4 Materials and Methods
3.4.1 Conceptual Framework for Advanced Remote Sensing for Economically Challenged Urban Areas
3.4.2 Datasets
3.4.3 Methods
3.5 Results and Discussions
3.6 Conclusions and Recommendations
References
4 Impact of Uncontrolled Tourism Development on Landscape Ecology of Purba Medinipur Coastal Region, West Bengal: A 4-C Framework and SWOC Analysis
4.1 Introduction
4.2 Background
4.3 Study Area
4.4 Materials and Methods
4.5 Result and Discussion
4.5.1 Changing Population Scenario
4.5.2 Changing LULC Scenario Due to Tourism in the Study Area
4.5.3 Major Causes for Tourism Development and Tourist Flow
4.5.4 Roles of Tourism for Landscape Transformation in the Study Area
4.5.5 Problem and Management Scenario for Tourism in Study Area
4.5.6 SWOC Analysis (Strength-Weakness-Opportunity-Challenges)
4.5.7 Essential Dimensions for Smart Tourism and Challenges in the Study Area
4.5.8 Main Challenges that Tourist Destinations Faced On
4.5.9 Ten Steps on the Way Forward Against Tourism Cum Urban Sprawling
4.6 Conclusion
References
5 Impact of Urban Heat Island: A Local-Level Urban Climate Phenomenon on Urban Ecology and Human Health
5.1 Introduction
5.2 Classification of UHI
5.3 Factors Contributing to Urban Heat Island Formation
5.4 Effects of UHI on Urban Ecology and Human Health
5.4.1 Urban Ecology
5.4.2 Human Health
5.5 Case Study
5.5.1 Background of Study
5.5.2 Materials and Methods
5.5.3 Results and Discussion
5.5.4 Conclusion and Recommendations
5.6 Recommendations
5.7 Conclusions
References
6 Identification of Environmental Epidemiology Through Advanced Remote Sensing Based on NDVI
6.1 Introduction
6.2 NDVI
6.3 Acceptance of Use of NDVI Globally
6.4 Case Study
6.4.1 Study Area
6.4.2 Methodology of Assessment
6.4.3 Result and Discussion
6.5 Conclusion
References
7 Assessment of Land Utilization Pattern and Their Relationship with Surface Temperature and Vegetation in Sikkim, India
7.1 Introduction
7.2 Study Area Description
7.3 Methodology
7.3.1 Data Sets Used
7.3.2 Land Use/Cover Classification and Accuracy Assessment
7.3.3 NDVI Computation
7.3.4 LST Computation
7.3.5 Correlation Analysis
7.4 Results and Discussion
7.4.1 Accuracy Assessment
7.4.2 LULC Change Dynamics
7.4.3 LST and NDVI Dynamics (1995–2005–2021)
7.4.4 Correlation Between LST Versus NDVI
7.5 Conclusion
References
8 Monitoring Land Use and Land Cover Change Over Bhiwani District Using Google Earth Engine
8.1 Introduction
8.1.1 Background
8.2 Study Area
8.3 Materials and Methods
8.4 Results and Discussion
8.5 Conclusion and Recommendation
References
9 Image and Perception of Royal Heritage and Eco-space of the Medium Towns in India: Reflection from Burdwan Royal Heritage Site
9.1 Introduction
9.2 Background of the Study
9.3 Study Area and its Rationale
9.4 Materials and Methods
9.4.1 Questionnaire Design and Sampling
9.4.2 Socioeconomic Profile of the Respondents
9.4.3 Methodology Adopted
9.5 Results and Discussions
9.5.1 Narrative Analysis
9.5.2 Exploratory Factor Analysis
9.5.3 Factor Loadings and Reliability Statistics
9.6 Conclusion and Recommendations
References
10 Governance and Floodplain Extent Changes of Yamuna River Floodplain in Megacity Delhi
10.1 Introduction
10.2 Statement of Problem
10.3 Study Area
10.4 Relevance of Floodplain in Megacity of Delhi
10.5 Literature Review
10.6 Data Sources
10.6.1 Floodplain Extent Changes
10.6.2 Governance/Planning
10.7 Analysis/Discussion
10.7.1 Floodplain Extent Changes
10.7.2 Governance
10.8 Conclusion
References
11 Assessing Urban Compactness Using Machine Learning and Earth Observation Datasets: A Case Study of Kolkata City
11.1 Introduction
11.2 Methodology
11.2.1 Study Area
11.2.2 Datasets
11.2.3 Methods
11.3 Results and Discussion
11.3.1 Land Use and Land Cover
11.3.2 Spatial Metrices to Assess Urban Compactness
11.4 Conclusions
References
12 Analysis of Ecological Vulnerability Behind the Land Conversion from Agriculture to Aquaculture of Purba Medinipur District in West Bengal, India
12.1 Introduction
12.2 Background
12.3 Study Area
12.4 Materials and Methods
12.4.1 Change Detection of Land Use
12.4.2 Image Processing
12.4.3 Soil Sample Collection
12.4.4 Water Sample Collection
12.4.5 Statistical Analysis
12.4.6 Perception Analysis
12.5 Result and Discussion
12.5.1 Ecological Impact Analysis
12.5.2 Change Detection of LULC
12.5.3 Soil Quality Assessment
12.5.4 Assessment of Water Quality
12.5.5 Loss of Bio-diversity
12.5.6 Ecological Cost Benefit (EECB) Analysis
12.6 Recommendations of Road Map for the Development and Sustainable Coping Strategies
12.7 Conclusion
References
13 Environmental Change Analysis Using Remote Sensing and GIS: A Study of Upper Baitarani Basin, Odisha
13.1 Introduction
13.2 Background
13.3 Study Area
13.4 Materials and Methods
13.4.1 Data Collection
13.4.2 Method
13.5 Result and Discussion
13.5.1 Land Use/Landcover
13.5.2 Normalized Different Vegetation Index (NDVI)
13.6 Conclusion and Recommendation
References
14 Mapping Urban Footprint Using Machine Learning and Public Domain Datasets
14.1 Introduction
14.2 Study Area
14.3 Datasets and Methodology
14.3.1 Remote Sensing Data
14.3.2 Socio-economic Data
14.3.3 Methods
14.4 Result and Discussion
14.4.1 Land Use and Land Cover (LULC)
14.4.2 Training and Test Sample
14.4.3 Land Use and Land Cover Using Random Forest
14.4.4 Land Use and Land Cover Using SVM-Linear
14.4.5 Land Use and Land Cover Using SVM-RBF
14.4.6 Built-Up Classification
14.4.7 Validation
14.5 Conclusion
References
Recommend Papers

Advanced Remote Sensing for Urban and Landscape Ecology
 9819930057, 9789819930050

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Advances in Geographical and Environmental Sciences

Sk. Mustak Dharmaveer Singh Prashant Kumar Srivastava   Editors

Advanced Remote Sensing for Urban and Landscape Ecology

Advances in Geographical and Environmental Sciences Series Editors Yukio Himiyama, Hokkaido University of Education, Asahikawa, Hokkaido, Japan Subhash Anand, Department of Geography, University of Delhi, Delhi, India

Advances in Geographical and Environmental Sciences synthesizes series diagnostigation and prognostication of earth environment, incorporating challenging interactive areas within ecological envelope of geosphere, biosphere, hydrosphere, atmosphere and cryosphere. It deals with land use land cover change (LUCC), urbanization, energy flux, land-ocean fluxes, climate, food security, ecohydrology, biodiversity, natural hazards and disasters, human health and their mutual interaction and feedback mechanism in order to contribute towards sustainable future. The geosciences methods range from traditional field techniques and conventional data collection, use of remote sensing and geographical information system, computer aided technique to advance geostatistical and dynamic modeling. The series integrate past, present and future of geospheric attributes incorporating biophysical and human dimensions in spatio-temporal perspectives. The geosciences, encompassing land-ocean-atmosphere interaction is considered as a vital component in the context of environmental issues, especially in observation and prediction of air and water pollution, global warming and urban heat islands. It is important to communicate the advances in geosciences to increase resilience of society through capacity building for mitigating the impact of natural hazards and disasters. Sustainability of human society depends strongly on the earth environment, and thus the development of geosciences is critical for a better understanding of our living environment, and its sustainable development. Geoscience also has the responsibility to not confine itself to addressing current problems but it is also developing a framework to address future issues. In order to build a ’Future Earth Model’ for understanding and predicting the functioning of the whole climatic system, collaboration of experts in the traditional earth disciplines as well as in ecology, information technology, instrumentation and complex system is essential, through initiatives from human geoscientists. Thus human geoscience is emerging as key policy science for contributing towards sustainability/survivality science together with future earth initiative. Advances in Geographical and Environmental Sciences series publishes books that contain novel approaches in tackling issues of human geoscience in its broadest sense — books in the series should focus on true progress in a particular area or region. The series includes monographs and edited volumes without any limitations in the page numbers.

Sk. Mustak · Dharmaveer Singh · Prashant Kumar Srivastava Editors

Advanced Remote Sensing for Urban and Landscape Ecology

Editors Sk. Mustak Department of Geography Central University of Punjab Bathinda, Punjab, India

Dharmaveer Singh Symbiosis Institute of Geoinformatics Symbiosis International (Deemed University) Pune, Maharashtra, India

Prashant Kumar Srivastava Institute of Environment and Sustainable Development Banaras Hindu University Varanasi, Uttar Pradesh, India

ISSN 2198-3542 ISSN 2198-3550 (electronic) Advances in Geographical and Environmental Sciences ISBN 978-981-99-3005-0 ISBN 978-981-99-3006-7 (eBook) https://doi.org/10.1007/978-981-99-3006-7 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

A dramatic increase in the world’s urban population is observed since the middle of the twentieth century. This historic shift from rural to urban culture is having a tremendous socioeconomic impact, particularly outside of Europe and North America where this tendency is more pronounced. Urban environments are under pressure from new developments brought on by this population expansion and are causing threats to long-term urban sustainability. More than half of the world’s population who live in urban areas face significant environmental and social challenges, particularly those related to the effects of climate change, water and sanitation, the encroachment of agriculture onto land, the degradation of green infrastructure, mobility caused by the rapid rural-to-urban transition, and other anthropogenic interventions. Achieving the goals of urban sustainability requires the assessment and monitoring of urban landscapes and natural resources in the context of changing climate and environmental conditions. Given the requirement for data from many themes at various time frames, it is challenging to provide solutions at numerous spatial and temporal dimensions. But now that remote sensing techniques have been developed and computational intelligence has advanced, problems with data availability have been greatly reduced. Landscape-level solutions to urban and environmental problems can be provided by integrating remotely sensed data with soft-computing techniques. The book “Advanced Remote Sensing for Urban and Landscape Ecology” that has used a variety of remote sensing data, such as microwave, hyperspectral, and VHR; mapping techniques, such as pixel and object-based machine learning; and geostatistical modelling techniques, such as cellular automata, entropy, and land fragmentation techniques, is able to address the challenges at the urban and landscape levels to support sustainable development goals (SDGs). Bathinda, India Pune, India Varanasi, India

Sk. Mustak Dharmaveer Singh Prashant Kumar Srivastava

v

Contents

1

2

3

4

Data and Urban Poverty: Detecting and Characterising Slums and Deprived Urban Areas in Low- and Middle-Income Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Monika Kuffer, Angela Abascal, Sabine Vanhuysse, Stefanos Georganos, Jon Wang, Dana R. Thomson, Anthony Boanada, and Pere Roca Investigation of Ecological Sustainability Through the Landscape Approach of Geospatial Technology: Study from New Town Project in Eastern India . . . . . . . . . . . . . . . . . . . . . . . . Anirban Kundu and Sk. Mafizul Haque

1

23

Advanced Remote Sensing for Sustainable Decent Housing for the Economically Challenged Urban Households . . . . . . . . . . . . . . F. N. Karanja and P. W. Mwangi

63

Impact of Uncontrolled Tourism Development on Landscape Ecology of Purba Medinipur Coastal Region, West Bengal: A 4-C Framework and SWOC Analysis . . . . . . . . . . . . . . . . . . . . . . . . . Manishree Mondal, Rabin Das, Chayon Chakraborty, Puja Karmakar, and Sk. Mustak

83

5

Impact of Urban Heat Island: A Local-Level Urban Climate Phenomenon on Urban Ecology and Human Health . . . . . . . . . . . . . . 113 Sangita Singh, Priya Priyadarshni, and Puneeta Pandey

6

Identification of Environmental Epidemiology Through Advanced Remote Sensing Based on NDVI . . . . . . . . . . . . . . . . . . . . . . 129 Vibhanshu Kumar, Birendra Bharti, Harendra Prasad Singh, Himanshu Kumar, and Sanjay Paul Kujur

vii

viii

Contents

7

Assessment of Land Utilization Pattern and Their Relationship with Surface Temperature and Vegetation in Sikkim, India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Shashi Sekhar, Nitu Singh, Sudhir Kumar Singh, Meenakshi Dhote, and Kumar Rajnish

8

Monitoring Land Use and Land Cover Change Over Bhiwani District Using Google Earth Engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Suraj Kumar Singh, Shruti Kanga, Bhartendu Sajan, Sayali Madhukarrao Diwate, and Gaurav Tripathi

9

Image and Perception of Royal Heritage and Eco-space of the Medium Towns in India: Reflection from Burdwan Royal Heritage Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Koyel Sarkar and Sanat Kumar Guchhait

10 Governance and Floodplain Extent Changes of Yamuna River Floodplain in Megacity Delhi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Shobhika Bhadu and Milap Punia 11 Assessing Urban Compactness Using Machine Learning and Earth Observation Datasets: A Case Study of Kolkata City . . . 229 Prosenjit Barman and Sk. Mustak 12 Analysis of Ecological Vulnerability Behind the Land Conversion from Agriculture to Aquaculture of Purba Medinipur District in West Bengal, India . . . . . . . . . . . . . . . . . . . . . . . . 251 Manishree Mondal, Ramu Guchhait, and Sk. Mustak 13 Environmental Change Analysis Using Remote Sensing and GIS: A Study of Upper Baitarani Basin, Odisha . . . . . . . . . . . . . . 283 Tapas Ranjan Patra, Priyanka Chakraborty, Diptimayee Naik, and Ashis Chandra Pathy 14 Mapping Urban Footprint Using Machine Learning and Public Domain Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Prosenjit Barman and Sk. Mustak

Editors and Contributors

About the Editors Dr. ir. Sk. Mustak is an Assistant Professor at the Department of Geography, the Central University of Punjab, Bathinda (India). His main research areas include applied earth observation, geoinformation, and artificial intelligence in urban and regional planning, and landscape ecology. He served as a research associate and research coordinator at the Ecoinformatics lab, Ashoka Trust for Research in Ecology and the Environment (ATREE), Bangalore, during 2018–2020. Dr. Mustak completed his M.Sc. in geoinformation science and earth observation from the Faculty of GeoInformation Science and Earth Observation (ITC), the University of Twente, the Netherlands; and his Ph.D. in urban geography with a specialization in advanced remote sensing and artificial intelligence from Pt. Ravishankar Shukla University, Raipur, India. He was awarded an ITC fellowship, many other grants, and international travel grants. He also received research grants from the Central University of Punjab, Bathinda, in 2020. He has presented 19 research papers at national and international conferences and published 15 research papers in national and international peer-reviewed journals, books, and proceedings. He is a reviewer for several journals. Dr. Dharmaveer Singh is an Assistant Professor at the Symbiosis Institute of Geoinformatics, Symbiosis International (Deemed University), Pune, India. He was a research Scientist-C at the National Institute of Hydrology, Roorkee, India, from 2016 to 2018. Dr. Singh completed Ph.D. in geoinformatics in 2015 at the Motilal Nehru National Institute of Technology, Allahabad. He received several academic awards, including the University Grants Commission (UGC) Early Career Research Grant in 2019, a post-doctoral fellowship from the Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences in 2016, and the Council of Scientific and Industrial Research—Junior Research Fellowship (National Eligibility Test) in 2009 and 2010. He held academic positions as a visiting scientist

ix

x

Editors and Contributors

at the Institute of International Rivers and Eco-Security (IIRES), Yunnan University in 2021–2022. His main research interests include water resources management, water security and economics, and climate change and modelling. He has published 30 research papers in peer-reviewed journals and conference proceedings. He is currently heading research projects sponsored by the UGC and is a reviewer for several scientific journals. Dr. Prashant Kumar Srivastava works at the Institute of Environment and Sustainable Development, Banaras Hindu University, as a faculty member in remote sensing. He received his doctoral degree from the Department of Civil Engineering, University of Bristol, Bristol, UK. He received several awards such as the National Aeronautics and Space Administration (NASA) Fellowship, University of Maryland Fellowship, Commonwealth Fellowship, and Early Career Research Award. He leads a number of projects funded by reputed agencies in India as well as abroad. He is also a collaborator with the NASA Jet Propulsion Laboratory on the Soil Moisture Active Passive (SMAP) mission as well as Scatterometer Satellite-1 (ScatSat-1), NASA-ISRO SAR Mission (NISAR), and Airborne Visible/Infrared Imaging Spectrometer (AVIRISNG) missions of India. He has published more than 180 peer-reviewed articles in journals and 9 books with renowned publishing houses such as Springer, Taylor and Francis, Wiley, and Elsevier. He also serves as a regional editor of Geocarto International, an associate editor of Journal of Hydrology, and more.

Contributors Angela Abascal Navarra Centre for International Development (NCID), Universidad de Navarra, Pamplona, Spain Prosenjit Barman Department of Geography, School of Environment and Earth Sciences, Central University of Punjab, Bhatinda, Punjab, India Shobhika Bhadu CSRD, School of Social Sciences, Jawaharlal Nehru University, New Delhi, India Birendra Bharti Department of Water Engineering & Management, Central University of Jharkhand, Ranchi, India Anthony Boanada Institute of Management in Latin America, University of St.Gallen, São Paulo, Brazil Chayon Chakraborty Department of Geography (UG & PG), Midnapore College (Autonomous), Midnapore, West Bengal, India Priyanka Chakraborty Department of Geography, Rajendra University, Prajna Vihar, Balangir, Odisha, India

Editors and Contributors

xi

Rabin Das Department of Geography (UG & PG), Bajkul Milani Mahavidyalaya, Midnapore, West Bengal, India Meenakshi Dhote School of Planning and Architecture, ENVIS RP, New Delhi, India Sayali Madhukarrao Diwate Centre for Climate Change and Water Research, Suresh Gyan Vihar University, Jaipur, Rajasthan, India Stefanos Georganos Geomatics Unit, Karlstad University, Karlstad, Sweden Ramu Guchhait Department of Geography, Midnapore College (Autonomous), Midnapore, West Bengal, India Sanat Kumar Guchhait Department of Geography, The University of Burdwan, Purba Bardhaman, India Shruti Kanga School of Environment and Earth Sciences, Department of Geography, Central University of Punjab, Bathinda, India F. N. Karanja Department of Geospatial Engineering & Space Technology, University of Nairobi, Nairobi, Kenya Puja Karmakar Department of Geography (UG & PG), Midnapore College (Autonomous), Midnapore, West Bengal, India Monika Kuffer Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, Enschede, The Netherlands Sanjay Paul Kujur Department of Geoinformatics, Central University of Jharkhand, Ranchi, India Himanshu Kumar Department of Water Engineering & Management, Central University of Jharkhand, Ranchi, India Vibhanshu Kumar Department of Water Engineering & Management, Central University of Jharkhand, Ranchi, India Anirban Kundu Department of Geography, University of Calcutta, Kolkata, India Sk. Mafizul Haque Department of Geography, University of Calcutta, Kolkata, India Manishree Mondal Department of Geography (UG & PG), Midnapore College (Autonomous), Midnapore, West Bengal, India Sk. Mustak Department of Geography, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India P. W. Mwangi Department of Spatial and Environmental Planning, Kenyatta University, Nairobi, Kenya Diptimayee Naik Department of Geography, Rajendra University, Prajna Vihar, Balangir, Odisha, India

xii

Editors and Contributors

Puneeta Pandey Department of Environmental Science and Technology, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India; Centre of Environmental Studies, University of Allahabad, Uttar Pradesh, Prayagraj, India Ashis Chandra Pathy Department of Geography, Utkal University, Vani Vihar, Bhubaneswar, Odisha, India Tapas Ranjan Patra Department of Geography, Rajendra University, Prajna Vihar, Balangir, Odisha, India Priya Priyadarshni Department of Environmental Science and Technology, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India Milap Punia CSRD, School of Social Sciences, Jawaharlal Nehru University, New Delhi, India Kumar Rajnish ENVIS Cell, Ministry of Environment, Forest and Climate Change, GoI, New Delhi, India Pere Roca Barcelona, España Bhartendu Sajan Centre for Climate Change and Water Research, Suresh Gyan Vihar University, Jaipur, Rajasthan, India Koyel Sarkar Department of Geography, The University of Burdwan, Purba Bardhaman, India Shashi Sekhar School of Planning and Architecture, ENVIS RP, New Delhi, India Harendra Prasad Singh Department of Water Engineering & Management, Central University of Jharkhand, Ranchi, India Nitu Singh Survey of India, Pushpa Bhawan, New Delhi, India Sangita Singh Department of Environmental Science and Technology, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India Sudhir Kumar Singh K. Banerjee Centre of Atmospheric and Ocean Studies, IIDS, Nehru Science Centre, University of Allahabad, Prayagraj, Uttar Pradesh, India Suraj Kumar Singh Centre for Sustainable Development, Suresh Gyan Vihar University, Jaipur, Rajasthan, India Dana R. Thomson Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, Enschede, The Netherlands Gaurav Tripathi Centre for Climate Change and Water Research, Suresh Gyan Vihar University, Jaipur, Rajasthan, India

Editors and Contributors

xiii

Sabine Vanhuysse Department of Geosciences, Environment & Society, Université libre De Bruxelles (ULB), Brussels, Belgium Jon Wang Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, Enschede, The Netherlands

Chapter 1

Data and Urban Poverty: Detecting and Characterising Slums and Deprived Urban Areas in Low- and Middle-Income Countries Monika Kuffer, Angela Abascal, Sabine Vanhuysse, Stefanos Georganos, Jon Wang, Dana R. Thomson, Anthony Boanada, and Pere Roca

Abstract A multidimensional characterisation of urban areas is essential to provide relevant data for monitoring deprived urban areas (urban poverty) beyond the dollar threshold (World Bank) or household characterisation (UN-Habitat). We present a holistic characterisation of deprivation through a framework composed of domains

M. Kuffer · J. Wang · D. R. Thomson Faculty of Geo-Information Science and Earth Observation (ITC), University of Twente, Enschede, The Netherlands e-mail: [email protected] J. Wang e-mail: [email protected] D. R. Thomson e-mail: [email protected] A. Abascal (B) Navarra Centre for International Development (NCID), Universidad de Navarra, Pamplona, Spain e-mail: [email protected] S. Vanhuysse Department of Geosciences, Environment & Society, Université libre De Bruxelles (ULB), Brussels, Belgium e-mail: [email protected] S. Georganos Geomatics Unit, Karlstad University, Karlstad, Sweden e-mail: [email protected] A. Boanada Institute of Management in Latin America, University of St.Gallen, São Paulo, Brazil e-mail: [email protected] P. Roca Barcelona, España e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_1

1

2

M. Kuffer et al.

and indicators for measuring urban poverty. It includes socio-economic and household characterisation (household-level) as well as the characterisation of physical and environmental conditions (area-level). In this chapter, we showcase the use of Earth Observation techniques to extract area-level data. The combination of Earth Observation and open geospatial data allows routine mapping and characterising essential aspects of urban deprivation related to the urban environment (e.g., contamination such as waste accumulations), urban morphology (e.g., unplanned urbanisation defined by built-up densities, street geometry, open/green spaces), and connectivity (e.g., the presence of infrastructures such as streetlights or road access). Such a mapping system provides meaningful information for classifying deprivation levels and discovering differences between and within deprived areas. Results are provided as an online tool for users to access information at the city and settlement scale in sub-Saharan African cities. The tool allows users to tailor information to support the improvement of living conditions for the rapidly growing number of urban inhabitants. Keywords Cities · Slums · Earth observation · Spatial analysis · Deep learning

1.1 Introduction A balanced housing demand–supply and access to employment, basic infrastructure, and services are often unattainable in urban areas, leading to the emergence and proliferation of “slums” or “informal settlements”, referred to as deprived urban areas (DUAs) in this chapter. Consequently, the majority of the sub-Saharan African urban population lives in DUAs. Policy responses to support DUA upgrading have been inadequate (UN-Habitat 2016), leaving out the poorest citizens residing in substandard housing, commonly located in areas that lack urban facilities, services, and infrastructure, and are often unplanned and located in hazardous and contaminated zones. These areas are frequently considered urban “anomalies” and are not included in city planning (Beukes 2015; Sliuzas et al. 2008). However, there is insufficient data on urban conditions to identify and characterise the poorest areas, and therefore there is a lack of data monitoring programs. Local actors, including local and national governments, non-governmental organisations (NGOs), and community groups, require up-to-date data to guide and monitor pro-poor policies and upgrading programs (Owusu et al. 2021) (Fig. 1.1). Generally, consistent and city-level data are urgently required to monitor the progress of the urban Sustainable Development Goal 11 (SDG 11), in particular SDG 11.1.1. on the “proportion of the urban population living in slums, informal settlements or inadequate housing” (UN-Habitat 2020a). However, data supporting its monitoring is not readily available. Available datasets on the SDG 11.1.1 indicator are country-level estimates that do not provide localised data (Kuffer et al. 2018; Kavvada et al. 2020). Thus, available data do not provide information about city-scale spatial patterns of DUAs and their dynamics. Geospatial and Earth Observation (EO)

1 Data and Urban Poverty: Detecting and Characterising Slums …

3

Fig. 1.1 Percentage of urbanisation rates (coloured by country) and the urban population living in slums (shown only for countries with >40% of the urban population living in slums) (Abascal et al. 2022a)

data have demonstrated their capabilities to map and monitor DUAs (Kuffer et al. 2020, 2021a; Liu et al. 2019). Data are increasingly available and accessible, and algorithm developments, big data, and cloud-based computation have been bridging the bottleneck to providing city-to-global mapping products. However, most EObased mapping studies on DUAs work at a city scale or below (Ajami et al. 2019; Badmos et al. 2018; Wurm et al. 2019) and do not deal with the main challenges of scalable, transferable, and low-cost solutions (Kuffer 2020). To date, EO data have not shown their capabilities to provide adequate monitoring instruments supporting knowledge on the location, characterisation, and dynamics of DUAs (Kuffer et al. 2021b), e.g., supporting local SDG 11.1.1 reviews. A recently published needs assessment of different user groups that require such data (Fig. 1.2) has shown that data are needed for a fine-grained spatial scale (e.g., gridded data at 100 m) with an update cycle of 1–2 years (Kuffer et al. 2021b). Generally, urban development questions require city-scale information and detailed neighbourhood analysis (Kuffer et al. 2018; Merodio Gómez et al. 2021). Two interrelated developments aim to produce city-scale and settlement-scale data on DUAs. The Integrated Deprived Area Mapping System (IDEAMAPS) network (IDEAMAPS 2022) leverages the strengths of current poverty geodata sources, i.e., census, field-based survey, and EO-based mapping, the latter through research carried out in the SLUMAP project (SLUMAP 2022). The combination of different mapping approaches aims at overcoming the limitations of individual approaches to DUA mapping, e.g., the deep knowledge of community-based mapping combined with the capacity to cover large areas with EO-based mapping. However, until recently, most EO-based studies (e.g., Wang et al. 2019; Williams et al. 2020) relied on veryhigh-resolution (VHR) satellite image acquisition (high-cost imagery and processing demands). To support routine and accurate mapping and characterisation of deprived urban areas, the IDEAMAPS network developed the Domain of Deprivation Framework to

4

M. Kuffer et al.

Fig. 1.2 User requirements of deprived (slum) related spatial information (Kuffer et al. 2021)

identify relevant geospatial and EO data for urban deprivation mapping and analysis (Abascal et al. 2022b). This framework builds on existing deprivation frameworks (e.g., the English Deprivation Index and the Bellagio Framework (McLennan et al. 2019; Gale et al. 2013)). The primary rationale for modelling deprivation not as a binary phenomenon but as a continuous layer is the high level of uncertainties of slum versus non-slum maps, as even local experts have difficulties agreeing on boundaries (Kohli et al. 2016). Existing deprivation mapping frameworks typically use census data, with availability issues and low temporal granularities, which quickly go out of date in fast-growing and transforming LMIC cities (Thomson et al. 2021). The IDEAMAPS Domains of Deprivation Framework groups locally meaningful DUA indicators into nine domains at three scales. Two domains reflect household deprivation (socio-economic status and housing characterisation). Four domains reflect arealevel deprivations (social hazards and assets, physical hazards and assets, unplanned urbanisation, and contamination). Three domains reflect aspects of deprivation that relate to the connectivity to the city (i.e., infrastructure, facilities and services, and governance). A guide for authorities and users (IDEAMAPS 2021) has been developed to support the operationalisation of all domains building on openly available geospatial data (e.g., night-time lights (see Kuffer et al. 2018) or sub-air pollution) and contextual image features (e.g., using Sentinel-2 imagery) (Fig. 1.3). The SLUMAP project that focused on sub-Saharan African cities investigated the potential of EO for mapping and characterising DUAs, aiming for cost efficiency. Experiments involving machine learning algorithms and satellite imagery with different spatial resolutions showed that city-level deprivation mapping could be achieved with open, cost-free Sentinel-1/2 images, while very-high-resolution images were requested to produce detailed settlement-level characterisation (Abascal et al. 2022; Georganos et al. 2021; Vanhuysse et al. 2021). At the city scale, the

1 Data and Urban Poverty: Detecting and Characterising Slums …

5

Fig. 1.3 IDEAMAPS domain of deprivation framework for LMIC cities (Abascal et al. 2022b)

location and extent of slums within a city were modelled, while detailed characteristics of the physical environment within slums were extracted at the settlement scale. The methodological objectives included scalability and transferability of the developments. IDEAMAPS and SLUMAP work on DUA models that utilise open geospatial and EO data. In particular, EO processes have advantages over traditional methods, such as cost-effectiveness, fine grain coverage, and high temporality. Thus, EO data allow for routine mapping of DUAs and characterising aspects related to the physical urban environment observed from space, connected to area-level domains. Such area-level information includes the urban environment (e.g., contamination and infrastructure to detect waste accumulations or other environmental hazards), unplanned urbanisation defined by the urban morphology (e.g., built-up densities, street width, and availability of open/green spaces), and urban facilities, services, and infrastructure (e.g., availability of streetlights, health services, and road access). EO approaches are commonly top-down, with no or limited user interactions. In contrast, our framework combines EO data with user engagement and includes data from local communities, acknowledging the importance of citizen science. Thus, the information needs and requirements of different user groups are the guiding principles for developing a flexible DUA mapping system. The deprivation data is also frequently analysed and disseminated in academic formats not easily accessible to local users (e.g., DUA communities, NGOs, or governments). To develop strategies for shortening the gap between inconsistent and non-available datasets in DUAs,

6

M. Kuffer et al.

the first step needs to develop an adequate understanding of user requirements at different scales (i.e., from the local community level to the national government and international organisations).

1.2 Methodology 1.2.1 Operationalising IDEAMAPS Domains of Deprivation Framework Through Semi-automated SLUMAP Processes The IDEAMAPS “domains of deprivation framework” allows the flexibility of integrating existing geostatistical models with innovative spatial-data cubes. The availability of geospatial data differs between locations, particularly in sub-Saharan African countries where available data are often outdated or have a low spatial resolution (coarse data). To mitigate the data scarcity issue, we use EO data to show their capacity to partially align with the IDEAMAPS framework by extracting relevant indicators. We use the city of Nairobi (Kenya) to explore the potential of highresolution (HR) and very-high-resolution (VHR) sensors (i.e., Sentinel-1/2, SPOT6/ 7, WorldView-3, and Google Earth images) to produce deprivation indicators. We employ Sentinel-1/2 images to reduce the costs of city-scale mapping, responding to user requirements for a low-cost mapping system. Working with publicly available satellite imagery allows for the development of standardised, transferable, and scalable mapping methods, i.e., supporting routine mapping of DUAs. We use a gridded mapping system of 100 by 100 m grid cells to avoid showing exact settlement boundaries that might contribute to stigmatisation. Furthermore, at the city scale, the overall probability of an area being deprived is displayed, which avoids binary labels (i.e., slums vs. non-slum areas). At the city scale, indicators can exhibit built-up densities, which is often a good proxy of deprivation levels. The local characterisation relies on the potential of VHR images to describe aspects of the urban morphology that constitute environmental and health issues (e.g., waste/garbage piles) and automate building mapping and detection (in support of local planning needs). Results are made available through a user-friendly WebGIS interface that provides information on the city- and settlement scale to address the accessibility requirements of users identified in earlier work (Kuffer et al. 2021) (Fig. 1.4).

1 Data and Urban Poverty: Detecting and Characterising Slums …

7

Fig. 1.4 IDEAMAPS/SLUMAP solution to address user requirements for spatial data on DUAs

1.2.2 Urban Deprivation Detection Using Open Software and Low-Cost Imagery at City Scale We assessed the potential of cost-free Sentinel-1/2 versus low-cost SPOT-7 for mapping the gridded probability of deprivation at the city scale (Vanhuysse et al. 2021). For this purpose, we implemented a machine learning workflow using Free and Open-Source Software (FOSS). This was done with GRASS GIS and R functions, employing Jupyter Notebook for automation. A morphological deprivation probability was computed, reflecting the probability for a grid cell to be part of a DUA (which is different from a degree of deprivation severity) based on morphological characteristics captured by EO. We opted for gridded mapping as this approach was successfully used in previous studies, although with very-high-resolution imagery (e.g., Duque et al. 2017). Besides, gridded layers can easily be combined with the rapidly increasing number of other open gridded mapping products (e.g., Facebook’s High-Resolution Population Layer, WorldPop, Global Human Settlement Layers (GHSL), and World Settlement Footprint (WSF)). We developed our workflow in the pilot study area of Nairobi, Kenya. Nairobi was selected because of data availability and the large number of inhabitants living in DUAs (around 60% of Nairobi’s population live in DUAs). Our study area covers the entire Nairobi metropolitan area, including continuous built-up areas extending outside the city boundaries but excluding the Nairobi National Park (the total area is 652 km2 ). We used Sentinel-1/2 and SPOT-7, i.e., publicly available images from Copernicus and low-cost commercial imagery (Fig. 1.5) and compared the results. The mapping approach also includes ancillary open global datasets (i.e., SRTM, Open Street Map (OSM), WSF 2019; Marconcini et al. 2020). In the first step, a vast number of image features were extracted (over 2000 spectral, spatial, and ancillary features). For optical imagery (S2 and SPOT7), these features included various vegetation indices, water or moisture indices, built-up indices, image transforms, texture metrics (e.g., Grey Level Co-occurrence Matrix (GLCM), Structural Feature Set), and a few metrics calculated on an unsupervised classification output (such as the Mean Patch Size). The image features of the SAR imagery (S1) included intensity, coherence, textures, and filtered bands. The list of ancillary features covered geomorphometric features, built-up, and street density. Statistics were calculated at the grid-cell level. Feature selection was used to reduce the high

8

M. Kuffer et al.

Fig. 1.5 The interface between deprived and non-deprived urban areas. Top left: GE imagery. Top right: SPOT7 (RGB). Bottom left: S2 (RGB). Bottom right: S1intensity (VV, VH, VV/VH) (Kuffer et al. 2021a)

feature dimensionality. We employed the Variable Selection Using Random Forest (VSURF) algorithm (Genuer et al. 2015). For classification, we used a classical machine learning approach (Random Forest) that allows working with a limited amount of labelled training data (as compared to deep learning). The classification scheme includes the following eight land-use/cover classes: (1) High to middensity-built area, (2) Low density-built area, (3) Industry/large structures, (4) Paved ground/Bare ground, (5) Vegetation, (6) Water, (7) Deprived urban areas (typical), and (8) DUAs (atypical). For the class “deprived urban areas”, we used two sub-types that have relevantly different morphological characteristics, based on the following definition of the deprived classes: . (7) Very high built-up densities in the form of compact arrangement of low-rise buildings that form “organic” patterns. The area has no structured street layout except for a few main streets (often at the boundary). Little or no vegetation. . (8) Areas shows variations in densities (high- to mid-densities) with compact arrangements of buildings that are more regular than in class 7.

1 Data and Urban Poverty: Detecting and Characterising Slums …

9

For training and testing, 3962 manually labelled samples (i.e., grid cells) were prepared for the target classes. We compared the performance of several feature combinations using standard accuracy metrics (i.e., precision, recall, and F1 score) for optimising the process.

1.2.3 Urban Deprivation Characterization: At Settlement-Scale On top of a city-scale analysis, we also characterise settlement-scale intra-deprived area environments (Georganos et al. 2021). This analysis included four main topics, (1) the accumulation of waste (i.e., garbage piles), (2) the built-up densities, (3) built-up morphology (i.e., morphological features of the built-up patterns), and (4) a bottom-up population estimate. For this level of analysis, we use VHR images. For the extraction of land cover/use information, we used superspectral image data acquired by the WorldView-3 satellite (8 multispectral and 8 SWIR bands). The main target classes are buildings, ground cover (bare soil and asphalt), vehicles, vegetation (tall and low vegetation), water, garbage, and shadow (residual class). For the image classification, a state-of-the-art processing framework utilising machine learning classifiers combined with Geographic Object-Based Image Analysis is deployed (Georganos et al. 2018a, c). Moreover, we analyse the produced land use/land cover models for deprived areas. The results are assessed, considering, besides quantitative assessment (i.e., overall accuracy), an analysis of interpretability and transferability (Georganos et al. 2018b). This assessment included, e.g., an analysis of a suitable grid size to reflect the urban morphological patterns in DUAs. Consequently, we compared indicators at different grid sizes (e.g., 25, 50, and 100 m). For example, we analysed a suitable grid size for analysing and visualising the density of garbage/ waste piles in settlements, one of the most severe environmental issues in deprived urban areas (Fig. 1.7a). A similar effort is presently ongoing in local communities collecting geodata about garbage piles in several deprived urban Nairobi regions (Fig. 1.6). Besides land cover/use information, we also extracted morphological information. This analysis builds on building footprints extracted from the baseline band set (red, green, and blue) of very-high-resolution imagery, as displayed on platforms such as Google Earth. We analysed whether morphological features differentiate DUAs from other parts of the city. This analysis entails two significant steps: (1) extracting building footprints using deep learning (a modified U-Net architecture) and (2) calculating morphological features based on the building patterns using the open-source Python library MOMEPY (2022). For the building footprint extraction training, we use a global training dataset provided by Wuhan University that contains labelled building footprints of a worldwide sample (Wuhan University 2020). The morphological analysis output provides a clustering result optimised in terms of cluster number. Building footprint extraction is required as an input to morphological

10

M. Kuffer et al.

Fig. 1.6 The living environment issues in deprived urban areas Community-based waste data collection ongoing work (Source https://slumap.ulb.be/news/trash_survey_foss4g/)

Fig. 1.7 Selection of image chips (top row) and corresponding building footprints, manually digitised (middle row) and predicted by CNN (bottom row), illustrating the challenges faced by automated methods in very densely built DUAs (Abascal et al. 2022c)

analysis, which is challenging in DUAs. Challenges are caused by the high densities and limited or no set-back between buildings, which causes clumps of buildings to be extracted (e.g., Fig. 1.7). Global or continental data layers of building footprints (e.g., Open Buildings by Google or Building Footprints by Microsoft) have significant omissions and large clumps in DUAs.

1 Data and Urban Poverty: Detecting and Characterising Slums …

11

1.2.4 SLUMAP Web-Based Portal: Towards Dissemination That Reaches Out to Local Users We developed a web portal that allows easy visualisation to optimise the user interaction with the mapping outputs. The data are packaged at city- and settlement scales, allowing for interactive visualisation. The city of Nairobi is used as a pilot, while other cities are presently processed (e.g., Accra, Dakar, Kampala, Kisumu, Ouagadougou, and Khartoum). The website allows users to switch between city scale to settlement scale information. Besides visualisation, users can make simple printouts of maps. Furthermore, comparing cities after implementing different cities will also be possible.

1.3 Results 1.3.1 Detecting Deprived Areas at City Scale Dimensionality reduction resulted in a dramatic decrease in the number of features required to achieve optimal accuracy and minimise the complexity of the application. Indeed, this allowed us to reduce processing time and supported the scalability of the method. The results were validated with an independent test set generated through visual interpretation. The validation focused on the two deprived classes (see deprivation definition of subclasses 7 and 8). Generally, the highest accuracy was obtained using SPOT7 and ancillary features. However, it did not differ substantially from the best combination of Sentinel 1/2 and ancillary features (Table 1.1). Consequently, the preference was given to using Sentinel-1/2 and ancillary features to support a low-cost mapping system. The morphological deprivation probability is the combined probability of the two classes of deprived areas (i.e., classes 7 and 8) (Fig. 1.8). The values show low Table 1.1 Precision, recall, and F1 scores of the best feature combinations employing Sentinel-1 (S1), Sentinel-2 (S2), SPOT7, and ancillary global features Class

Metric

S2 S1

S2 S1 Ancillary

SPOT7

SPOT7 Ancillary

Class 7

Precision

0.94

0.96

0.86

0.94

Class 7

Recall

0.89

0.89

0.89

0.93

F1

0.91

0.92

0.87

0.94

Precision

0.79

0.84

0.79

0.88

Recall

0.82

0.89

0.78

0.89

F1

0.80

0.86

0.79

0.89

12

M. Kuffer et al.

Fig. 1.8 Gridded deprivation probability (Nairobi, Kenya). Left: From S1-2 and ancillary open geodata. Right: From SPOT7 and ancillary geodata (Adapted from Kuffer et al. 2021a)

probability values for areas that are commonly understood as non-deprived areas (e.g., planned urban areas with regular road patterns and vegetation cover) and high probability values for the well-known deprived urban areas (locally known as slums or informal areas—such as Kibera/Kibra), which are characterised by high built-up density and limited or no green spaces.

1.3.2 Intra- and Inter-urban Deprivation Characterisation Through Land Use and Land Cover Indicators The example of the Mathare settlement in Nairobi (Fig. 1.9) shows the spatial patterns of land use/cover. This base map allows the extraction of information (indicators) about the urban environment that links to the domains of deprivation. The environmental conditions of DUAs are dominated by overcrowding and the absence of green and open space. Indicators that can be extracted from land cover/use information, e.g., relating to built-up densities, availability of roads that can be accessed by vehicles (for instance, cars cannot access high-density built-up areas, causing problems for infrastructure supply and in case of emergencies such as fires). Furthermore, environmental degradation is highlighted by the massive accumulation of waste (Fig. 1.9). Several indicators can also serve as socio-economic proxies (Engstrom et al. 2017), e.g., the presence of vehicles (Fig. 1.9) relates to accessibility and socio-economic activity. Concerning the classified maps, the overall accuracy (OA) using all WV3 bands (multispectral + shortwave infrared) surpassed 87%, exhibiting a strong potential to capture fine details of the urban environment. The produced indicators have been summarised at a coarser grid level to provide aggregated information on different aspects of the urban environment (Fig. 1.9). The gridded map shows the spatial patterns of waste concentration in Mathare. The patterns mostly follow the area’s drainage system, i.e., in case of rain, garbage flows into the central drainage

1 Data and Urban Poverty: Detecting and Characterising Slums …

13

Fig. 1.9 Land use/cover mapping in deprived areas in Mathare, Nairobi, highlights garbage piles, lack of openness, detection of vehicles, and gridded garbage density (%) (Adapted from Georganos et al. 2021; Kuffer et al. 2021)

systems, polluting the water and causing obstructions to the drainage. These obstructions increase the flood severity, often causing a massive loss of properties of an already very vulnerable population.

1.3.3 Intra-urban Deprivation Characterisation Through Morphology Indicators The morphological analysis provides a consistent method to extract similar morphological clusters. This can be done across cities, as shown in Fig. 1.10, to understand the comparability of morphologies of deprived urban areas across SSA cities. Figure 1.11 shows morphological diversity within Nairobi city with two samples of building footprints from a slum/deprived urban area and a planned/non-deprived urban area. The building configurations and shapes show significant differences

14

M. Kuffer et al.

Fig. 1.10 Urban deprivation clusters based on building morphology indicators

across the city. These differences can be measured with morphological metrics and are captured by different morphological clusters. Once the morphological building patterns are explicitly measured, clusters of similar building patterns are classified. The result is morphological clusters that show similar built-up morphologies. The morphological group highlighted in red

Fig. 1.11 a Building footprints extracted in recognised local slums. b Building footprints extracted in non-slums. c Clusters based on morphological metrics (Nairobi)

1 Data and Urban Poverty: Detecting and Characterising Slums …

15

(Fig. 1.11b) reflects the distribution of deprivation areas delineated by local actors (locally known as slums).

1.3.4 SLUMAP Web-Based Data Portal While users do not necessarily have access to and knowledge of working with EO data, GIS software is commonly used across diverse local actors. To make the geodata accessible, we package them in a web-based visualisation tool (Roca and SLUMAP 2022). The tool provides easy access to data processed at different levels. Gridded information about the deprivation probability and built-up densities at the city scale is combined with the highest resolution gridded population data (Facebook; see Fig. 1.12). We enable users to obtain estimates of the most deprived population from a specific area. At the settlement scale, for several DUAs, more detailed data is provided about the land use/cover, i.e., the built-up environment and aspects of environmental conditions (e.g., vegetation presence and garbage pile locations; see Fig. 1.13b). To further develop deprivation indicators, the land use/cover information has been developed into gridded indicators to develop deprivation indicators (Fig. 1.13a) further. These indicators include the density of vegetation cover, garbage piles, and build-up for each grid cell. This is combined with a bottom population estimate using local population data from DUAs in Nairobi. The web application allows users to change class boundaries in the case of continuous values, thus generating different maps, legends, and statistics. In addition, the user can activate/deactivate various categories of the legend and filter data by a defined range of values.

Fig. 1.12 SLUMAP web visualisation of morphological deprivation probability at city scale

16

M. Kuffer et al.

Fig. 1.13 a Mathare settlement build-up percentage, b Mathare zoom-in land cover classes

1.4 Discussion 1.4.1 The Importance of Open-Access Data that Deals with Data Ethics and Privacy The results of DUA mapping done within SLUMAP allow concluding that Sentinel images, which are cost-free datasets with wide temporal availability, are a valuable option for mapping some aspects of multidimensional deprivation at the city scale, e.g., the morphological deprivation probability. When well-trained and paired with additional features from available global datasets (e.g., WSF (Marconcini et al. 2020)), models using Sentinel-1/2 can reach accuracy levels close to models based on commercial HR imagery. The main advantages of Sentinel images are that they allow for frequent updates, are widely available and accessible, and can be processed with limited computational resources (compared to VHR images), thereby meeting users’ requirements (Kuffer et al. 2021). Geo-ethics and the potential implication of mapping vulnerable groups are significant when mapping poverty. In this respect, aggregating metrics at the grid level

1 Data and Urban Poverty: Detecting and Characterising Slums …

17

could be preferable to using disaggregated building metrics that expose citizens to personal identification, or administrative divisions, which can potentially cause the well-known modifiable areal unit problem (MAUP) (Openshaw 1984; Grippa et al. 2019). We aim to link our data environment with other platforms that provide deprivation data at a grid level, such as WUDAPT or WorldPop (Chi et al. 2021; Stewart and Oke 2012; Bondarenko 2020). Combining datasets that highlight deprivation aspects in a data-poor environment can shed light on topics that have been systematically omitted and allow to visualise the scale of the required solutions. For example, sanitation and water solutions (besides the main other infrastructure solutions) are urgently needed. The absence of improved sanitation and safe water is a major factor that reduces the life expectancy in such areas by 10–20 years. For designing locally relevant—low-cost solutions, besides co-designing such solutions, data that guide decision-making about technical requirements are crucial (Friesen et al. 2018).

1.4.2 Departing from Binary Slum Versus Non-slum Maps Towards Characterising Living Conditions Urban deprivation is very diverse (multidimensional) and cannot be simplified into binaries of slum and non-slum conditions (Baud et al. 2008, 2009; Georganos et al. 2021). Even in field experience (Fig. 1.14), mapping often experiences challenges in drawing the exact boundaries between slum and non-slum areas (Kohli et al. 2016; Pratomo et al. 2016). Furthermore, details on the variation in living conditions, local needs, and priorities are essential to use local resources to improve living conditions economically.

1.4.3 Understanding Intra- and Inter-urban Deprivation Diversity in Support of Pro-poor Policies Efforts to classify intra-slum variability according to its characteristics highlight the existence of spatially observable urban physical differences (Georganos et al. 2021; Kuffer et al. 2021; Noble and Wright 2013). The differences related to building density, absence or presence of vegetation, vehicles, and waste piles are the first step to better understanding the internal structure of deprived areas and provide meaningful indicators to support pro-poor policies and evidence-based policymaking for sustainable cities. Such differences have a locational-geographic dimension (Kuffer et al. 2017). For example, inner-city DUAs often have very high built-up and population densities, e.g., structures might have several floors, while peri-urban DUAs have more commonly lower built-up densities but are much more deprived in terms of access to infrastructure and services. However, densities also differ between cities, e.g., DUAs in Kisumu (Kenya) have much lower densities (on average) than DUAs in

18

M. Kuffer et al.

Fig. 1.14 Satellite and ground photos from diverse deprived urban areas in Nairobi City

Nairobi (Kenya; Karanja 2010; SLUMAP 2022). Upgrading infrastructure requires understanding and related data (Friesen et al. 2018). For example, the provision of infrastructure (e.g., sanitation) is more accessible in lower density DUAs, while in very-high-density DUAs, many public services (e.g., waste collection trucks or fire brigades) cannot enter, and local solutions (e.g., the use of handcars) need to be developed (Corburn et al. 2020; Wanjiru 2021). It is essential to understand that local solutions that are co-designed and locally owned are generally more successful and scalable than top-down interventions that are not considering the needs of communities (Kotadiya et al. 2018; Patel et al. 2015).

1.5 Conclusion DUAs emerge with the rapid urbanisation occurring in LMICs, ineffective planning, and a lack of affordable housing options (among other factors; see Cities Alliance 2021 and UN-Habitat 2020b). The proposed Integrated Deprived Area Mapping System framework (IDEAMAPS 2022) provides a flexible gridded mapping system to be operationalised in LMIC cities to uncover urban deprivation. SLUMAP showcases the potential of EO data for, on the one hand, producing city-scale maps that localise the diversity of deprivation and, on the other hand, unravel their characteristics across and within DUAs. A user-friendly web was created to provide relevant urban data to local actors. The proposed approach has the advantage of being scalable and transferable and allows for local adaptations in the form of a user-centred mapping approach. Results support cross-disciplinary information needs on DUAs

1 Data and Urban Poverty: Detecting and Characterising Slums …

19

and show EO data’s potential to be combined with geospatial data for local SDG monitoring.

References Abascal A, Vanhuysse S, Grippa T, Rodriguez I, Georganos S, Wang J, Kuffer M, Martinez-Diez P, Santamaria-Varas M, Wolff E (2022a) AI Perceives Like a Local: Unveiling Urban Deprivation Levels Using Satellite Imagery [Manuscript submitted for publication]. Navarra Centre for International Development, University of Navarra Abascal A, Rothwell N, Shonowo A, Thomson DR, Elias P, Elsey H, … Kuffer M (2022b) “Domains of deprivation framework” for mapping slums, informal settlements, and other deprived areas in LMICs to improve urban planning and policy: a scoping review. Comput Environ Urban Syst 93:101770. https://doi.org/10.1016/j.compenvurbsys.2022.101770 Abascal A, Rodríguez-Carreño I, Vanhuysse S, Georganos S, Sliuzas R, Wolff E, Kuffer M (2022c) Identifying degrees of deprivation from space using deep learning and morphological spatial analysis of deprived urban areas. Comput Environ Urban Syst 95(2013). https://doi.org/10.1016/ j.compenvurbsys.2022.101820 Ajami A, Kuffer M, Persello C, Pfeffer K (2019) Identifying a slums’ degree of deprivation from VHR images using convolutional neural networks. Remote Sens 11(11):1282. https://doi.org/ 10.3390/rs11111282 Baud I, Sridharan N, Pfeffer K (2008) Mapping urban poverty for local governance in an Indian mega-city: the case of Delhi. Urban Stud 45(7):1385–1412. https://doi.org/10.1177/004209800 8090679 Baud I, Pfeffer K, Sridharan N, Nainan N (2009) Matching deprivation mapping to urban governance in three Indian mega-cities. Habitat Int 33(4):365–377. https://doi.org/10.1016/j.habitatint.2008. 10.024 Beukes A (2015) Making the invisible visible: generating data on ‘slums’ at local, city and global scales. International Institute for Environment and Development, London, UK Badmos OS, Rienow A, Callo-Concha D, Greve K, Jürgens C (2018) Urban development in West Africa—monitoring and intensity analysis of slum growth in Lagos: linking pattern and process. Remote Sens 10(7):1044. https://doi.org/10.3390/rs10071044 Bondarenko M, Kerr D, Sorichetta A, Tatem A (2020) Census/projection-disaggregated gridded population datasets, adjusted to match the corresponding UNPD 2020 estimates, for 183 countries in 2020 using Built-Settlement Growth Model (BSGM) outputs. University of Southampton. https://doi.org/10.5258/SOTON/WP00685 [Dataset] Chi G, Fang H, Chatterjee S, Blumenstock JE (2021) Micro-estimates of wealth for all low-and middle-income countries. https://arxiv.org/ftp/arxiv/papers/2104/2104.07761.pdf Cities Alliance (2021) The challenge of slums—an overview of past approaches to tackle it. Cities Alliance, Brussels Corburn J, Vlahov D, Mberu B, Riley L, Caiaffa WT, Rashid SF, … Ayad H (2020) Slum health: arresting COVID-19 and improving well-being in urban informal settlements. J Urban Health. https://doi.org/10.1007/s11524-020-00438-6 Duque JC, Patino JE, Betancourt A (2017) Exploring the potential of machine learning for automatic slum identification from VHR imagery. Remote Sens 9(9):895. http://www.mdpi.com/20724292/9/9/895 Engstrom R, Hersh JS, Newhouse DL (2017) Poverty from space: using high-resolution satellite imagery for estimating economic well-being. Washington, DC. http://documents.worldbank. org/curated/en/610771513691888412/Poverty-from-space-using-high-resolution-satellite-ima gery-for-estimating-economic-well-being

20

M. Kuffer et al.

Friesen J, Taubenböck H, Wurm M, Pelz PF (2018) The similar size of slums. Habitat Int 73:79–88. https://doi.org/10.1016/j.habitatint.2018.02.002 Genuer R, Poggi J-M, Tuleau-Malot C (2015) VSURF: an R package for variable selection using random forests. R J 7/2. ISSN: 2073-4859 Georganos S, Grippa T, Lennert M, Vanhuysse S, Johnson BA, Wolff E (2018a) Scale matters: spatially partitioned unsupervised segmentation parameter optimization for large and heterogeneous satellite images. Remote Sens 10(9):1440. https://doi.org/10.3390/rs10091440 Georganos S, Grippa T, Vanhuysse S, Lennert M, Shimoni M, Kalogirou S, Wolff E (2018b) Less is more: optimizing classification performance through feature selection in a very-high-resolution remote sensing object-based urban application. Giscience Remote Sens 55(2):221–242. https:/ /doi.org/10.1080/15481603.2017.1408892 Georganos S, Grippa T, Vanhuysse S, Lennert M, Shimoni M, Wolff E (2018c) Very high resolution object-based land use-land cover urban classification using extreme gradient boosting. IEEE Geosci Remote Sens Lett 15(4):607–611. https://doi.org/10.1109/LGRS.2018.2803259 Georganos S, Abascal A, Kuffer M, Wang J, Owusu M, Wolff E, Vanhuysse S (2021) Is it all the same? Mapping and characterizing deprived urban areas using worldview-3 superspectral imagery. A case study in Nairobi, Kenya. Remote Sens 13(24):4986. https://www.mdpi.com/ 2072-4292/13/24/4986 Grippa T, Linard C, Lennert M, Georganos S, Mboga N, Vanhuysse S, Gadiaga A, Wolff E (2019) Improving urban population distribution models with very-high resolution satellite information. Data 4(1):13. https://www.mdpi.com/2306-5729/4/1/13 IDEAMAPS (2021). https://ideamapsnetwork.org/toolkit-goverment. Accessed 11 July 2022 IDEAMAPS (2022). https://ideamapsnetwork.org/. Accessed 11 July 2022 Karanja I (2010) An enumeration and mapping of informal settlements in Kisumu, Kenya, implemented by their inhabitants. Environ Urban 22(1):217–239. https://doi.org/10.1177/095624780 9362642 Kavvada A, Metternicht G, Kerblat F, Mudau N, Haldorson M, Laldaparsad S, … Chuvieco E (2020) Towards delivering on the sustainable development goals using earth observations. Remote Sens Environ 247:111930. https://doi.org/10.1016/j.rse.2020.111930 Kohli D, Stein A, Sliuzas R (2016) Uncertainty analysis for image interpretations of urban slums. Comput Environ Urban Syst 60:37–49. https://doi.org/10.1016/j.compenvurbsys.2016.07.010 Kotadiya R, Kuffer M, Sliuzas RV, Patel S (2018) Post-resettlement socio-economic dynamics. In: van Maarseveen MFAM, Martinez JA, Flacke J (eds) The case of Ahmedabad, India. CRC Press, Boca Raton, pp 245–266 Kuffer M (2020) Digitalization and urban development in the Global South: towards reliable population data in deprived urban areas. In: Österreichische Entwicklungspolitik 2020: Digitalization for Development? Challenges for Developing Countries. Österreichische Forschungsstiftung für Internationale Entwicklung, pp 73–81 Kuffer M, Orina F, Sliuzas R, Taubenböck H (2017, 6-8 March). Spatial patterns of slums: comparing African and Asian cities Joint urban remote sensing event (JURSE), Dubai, United Arab Emirates Kuffer M, Pfeffer K, Sliuzas R, Taubenböck H, Baud I, Maarseveen MV (2018) Capturing the urban divide in nighttime light images from the international space station. IEEE J Sel Top Appl Earth Obs Remote Sens 1–9. https://doi.org/10.1109/JSTARS.2018.2828340 Kuffer M, Thomson DR, Boo G, Mahabir R, Grippa T, Vanhuysse S, … Kabaria C (2020) The role of earth observation in an integrated deprived area mapping “system” for low-to-middle income countries. Remote Sens 12(6):982. https://doi.org/10.3390/rs12060982 Kuffer M, Vanhuysse S, Georganos S, Wang J (2021a) Meeting user requirements for mapping and characterizing deprived urban areas in support of pro-poor policies. GI_Forum 9(1):85–93. https://doi.org/10.1553/giscience2021_01_s85 Kuffer M, Wang J, Thomson DR, Georganos S, Abascal A, Owusu M, Vanhuysse S (2021b) Spatial information gaps on deprived urban areas (slums) in low-and middle-income countries: a usercentered approach. Urban Sci 5(4):72. https://doi.org/10.3390/urbansci5040072

1 Data and Urban Poverty: Detecting and Characterising Slums …

21

Kuffer M, Grippa T, Persello C, Taubenböck H, Pfeffer K, Sliuzas R (2021c) Mapping the morphology of urban deprivation. In: Urban remote sensing, pp 305–323 Liu R, Kuffer M, Persello C (2019) The temporal dynamics of slums employing a CNN-based change detection approach. Remote Sens 11(23). https://doi.org/10.3390/rs11232844 Marconcini M, Metz-Marconcini A, Üreyen S, Palacios-Lopez D, Hanke W, Bachofer F, … Strano E (2020) Outlining where humans live, the World Settlement Footprint 2015. Sci Data 7(1):242. https://doi.org/10.1038/s41597-020-00580-5 McLennan AD, Noble S, Noble M, Plunkett E, Wright G, Gutacker N (2019) The English indices of deprivation 2019: technical report. https://dera.ioe.ac.uk/34259/1/IoD2019_Technical_Rep ort.pdf. Accessed 11 July 2022 Merodio Gómez P, Juarez Carrillo OJ, Kuffer M, Thomson DR, Olarte Quiroz JL, Villaseñor García E, … Brito PL (2021) Earth observations and statistics: unlocking sociodemographic knowledge through the power of satellite images. Sustainability 13(22):12640. https://www. mdpi.com/2071-1050/13/22/12640. Accessed 11 July 2022 MOMEPY (2022). http://docs.momepy.org/en/stable/. Accessed 11 July 2022 Noble M, Wright G (2013) Using indicators of multiple deprivation to demonstrate the spatial legacy of apartheid in South Africa. Soc Indic Res 112(1):187–201. https://doi.org/10.1007/s11 205-012-0047-3 Openshaw S (1984) The modifiable areal unit problem. Geobooks. https://ci.nii.ac.jp/naid/100244 64407/en/ Owusu M, Kuffer M, Belgiu M, Grippa T, Lennert M, Georganos S, Vanhuysse S (2021) Towards user-driven earth observation-based slum mapping. Comput Environ Urban Syst 89:101681. https://doi.org/10.1016/j.compenvurbsys.2021.101681 Patel S, Sliuzas R, Mathur N (2015) The risk of impoverishment in urban development-induced displacement and resettlement in Ahmedabad. Environ Urban 27(1):231–256. https://doi.org/ 10.1177/0956247815569128 Pratomo J, Kuffer M, Martínez J, Kohli D (2016) Uncertainties in analyzing the transferability of the generic slum ontology. Paper presented at the GEOBIA 2016: solutions and synergies, Enschede, The Netherlands. https://doi.org/10.3990/2.428 Roca P, SLUMAP (2022) SLUMAP web-based data portal. https://pere.gis-ninja.eu/slumaps/slu maps_dev.html!. Accessed 11 July 2022 Sliuzas R, Mboup G, de Sherbinin A (2008) Report of the expert group meeting on slum identification and mapping. Enschede, The Netherlands SLUMAP (2022). https://slumap.ulb.be/. Accessed 11 July 2022 Stewart ID & Oke TR (2012) Local climate zones for urban temperature studies. B Am Meteorol Soc 93(12):1879–1900. https://doi.org/10.1175/bams-d-11-00019.1 Thomson DR, Gaughan AE, Stevens FR, Yetman G, Elias P, Chen R (2021) Evaluating the accuracy of gridded population estimates in slums: a case study in Nigeria and Kenya. Urban Sci 5(2):48 UN-Habitat (2016) Slums Almanac 2015–16. Tracking improvement in the lives of slum dwellers. Nairobi, Kenya UN-Habitat (2020a) Metadata indicator 11.1.1. https://unstats.un.org/sdgs/metadata/files/Metadata11-01-01.pdf. Accessed 11 July 2022 UN-Habitat (2020b) World cities report—the value of sustainable urbanization. United Nations Human Settlements Programme, Nairobi Vanhuysse S, Georganos S, Kuffer M, Grippa T, Lennert M, Wolff E (2021) Gridded urban deprivation probability from open optical imagery and Dual-Pol Sar data. In: Proceedings of the 2021 IEEE international geoscience and remote sensing symposium IGARSS, pp 2110–2113 Wang J, Kuffer M, Roy D, Pfeffer K (2019) Deprivation pockets through the lens of convolutional neural networks. Remote Sens Environ 234:111448. https://doi.org/10.1016/j.rse.2019.111448 Wanjiru N (Producer) (2021) Community voices #1: waste management solutions. https://viceve rsaonline.nl/2021/09/10/community-voices-1-waste-management-solutions/. Accessed 11 July 2022

22

M. Kuffer et al.

Williams TKA, Wei T, Zhu X (2020) Mapping urban slum settlements using very high-resolution imagery and land boundary data. IEEE J Select Topics Appl Earth Obs Remote Sens 13:166–177. https://doi.org/10.1109/JSTARS.2019.2954407 Wuhan University (2020). gpcv.whu.edu.cn/data/building_dataset.html. Accessed 2 Mar 2020 Wurm M, Stark T, Zhu XX, Weigand M, Taubenböck H (2019) Semantic segmentation of slums in satellite images using transfer learning on fully convolutional neural networks. ISPRS J Photogramm Remote Sens 150:59–69. https://doi.org/10.1016/j.isprsjprs.2019.02.006

Chapter 2

Investigation of Ecological Sustainability Through the Landscape Approach of Geospatial Technology: Study from New Town Project in Eastern India Anirban Kundu and Sk. Mafizul Haque

Abstract The uncontrolled growth of urban population in developing nations has led to a situation of rapid alterations of peri-urban land. It immediately needs costeffective monitoring and evaluates the policy for ecological sustainability. This study tries to investigate the spatio-temporal dynamics of land transformation patterns coupled with a holistic apprehension of landscape sustainability induced by different facets of ecological configurations of different land utilization, applying modernday’s techniques from geospatial platform and landscape metrics within a multicriteria framework of Analytical Hierarchy Process (AHP) in one of the new town project areas of Easter India. Computation of Landsat 5 TM and Landsat 8 OLI images for the years 1988, 2000, 2010 and 2020 reveals that there has been a steady growth in urban environment (1.35%/year) at a cost of altering natural resources. Use of different landscape metrics like Percentage of Landscape (PLAND), Largest Patch Index (LPI), Number of patches (NP) and Mean Euclidean Nearest Neighbour Distance (ENN_MN) within a fishnet grid space of 1 km × 1 km dimension reveals the areal extension and connectivity between vegetative landscape and water bodies has decreased with a steady increase of fragmentation until 2010, whereas the situation is completely reversed for builtscapes, especially in Action Areas I, II and III of New Town Rajarhat. These results certainly depict the chances of degradation of ecological sustainability which has also been observed using a land usebased Composite Ecological Sustainability Index (LUCESI). A steady decrement in LUCESI has been observed for Action Areas-I and II during 1988–2010. Contrarily, the state authority’s rapid greening and blueing programs have pertinently improved the situation in 2020. Therefore, this study certainly proves the role of authority in cityscape planning and calls for sustainable and radical policy interventions by different stakeholders in the future to accomplish a more ecologically sustainable township planning. A. Kundu · Sk. Mafizul Haque (B) Department of Geography, University of Calcutta, 35, B. C. Road, Kolkata, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_2

23

24

A. Kundu and Sk. Mafizul Haque

Keywords Ecological sustainability · Landscape metrics · AHP · New Town · LUCESI

2.1 Introduction The rapid increase of population has commanded to a situation of accession in the global urban population which crossed the rural share. The growth in the urban population is expected to become almost 58% by 2070, while a decremental trend of 18% is expected to observe for the rural areas (Habitat 2022). Moreover, a study by Montgomery (2008) has particularly affirmed that expected population growth is mostly going to take place in urban areas of developing countries. In India, the urban population had increased by 69% after the economic liberalization in 1991 (Chadchan and Shankar 2012). Such an enormous increment in urban population certainly demands living space which has further resulted in an increase in total number of census towns and small urban cities. The share of the urban population in census towns in India has increased from 7.6% in 2001 to 14.5% in 2011 (Jain 2018). Such transformation in the demography and increment in the urbanization process has certainly led to a rapid transformation of the land use practices in urban areas, especially in newly developed urban spaces. In such a scenario, urban sprawl has become a contemporary issue of unorganized urban planning; all major metropolitans of India like Delhi (Jain et al. 2016; Sharma and Joshi 2013), Mumbai (Shafizadeh Moghadam and Helbich 2013), Kolkata (Haque 2013; Rahaman et al. 2018) and Chennai (Aithal and Ramachandra 2016; Padmanaban et al. 2017) are experiencing a concerning amount of urban expansion along in peripheral regions. All are overbounded in nature (Haque et al. 2020). This unplanned and chaotic manner of growth duly affects predominantly rural land utilization patterns of the peripheral regions of metropolitans coupled with the destruction of ecological sustainability. The sheer transformation of permeable landscapes to impervious built-up spaces has resulted in a greater degree of land fragmentation (Haque 2020) and loss of connectivity between physical landscapes. Contrarily, the 11th Sustainable Development Goal (SDG 11) has adroitly accounted for building sustainable cityscapes for its dwellers to promote better and liveable townscape planning in near future (United Nations 2016). Moreover, the 13th, 14th and 15th Sustainable Development Goals have also iterated about combatting climate-induced risks, conserving, protecting, restoring and promoting the sustainable use of ecosystems (United Nations 2016). But the unorganized process of urban sprawl in India has certainly countered all these SDGs, and therefore it is a major policy concern in contemporary times. Smart city program can be a lubricating policy in ameliorating such an organized manner of township planning from a sustainability perspective. Smart cities can be defined as cities with enhanced workability, efficiency, liveability and sustainability regarding all types of its functions. Moreover, Government of India also started a “Smart City Mission” in 2015 and selected 100 possible township projects all over India that could be established as

2 Investigation of Ecological Sustainability Through the Landscape …

25

smart cities and sanctioned a certain amount of budgetary allocations for each of them (Kumar et al. 2018). Therefore, it is extremely necessary to investigate the changes in landscape horizons coupled with comprehended ecological sustainability of all these township projects in order to generate a better idea of the performance of these new township projects. Interestingly, most of the published studies deal with the spatio-temporal dynamics of land use change in the new township projects only from a landscape perspective. But there is a gap in discourse of knowledge that tries to envisage the ecological sustainability of cities along landscape development regime. Therefore, the idea of sustainability under the paradigm of landscape dynamics should be addressed in this attempt. This study tries to enquire about the performance of a smart city, recently declared as a green city project located in the eastern part of India, i.e. New Town Rajarhat and its ecological sustainability underpinned by changes in the different aspects of landscape. The geospatial approach has been proved to be effective for determining the landscape patterns in urban areas since last four decades. Due to comparatively larger areal coverage and repetitive monitoring system, remotely sensed optical images can depict a detailed idea of the changing land use patterns in a multidate manner. Therefore, recent studies (Chanu et al. 2021; Halder et al. 2021a, b; Paul et al. 2021; Shahfahad et al. 2021) have used optical remote sensing-based classification techniques for determining the land use dynamics of urban areas in pan India. On the other hand, the study of different aspects of landscape dynamics is also a popular juxtapose to apprehending the structure, pattern and configuration of different land use practices. In this regard, landscape metrics can quantify the spatial patterns of landscape dynamics of an area, and it has been effectively used in different previous studies (Atasoy 2018; Cai et al. 2016; Kong et al. 2007; Marulli and Mallarach 2005; O’Neill et al. 1988). Interestingly, the combined imprint of the remote sensing approach and landscape-based spatial structure analysis could yield a quantified idea of ecological sustainability, but such an integrated framework is still limited in the discourse of landscape studies. The Analytical Hierarchy Process (AHP) in the large spectrum of Multi-Criteria Decision Making (MCDM) regime is considered to be a useful tool for assimilating different factors responsible for an event within a singular framework and cultivating a comprehended and holistic idea of the event (Saaty 1987). The use of such a powerful tool of AHP is very limited in the realm of landscape studies. Therefore, this study has tried to construct a novel methodological framework that would accommodate all the different facets of landscape patterns of varying land use classes within a singular quantifiable index to contemplate a detailed understanding of the spatio-temporal dynamics of ecological sustainability in the New Town Rajarhat (NTR) region within the period of 1988–2020. The main objectives of this work are as follows: a. To identify the spatial changes in the land use pattern in the NTR region, b. To determine and quantify the different aspects of landscape configurations (i.e. expansion, clustering, fragmentation and connectivity) of varying land utilization classes in the study area,

26

A. Kundu and Sk. Mafizul Haque

c. To comprehend the spatial dynamics of ecological sustainability owing to different landscape configurations having holistic apprehension of sustainability.

2.1.1 Background of the Landscape Dynamics in the Study Area The study has been conducted in the New Township project in the Rajarhat area which is situated along the north-east of the Kolkata metro core and is considered to be the first-ever self-reliant smart-city project in West Bengal (Kundu 2016). The NTR extends from 22° 30' N to 22° 38' N and 88° 26' E to 88° 34' E and comprises nearly 28 km2 area that extends within the two community development (CD) blocks, namely Rajarhat (situated in North 24 Parganas district) and Bhangar-II (located in South 24 Parganas district). Bidhannagar is located in the western part of NTR, whereas the urban centre of Barasat is located in the north. The East Kolkata Wetlands is in the south-west part of this study region. The Krisnapur Canal (locally known as the Keshtopur Khal) is flowing through the middle of the NTR region and serves as the prime sewage canal of the region. The NTR is one of the planned township projects in West Bengal, launched in the year 1993 as an alternative residential module and substitutive business centre to overcome the burgeoning pressure of the increasing population in the core of Kolkata (Kundu 2016). After a political non-compliance between the central and state government of West Bengal, it was converted into a Green City project by the state government of West Bengal and is considered to be the first-ever green city project in India (Ghosal 2016b). All the civic services and civic amenities in NTR are administered by the New Town Kolkata Development Authority (NKDA), whereas all the infrastructural, housing and other developmental activities are governed by the West Bengal Housing Infrastructure Development Corporation (WBHIDCO). From an economic standpoint, WBHIDCO has developed NTR as a Fintech hub where nearly about 25 financial institutions (including banking and non-banking) have afforded spaces till 2019 (Milleniumpost 2019). The NTR is segmented into four action areas (Action Area I, Action Area II, Action Area III and Action Area IV) and one central business district (CBD). Since its outset, largescale transportation projects have been set up in NTR to bolster the communication and accessibility of the region with the core city of Kolkata and other nearby urban centres (e.g. Madhyamgram, Barasat, Dunlop and Barahnagar). Presently, NTR is well connected with the Eastern Metropolitan Bypass and Kolkata-Basanti highway. The New Garia-Kolkata Airport metro line (Line 6) of the Kolkata Metro Railway is under construction, and it will further reinforce the accessibility of the region in the near future (Figs. 2.1 and 2.2). The NTR had been set up in an extremely rural landscape, and a large-scale rural–urban transformation of the landscape had taken place in the last 30 years in the region. The consistent increment in the impervious built-up spaces has predominantly been done with encroachment in the vegetative areas, arable cultivable areas and water bodies (Haque 2020; Haque et al. 2020; Mitra and Banerji 2018). Such an

2 Investigation of Ecological Sustainability Through the Landscape …

27

Fig. 2.1 a Relative position of the NTR region with respect to India and b Kolkata; c administrative configuration of the NTR region; d distribution of fishnets (1 km × 1 km) along the entire NTR region

28

A. Kundu and Sk. Mafizul Haque

Fig. 2.2 Conceptual framework of the adopted methods in this study

intriguing nature of landscape transformation has certainly put ecological sustainability into question. Although the Environmental Impact Assessment (EIA) report of this project claimed that the NTR will reform its green landscape through urban forestry and an extensive plantation program and will keep 30% of its total area under green cover (WBHIDCO 1999), the real situation should be investigated from an ecological sustainability context in the landscape realm. This study therefore would try to quantify the spatio-temporal dynamics of major facets of landscape patterns of different land use classes and would further contemplate a holistic understanding of ecological sustainability from a landscape regime in the NTR region.

2.2 Database and Methodology 2.2.1 Data Consideration To obtain the spatio-temporal changes in the land use pattern and its different landscape dynamics, this study has used satellite-based multispectral images, which have been obtained for the years 1988, 2000, 2010 and 2020 (Table 2.1), and extracted from the USGS, Earth Explorer repository. All images are of January because of the

2 Investigation of Ecological Sustainability Through the Landscape …

29

more possibility of cloud-free as well as seasonal variability of land use patterns has also been avoided by acquiring images of a particular time juncture. Cloud-free data from Landsat 5 TM images (for 1988, 2000 and 2010) and Landsat 8 OLI images (for 2020) have been used in the study. Both the sensors have several optical bands (6 for TM and 8 for OLI) containing the reflectance from land surface elements in different electromagnetic spectrums with varying wavelengths, which enable them to differentiate between the land surface signatures. Moreover, the repetitive monitoring system and a spatial resolution of almost 30 m of both the sensors endow them to monitor the land use pattern at a relatively broader spatial extent with more details and less time intervals. After the images were obtained, they were radiometrically calibrated and atmospherically corrected using the FLAASH module in the ENVI environment to avert the impacts of terrain roughness and atmospheric interference on radiant energy. Further analyses of the images have been discussed in the following sub-sections. Table 2.1 Details catalogue of the remote sensing images used Satellite sensor

Date

Landsat 5 TM

8th November 1988 9th November 2000 5th November 2010

Landsat 8 OLI

16th November 2020

Band configuration Band

Wavelength (µm)

Band 1

0.45–0.52

Band 2

0.52–0.60

Band 3

0.63–0.69

Band 4

0.76–0.90

Band 5

1.55–1.75

Band 6

10.40–12.50

Band 7

2.08–2.35

Band 1

0.43–0.45

Band 2

0.45–0.51

Band 3

0.53–0.59

Band 4

0.64–0.67

Band 5

0.85–0.88

Band 6

1.57–1.65

Band 7

2.11–2.29

Band 8

0.50–0.68

Band 9

1.36–1.38

Band 10

10.6–11.19

Band 11

11.50–12.51

Spatial resolution (m) 30

120 30

100

30

A. Kundu and Sk. Mafizul Haque

2.2.2 Extraction of Land Use Land Cover Features in Change Dynamics of Urban Environment For delineating the spatio-temporal variations in the land use land cover (LULC) patterns in the NTR region, firstly a supervised image classification scheme has been obtained. A supervised classification scheme predicts the class of any particular pixel based on the user-assigned training samples which the user defines with the help of his prior knowledge. Therefore, the accuracy of such a scheme remains relatively higher compared to other unsupervised schemes. Therefore, the preprocessed multispectral images were first clipped out using the vector shapefile of the study area. Thereafter, using a detailed visual interpretation and pre-acquired knowledge of the authors, 5 dominant LULC classes (e.g. Built-up area, Water body, Vegetation, Agricultural land and Barren land with sparse vegetation) have been identified in the entire NTR region. Next, an optimal amount of training samples were collected as described by Lillesand and Kiefer (1979), for each of the LULC classes to train the classification scheme. Finally, using the Maximum Likelihood Classifier (MLC) algorithm, the NTR region has been classified and the LULC classes of the region for each year have been extracted. The total number of pixels occupied by each class in each year was further tabulated and the percentage-wise areal distribution of each class was calculated using Eq. 2.1 (modified after Mitra and Banerji 2018): Percentage of area of each class Total number of pixels occupied by each class × 100 = Total number of pixels occupied by all classes

(2.1)

Moreover, the yearly rate of change in the percentage area of each LULC class was also calculated using Eq. 2.2 (after Fei and Zhao 2019): Yearly rate of change (%) of each class Area (%) of each class in final year − Area (%) of each class in initial year = Total number of years (2.2) Finally, the extracted LULC images were used for further landscape-level analysis.

2 Investigation of Ecological Sustainability Through the Landscape …

31

2.2.3 Landscape Metrics to Represent Different Aspects of Landscape Configuration in Ecological Sustainability Landscape metrics can be defined as algorithms that can quantify the spatial structure and pattern of a particular land use class within a space–time compendium. The inherent computing formula of any landscape metric enables it to determine a certain aspect of landscape configuration. In this study, we have selected four landscape metrics to represent four major facets of landscape configuration (i.e. expansion, clustering, fragmentation and connectivity) in the NTR region. A detailed discussion of the selected landscape metrics has been done in Table 2.2. Moreover, the idea of ecological sustainability depicted by any particular landscape metric varies over different LULC classes. For example, landscape fragmentation of vegetative land depicts a lesser degree of ecological sustainability whereas fragmentation of builtup areas certainly portrays a relatively higher degree of sustainability. Therefore, it is extremely important to perceive the underlying characters of each landscape metric concerning different LULC classes. Reasons for the selection of the landscape metrics with respect to the different facets of landscape configurations are as follows. The areal coverage of any landscape is the prime factor that determines the impact of the landscape on the ecological quality. Therefore, the first landscape metric selected in this regard is the Percentage of Landscape (PLAND), which helps to Table 2.2 Adopted landscape configuration and their detailed discussion Landscape configuration

Selected metric

Expansion

Percentage of Percentage of the total 0 < PLAND ≤ 100 landscape landscape comprised by (per cent) (PLAND) a particular class

Weng (2007)

Clustering

Largest patch Percentage of landscape 0 < LPI ≤ 100 index (LPI) acquired by the largest (per cent) patch of a particular class

Araya and Cabral (2010), Fei and Zhao (2019)

Fragmentation

Number of patches (NP)

Total number of patches NP ≥ 1 of a particular class (none)

Jha et al. (2005), Kamusoko and Aniya (2007)

Connectivity

Mean Euclidean nearest neighbour distance (ENN_MN)

The mean Euclidean distance of a particular land use class with its nearest neighbouring class of the same type

Magle et al. (2009), Mu et al. (2021)

Explanation

Range (unit)

ENN_MN > 0 (meters)

References

32

A. Kundu and Sk. Mafizul Haque

understand the expansion of any certain landscape within a particular area. PLAND can be described as the percentage of a certain category of landscape within a specific area and is calculated using Eq. 2.3 (McGarigal and Marks 1995): En PLAND =

j=1

ai j

A

× 100

(2.3)

where aij = the area (m2 ) of any certain LULC class and A = the total area of the landscape (m2 ). Besides the areal expansion of a certain landscape, the nature of concentration and dispersion of landscape also play a vital role in determining ecological sustainability. A densely clustered patch of water body or dense vegetative land certainly promotes a higher degree of sustainability compared to a dispersed patch. The converse situation is true for a patch of built-up area or barren land. To ascertain the landscape cluster of different LULC classes, we have applied the Largest Patch Index (LPI) metric using Eq. 2.4 (McGarigal and Marks 1995), which determines the percentage of landscape acquired by the largest patch of a particular landscape within an area: LPI =

max(ai j ) × 100 A

(2.4)

where max(aij ) = area (m2 ) of the largest patch of a certain LULC class and A = total area (m2 ) of the entire landscape. To portray the landscape fragmentation and connectivity of different LULC classes, we have selected the Number of Patches (NP) metric and Mean Euclidean Nearest Neighbour Distance (ENN_MN) metric, respectively. Both the metrics have been calculated using Eq. 2.5 (McGarigal and Marks 1995) and Eq. 2.6 (McGarigal and Marks 1995), respectively: NP = n i j

(2.5)

ENN_MN = h i j

(2.6)

where nij = total number of patches of a certain LULC class within a particular area and hij = the mean distance (m) between the centroids of two consecutive patches of a certain LULC class within an area. For determining the landscape metric of each LULC class of the individual year, first, a fishnet grid of 1 km ×1 km has been produced along the entire NTR region in the ArcGIS environment. The selection of 1 km2 grid space lies in the fact that it measures all landscape metrics within a unit areal space. Next, the LULC map of the respective year has been extracted using the fishnet gridding vector file, and a batch file consisting of all the spatial information of all the 1 km2 raster images of the LULC map has been created in the Python environment. Thereafter, the batch file is loaded in FRAGSTAT 4 environment and the selected landscape metrics for

2 Investigation of Ecological Sustainability Through the Landscape …

33

all the classes were calculated and tabulated in MS Excel. Finally, the MS Excel file containing landscape metrics values of all the metrics is incorporated again in the ArcGIS environment and using an interpolation method (IDW), landscape metric maps of all the LULC classes for the respective year have been created.

2.2.4 Application of Fuzzy Membership Functions for Normalization of Landscape Metric Rasters The concept of ecological sustainability perceived from any particular landscape metric of a particular LULC class completely depends on the author’s cognition and judgement. Therefore, it is quite difficult for the author to set up some threshold values as a limit for considering the degree of sustainability. In such a case, the use of fuzzy logic and fuzzy membership functions, propounded by Zadeh (2013), are quite effective as these functions consider all the possible values of a particular variable in the classification scheme and are thereby independent of any threshold value. Moreover, the fuzzy membership functions limit the values of the landscape metrics within a range of 0–1, hence, it also serves as a normalization technique for the landscape metric maps. After acquiring the landscape metric maps for all the LULC classes of all years, fuzzy membership functions have been applied to all the landscape metrics. After a detailed review of the literature, two fuzzy membership functions (i.e. small and large) have been selected to fuzzify the landscape metric maps. The small fuzzy membership function tends to incorporate the smaller values of a particular set within the transformed membership sets, whereas the large function tends to include the larger values. After careful consideration and the efficacy of a particular landscape metric of a particular LULC class to identify ecological sustainability, the fuzzy membership functions for all the landscape metric maps for all the LULC classes were selected. The details of the fuzzy membership functions have been discussed in Table 2.3.

2.2.5 Application of Analytical Hierarchy Process (AHP) for Determining Individual Land Use-Based and Composite Land Use-Based Ecological Sustainability The landscape metric maps of each LULC class for a specific year illustrate the different facets of landscape configuration but for getting a concrete idea of ecological sustainability due to the landscape structure of each LULC class, it is important to combine all facets of landscape structures within a unified framework. Moreover, the perception of ecological sustainability generated from each LULC class should also be coalesced for a holistic and composite idea of sustainability in the

34

A. Kundu and Sk. Mafizul Haque

Table 2.3 Selected fuzzy membership function for different landscape metrics of different LULC classes LULC class

Nature of selected fuzzy membership Percentage of landscape (PLAND)

Largest patch index (LPI)

Number of patches (NP)

Mean Euclidean nearest neighbour distance (ENN_ MN)

Built-up area

Small

Small

Large

Large

Water body

Large

Large

Small

Small

Vegetation

Large

Large

Small

Small

Agricultural land

Large

Large

Small

Small

Barren land with sparse vegetation

Small

Small

Large

Large

NTR region. A multi-criteria decision framework is an important tool in such cases and the AHP, formulated by Saaty (1987), has been considered in this study. The AHP framework incorporates all the causative factors that drive any event within a comparative prioritization scheme and using the relative importance scale by R. W. Saaty in 1987, it computes the relative weightage of the factors. Hence, AHP is a famous tool for distinguishing the spatial dynamics of events and has been applied in different studies. The AHP process has been implemented through two stages in this study. First, for determining the relative weightage of the landscape metrics in their ability to depict ecological sustainability, the four selected landscape metrics have been arranged in a 4 × 4 matrix and the relative importance of each metric with respect to others has been determined using Table 2.4 and the consistency ratio for the weightage scheme has also been calculated (using Table 2.5) to decide the acceptance of the weightage scheme. After acquiring the relative weights, all the fuzzified landscape metric rasters of a particular LULC class of a particular year have been multiplied with their respective weights using the raster calculator tool in ArcGIS and weighted fuzzy landscape metric rasters have been created. Finally, all the weighted fuzzy landscape metric rasters are aggregated using Eq. 2.7 to compute land use-based ecological sustainability index [modified after the Additive Ratio Assessment (ARAS) method propounded by Zavadskas and Turskis (2010)]: ) ( ) ( ESIx = wPLAND × PLANDfuzzy + wLPI × LPIfuzzy ( ) ( ) + wNP × NPfuzzy + wENN_MN × ENN_MNfuzzy

(2.7)

where ESIx refers to the Ecological Sustainability Index for LULC class x; wPLAND , wLPI , wNP , wENN_MN represent the relative weightage of PLAND, LPI, NP and ENN_ MN respectively and PLANDfuzzy , LPIfuzzy , NPfuzzy , ENN_MNfuzzy refer to fuzzified

2 Investigation of Ecological Sustainability Through the Landscape … Table 2.4 Scale of relative importance of one parameter over others, after Saaty (1987)

35

Intensity

Description

1

Equal importance

3

Moderate importance

5

Strong importance

7

Very strong importance

9

Extreme importance

2, 4, 6, 8

Intermediate values between two adjacent judgements

Table 2.5 Random index table propounded by Saaty (1987) n

1

2

3

4

5

6

7

8

9

10

Random index

0.00

0.00

0.58

0.90

1.12

1.24

1.32

1.41

1.45

1.49

rasters of PLAND, LPI, NP and ENN_MN, respectively. The same method for all LULC classes and all years is then repeated. After acquiring the land use-based ecological sustainability indices (ESIx ) of a particular year, the final aim of this paper is to combine all the ESIx ’s to contemplate a comprehensive idea of the land use-based composite ecological sustainability index (LUCESI). In such context, AHP has been used for the second time and the relative weightage of all the LULC classes with respect to their importance in perceiving ecological sustainability has been calculated using Table 2.4 and the consistency of the weightage scheme has also been calculated using Table 2.5. After depicting the weights, all the land use-based ecological sustainability indices were normalized following Eq. 2.8 and each weighted normalized index was multiplied by their respective weightage: ESInormalized =

ESIactual − ESImin ESImax − ESImin

(2.8)

where ESInormalised = normalized land use-based ecological sustainability index, ESIactual = pixel value of raster of any land use-based ecological sustainability index, ESImax = maximum value of all the pixels of the raster of any land use-based ecological sustainability index and ESImin = minimum value of all the pixels of the raster of any land use-based ecological sustainability index. Finally, all the normalized ESIx ’s of the respective year are combined using Eq. 2.9 [modified after the Additive Ratio Assessment (ARAS) method propounded by Zavadskas and Turskis (2010)] and a final Land Use-based Ecological Sustainability Index (LUCESI) for a particular year has been apprehended: ( ) ( ) LUCESI = wbuiltup area × BUAESI + wwater body × WBESI ( ) ( ) + wdense vegetation × VESI + wagricultural land × AGLESI

36

A. Kundu and Sk. Mafizul Haque

+ (wbarren land × BALESI)

(2.9)

where LUCESI refers to the Land Use-based Ecological Sustainability Index; wbuiltup area , wwater body , wdense vegetation , wagricultural land and wbarren land represent the relative weights of built-up area, water body, dense vegetation, agricultural land and barren land respectively, and BUAESI, WBESI, WESI, AGLESI and BALESI depict the ecological sustainability indices based of varying LULC classes (i.e. built-up area, water body, dense vegetation, agricultural land and barren land, respectively). The same steps have been incorporated for all the years.

2.2.6 Validation of the Proposed LUCESI Model The proposed LUCESI model is a novel, comprehended, quantifiable index that can discern the holistic idea of ecological sustainability from a landscape regime. But it is extremely necessary to contemplate the robustness, accuracy, usefulness and predictive ability of the LUCESI model. Therefore, validation of the proposed LUCESI model is important. In such a context, we have used the Normalized Difference Vegetation Index (NDVI) as the actual predictor data. NDVI can delineate the quality and health of vegetation in a particular satellite image pixel. It is a wellestablished model for spatially predicting ecological characteristics and has been effectively used in different studies (Li et al. 2017; Ma et al. 2020; Pettorelli et al. 2005; Zhang et al. 2012). NDVI has been calculated using Eq. 2.10, after Rouse et al. (1973): NDVI =

NIR − RED NIR + RED

(2.10)

where NIR = spectral reflectance in the Near-infrared region and RED = spectral reflectance in the visible red region. Therefore, a good association between the NDVI and LUCESI can certainly depict the robustness of the LUCESI model. For validation of the LUCESI model, first, the NDVI images of all the years have been created in the ArcGIS environment. Next, 3995 sample points at a regular interval of 100 m have been generated along the entire NTR region, and the NDVI value of each point of a particular year and corresponding LUCESI value have been extracted and a scatterplot between NDVI and LUCESI of all the points has been created to depict the overall association between the two. For a more statistical explanation, we have used the Chi-square test of association between NDVI and LUCESI. The Chi-square test yields the comparison between the variance of simulated and actual output (Ringuest 1986) and thus predicts whether a significant association between the two is present or not. Therefore, this is a powerful yet simple tool for validating the simulated data. For computing the association between NDVI and LUCESI, the NDVI and LUCESI of rasters of each year have been reclassified into 5 classes (Very High, High, Moderate, Low and Very Low) in ArcGIS and the classes

2 Investigation of Ecological Sustainability Through the Landscape …

37

corresponding to the 3995 points were extracted. Points constituting NoData values (−9999) were dropped off. Finally, the data was loaded in the STATA environment and using the Chi-square test, we have comprehended whether the association between NDVI and LUCESI is significant or not. All the steps were reiterated for all the consecutive years.

2.3 Results and Analysis 2.3.1 Spatio-temporal Dynamics of Land Use and Land Cover in the NTR Region The spatial distribution of LULC classes in the NTR region depicts a steady change in the land use classes, especially in the impervious built-up areas in the last 32 years (Fig. 2.3). In the year 1988, the New Township project was not started, hence, the primary landscape characteristics of the water environment can be firmly portrayed in the image. At this time, the NTR region comprises mostly vegetation patches (35.55%) and agricultural land (27.83%), which together constituted more than 50% of the NTR region. Even, the water bodies also contained 12.64% of the total area. The built-up areas were bounded in the southern (presently in Action Area IV) portion due to the dominance of agriculture practices. The concentration of water bodies and vegetative lands was mostly noticed in the northern portion in Action Area II and north-western portion in Action Area I. Subsequently, the landscape characteristics have changed with time, mostly at the cost of a reduction in natural resources. In 2000, the New Township project was just started, and activities of vegetation destruction and water body encroachment started to convert the piece of land for suitable construction purposes. Therefore, the maximum concentration of bare surface with sparse shrubs (45.35%) was attributed during this time in the NTR region. The construction of builtup areas continued after 2000 at a faster rate, and a striking increment in the built-up areas has been observed in this region in 2010 (at a rate of 1.37% year−1 ) and 2020 (at a rate of 3.20% year−1 ), in almost all parts of the study area. A prominent decrement in water bodies (0.12% year−1 ), vegetation patch (1.39% year−1 ) and agricultural land (1.23% year−1 ) has also been observed during 2000–2010. Interestingly, there has been an increment in the areal coverage of vegetative land (1.64% year−1 ) and water bodies (0.08% year−1 ) during 2010–2020. This can be attributed to the shift in the focal planning paradigm of the region (i.e. from Smart City to Green City). As the barren lands created for construction purposes were converted into built-up areas, the amount of this land use decreased by almost 37.94% (at a rate of 3.79% year−1 ) during this decade. The overall changes in the land use classes in the NTR region show that the builtup areas have increased the most (at a rate of 1.39% year−1 ) in 32 years, whereas all the other LULC classes have shown an overall decrement in their areal coverage. This

38

A. Kundu and Sk. Mafizul Haque

Fig. 2.3 Distribution of land use and land cover classes in the NTR region for the years 1988, 2000, 2010 and 2020

2 Investigation of Ecological Sustainability Through the Landscape …

39

decremental trend is most prominent in agricultural land use (R2 = 0.83) followed by water bodies (R2 = 0.68) and vegetation patches (R2 = 0.34), whereas the incremental trend of built-up areas is also quite prominent (R2 = 0.76) (Figs. 2.4 and 2.5). The spatial trend of changes in the LULC classes in different Action Areas of the NTR region (Fig. 2.6) shows that the maximum increment of built-up areas has occurred in Action Area I, whereas the change is nearly the same along Action Area II and Action Area III (Fig. 2.6a). In the case of water body distribution, its amount decreased for all Action Areas except Action Area II where a gradual increment has been observed during 2000–2020 (Fig. 2.6b). The trend of changes in the vegetative patch (Fig. 2.6c) and agricultural land (Fig. 2.6d) is almost the same for Action Areas I, II and III, whereas Action Area IV depicts an increase in both the land use during 2010–2020.

Fig. 2.4 Trend of change in the LULC classes in the NTR region during 1988–2020

Fig. 2.5 Boxplot of rate of change of different land use classes

40

A. Kundu and Sk. Mafizul Haque

Fig. 2.6 Areal distribution (%) of a built-up area, b water body, c vegetation, d agricultural land and e barren land with sparse vegetation in different Action Areas in the NTR region during 1988, 2000, 2010 and 2020

2 Investigation of Ecological Sustainability Through the Landscape …

41

2.3.2 Spatio-temporal Distribution of Different Facets of Landscape in Ecological Sustainability After the spatio-temporal analysis of the land use pattern of the NTR region, the configuration and structure of the landscape of different LULC classes have been depicted using 4 landscape metrics that try to contemplate 4 different facets of landscape dynamics of the region. These facets of landscape patterns enable us to synthesize the underlying attributes of ecological sustainability. Hence, using the methodology described in Rouse et al. (1973, Sect. 3.2), the selected metrics have been apprehended. The spatio-temporal distribution of PLAND of built-up area in 1988 shows a higher (14.01%) value in the southern part of the NTR region. This trend completely shifted to the north-western part with a comparatively lower value (9.18%) of PLAND in 2000. After 2000, the urbanization in the NTR region increased significantly and the PLAND value of the built-up areas increased rapidly with the highest value of 43.41% and 87.67% in 2010 and 2020, respectively. This transformation in the builtup areas has been done with encroachment on the water bodies, vegetative lands and agricultural lands, and this trend has also been reflected in the PLAND distribution of the 3 land use classes. The north-western part of NTR had a higher PLAND value of water body (40.07%) and vegetative land (71.76%) in 1988, which decreased to 36.15% and 62.71% respectively in 2000, 30.96% and 34.83% in 2010. Moreover, the spatial dynamics of PLAND of water bodies and dense vegetative lands also changed within this period (Figs. 2.7, 2.8, and 2.9). Interestingly, the PLAND values of water bodies and vegetative land increased to a small extent in the year 2020 (Fig. 2.10). Like PLAND, the spatial distribution of LPI of built-up areas in the NTR region shows a higher value (11.33%) in the southern part in 1988, whereas the trend shifted towards the north-western part in 2000 with a value of 5.56%. After 2000, the LPI values for built-up areas increased and became almost 40.73 and 87.67% in 2010 and 2020. The spatial trend of higher LPI of built-up areas also expands from the north-eastern to the central, northern, eastern and western parts of the NTR region in 2010 and 2020. This increment in LPI values of built-up areas certainly proves that the tendency of unit area of landscape occupied by only built-up classes has increased which further promotes the fact of destruction of ecological richness in the region. This decremental tendency of ecological richness has also been observed in the spatio-temporal dynamics of the LPI values of water bodies, vegetation and agricultural lands. The higher LPI of water bodies (31.66%) in the year 1988 decreased in 2000 with the highest LPI value of 27.76%, although this trend reversed in 2010 and 2020 when the highest LPI of water bodies increased to 29.92% and 38.22%, respectively. In the case of vegetative lands, the highest LPI value was 67.07% in 1988 which decreased to 61.67% and 30.84% in 2000 and 2010, respectively. The highest LPI of vegetative lands also increased to 58.58% in 2020. The higher degree of LPI of agricultural lands was mostly concentrated in the southern and western parts of NTR in 1988 and 2000, whereas this spatial trend was limited to only the

42

A. Kundu and Sk. Mafizul Haque

Fig. 2.7 Spatial distribution of different landscape metrics of varying LULC classes of the year 1988 (Each column represents a particular landscape metric, and each row represents a particular LULC class.)

2 Investigation of Ecological Sustainability Through the Landscape …

43

Fig. 2.8 Spatial distribution of different landscape metrics of varying LULC classes of the year 2000 (Each column represents a particular landscape metric, and each row represents a particular LULC class.)

44

A. Kundu and Sk. Mafizul Haque

Fig. 2.9 Spatial distribution of different landscape metrics of varying LULC classes of the year 2010 (Each column represents a particular landscape metric, and each row represents a particular LULC class.)

2 Investigation of Ecological Sustainability Through the Landscape …

45

Fig. 2.10 Spatial distribution of different landscape metrics of varying LULC classes of the year 2020 (Each column represents a particular landscape metric, and each row represents a particular LULC class.)

46

A. Kundu and Sk. Mafizul Haque

southern part after 2000. The higher degree of LPI distribution of barren land with sparse vegetation was mainly concentrated along the southern part of the NTR region in 1988, whereas the trend shifted towards the north-western part due to land transformation into barren lands for further land preparation for construction purposes. After 2010, the higher LPI of barren land also decreased as most of the construction in the barren lands had been completed and had been transformed into built-up areas. After comprehending the landscape expansion and landscape clustering of different LULC classes, it is important to understand the degree of landscape fragmentation and landscape connectivity of the LULC classes for a holistic understanding of the ecological sustainability of the NTR region. Land fragmentation depicts the breaking up of larger patches of landscapes into smaller ones, whereas connectivity of landscape reflects the degree to which energy flow is maintained within two consecutive landscape patches. These two facets are important for determining ecological sustainability as a larger landscape patch of water body, vegetation and agricultural land certainly promotes a higher degree of species richness as well as higher connectivity between the patches depicting a higher degree of the flow of energy, nutrients and species. On the contrary, the converse logic is true for built-up patches and barren land patches. The high NP of built-up areas in 1988 mostly was concentrated along the southern part of the NTR region as the concentration of major built-up areas was along this part. As the New Township was started by 2000, the higher NP values shifted towards the north-western part. No ENN_MN values were generated for these two years for builtup areas due to the very low concentration of built-up patches. Although the higher NP values were concentrated along the north-western part in 2010, also the lower ENN_MN was concentrated in this area. This certainly proves the new construction (although fragmented) of built-up areas along this part. This depiction of continuous expansion of built-up areas was more profound in the year 2020, where the concentration of low NP and ENN_MN values were along the north-western and central parts of the region. The spatio-temporal distribution of NP and ENN_MN of water bodies and vegetative lands shows a tendency for the destruction of ecological sustainability in the NTR region. The lower degree of NP and ENN_MN of water bodies in the north-western part of the region in 1988 transformed into a very high degree of NP and ENN_MN in the years 2000 and 2010. Although there has been a moderatelow degree of NP of water bodies in the north-western part in 2020, the ENN_MN remains higher. In the case of vegetative lands and agricultural lands, the situation is almost the same as in water bodies. The north-western, central and northern parts of the NTR region show a lesser degree of fragmentation and a higher degree of connectivity between vegetative lands in 1988, whereas the situation completely reversed during 2000 and 2010. Despite this fact, there is a certain improvement in the situation by 2020.

2 Investigation of Ecological Sustainability Through the Landscape …

47

2.3.3 Spatio-temporal Distribution of Individual Land Use-Based Ecological Sustainability and Composite Land Use-Based Ecological Sustainability The spatio-temporal dynamics of different landscape metrics of different LULC classes certainly interpret the different landscape characteristics and structures, which, in turn, has also provided the varying facets of ecological sustainability spawned by individual land use classes in the NTR region. Therefore, it is important to combine these facets to engender a comprehended idea of ecological sustainability caused by different LULC classes. Moreover, the idea of sustainability established from individual LULC classes should also be integrated in order to develop an overall exhaustive perception of sustainability in the region. Therefore, we have applied the multi-criteria framework of the Analytical Hierarchy Process (AHP), in this regard. The core methodology of AHP ascertains the relative weightage of different parameters responsible for a certain event using a relative importance score. After a detailed discussion with experts from the field of ecology and regional planning, we computed the relative weightage of the different landscape metrics based on their ability to reciprocate ecological sustainability. Accordingly, the PLAND depicting landscape expansion obtained the highest weightage of 0.54, followed by LPI (0.26), NP (0.12) and ENN_MN (0.07). The consistency ratio in this weightage scheme is 0.033 which can be considered as a good weightage scheme. After acquiring the relative weights, the next step is to combine the weighted landscape metrics into a singular additive framework for depicting individual land use-based ecological sustainability indices. But the major problem in this regard is the metrics constitute different units, and it is quite difficult and vague to consider a threshold value for any of the metrics to classify them as better or worse representatives of ecological sustainability. Therefore, we have used the fuzzy membership functions for all the metrics which makes them unit free as well as considers all the metric values within the set (limiting their value from 0 to 1) based on their relation to ecological sustainability. Metrics having positive relation with ecological sustainability were fuzzified as “Large” members and metrics with negative relation were fuzzified as “Small” members. After fuzzification, the weightage of each landscape metric was multiplied with the respective fuzzified raster, and a weighted fuzzified landscape metric raster has been generated. Finally, all the weighted fuzzified raster of the 4 landscape metrics (PLAND, LPI, NP and ENN_MN) of each LULC class were summed up to develop Built-up Area-based Ecological Sustainability Index (BUAESI), Water Body-based Ecological Sustainability Index (WBESI), Vegetation-based Ecological Sustainability Index (VESI), Agricultural Land-based Ecological Sustainability Index (AGLESI) and Barren Land-based Ecological Sustainability Index (BALESI) respectively for successive years (Fig. 2.11; Table 2.6). The distribution of BUAESI in 1988 shows a low sustainability score along the southern part of the NTR region, whereas a small patch of low BUAESI started to grow along the north-western part from 2000 and it expands along the Action Area II in 2010 and finally, it expands upto Action Area III in 2020. The distribution of

48

A. Kundu and Sk. Mafizul Haque

Fig. 2.11 Individual land use-based ecological sustainability index and their spatio-temporal distribution

2 Investigation of Ecological Sustainability Through the Landscape …

49

Table 2.6 Computation of relative weightage of different landscape metrics using AHP, after Saaty (1987)

Expansion (PLAND)

Expansion (PLAND)

Clustering (LPI)

Fragmentation (NP)

Connectivity (ENN_MN)

Relative weightage

1

3

4

6

0.54488

Clustering (LPI) 0.333333

1

3

4

0.26453

Fragmentation (NP)

0.25

0.333333

1

2

0.11951

Connectivity (ENN_MN)

0.166667

0.25

0.5

1

0.07106

higher WBESI was along Action Area I in 1988. The WBESI started to decrease for all the parts of NTR after 1988 and higher WBESI values become more scarce and left as small patches along Action Area III in 2000 and along Action Area II in 2010 and 2020. In the case of VESI, Action Area I and Action Area II constitute higher VESI values in 1988 and 2000, whereas the scenario completely changes in 2010 and most of the parts of NTR come under low VESI classes in 2010 and 2020 except a small patch of higher VESI in the east of Action Area III in 2010 and the east of Action Area II in 2020. The AGLESI is mostly low for all the years in the NTR region and higher AGLESI values were concentrated as patches in the southern part of Action Area IV and the central part of Action Area III in 1988. AGLESI along this central part started to decrease in 2000, and it comes under low AGLESI in 2010 and 2020. Finally, the BALESI were high for most of the part of NTR in 1988 which started to decrease by 2000 mostly along Action Area I and Action Area II in 2000 and 2010. In the year 2020, the BALESI becomes higher again for most of the region, but this increment does not depict any betterment of ecological sustainability rather the conversion of barren lands to built-up areas is responsible for this change. After depicting the individual land use-based ecological sustainability indices for individual years, the final goal of this study is to combine all these indices within a singular framework to comprehend a Land use-based Composite Ecological Sustainability Index (LUCESI). To apprehend the LUCESI, all the individual land use-based sustainability indices (BUAESI, WBESI, VESI, AGLESI and BALESI) were given relative weightage based on their importance in depicting ecological sustainability, using the AHP process. In this regard, VESI was given the highest priority (0.46) as vegetative lands constitute the most ecological richness and are the prime agent of ecological sustainability. The WBESI was given the second-highest weightage (0.29) followed by agricultural land (0.14), barren lands (0.07) and built-up areas (0.04). The detailed weightage scheme has been discussed in Table 2.7. Afterward using Eqs. 2.8 and 2.9, the final LUCESI values were attributed. The spatial distribution of LUCESI in 1988 depicts a comparatively higher distribution of LUCESI values along Action Area I and Action Area II. All the landmark areas in Action Area I (e.g. Mahishghot, Koch Pukur Biswa Bangla Gate) and Action Area II (e.g. Eco Park, Hatiara, Noapara, Gopalpur and Rekjuani) belong to the Very

50

A. Kundu and Sk. Mafizul Haque

Table 2.7 Computation of relative weightage of different land use-based ecological sustainability indices using AHP, after Saaty (1987) VESI

WBESI

AGLESI

BALESI

BUAESI

Relative weightage

VESI

1

2

4

7

8

0.462128

WBESI

0.5

1

3

5

6

0.291843

AGLESI

0.25

0.333333

1

3

5

0.142524

BALESI

0.142857

0.2

0.333333

1

3

0.066941

BUAESI

0.125

0.166667

0.2

0.333333

1

0.036563

high, High LUCESI classes in this time. The southern part of the NTR region (Action Area III and Action Area IV) belong to the Very low, Low LUCESI classes in 1988. The situation completely changes in 2000, when most of the part of Action Area I belonged to the Very Low, Low LUCESI classes, and the very small patch of the Moderate LUCESI class in 1988 along the northern part of Action Area II expanded extremely and a major part of this region belonged to Low LUCESI class with some patches of Very Low class along Hatiara, Noapara, Gopalpur and Eco Park. The situation in Action Area III remained almost the same as the year 1988, and Action Area IV gained some area under Moderate High LUCESI classes. The expansion of the Very Low–Low LUCESI classes along Action Area I and Action Area II continued rapidly in 2010, and more than 90% of the area in these two administrative units goes under the Very Low LUCESI class. A very small patch of Moderate, High LUCESI class remained in the western part of Action Area II. The situation for Action Area III again remained the same in 2010 and Action Area IV gained more spaces compared to 2000 in the Moderate–High LUCESI class, mostly along Suk Pukuria and Hatisala. In 2020, the situation does not change for Action Area I, whereas a major part of Action Area II goes under the Very High–High LUCESI classes in this year, mostly along Eco Park and Kadampukur. Action Area IV gained more areas in the Moderate–High LUCESI classes, and the Low LUCESI class along the southern part of the NTR region completely transformed into Moderate–Low LUCESI classes in 2020.

2.3.4 Measurement of Association Between LUCESI and NDVI For validating the LUCESI model, we have calculated the NDVI values for consecutive years using Eq. 2.10 and following the methods described in Rouse et al. (1973, Sect. 3.6), we have employed the scatterplots between the NDVI and LUCESI of consecutive years. In this regard, LUCESI has been attributed on the x-axis, whereas NDVI has been depicted on the y-axis. Figure 2.13 has adroitly shown that there is a positive relationship of varying nature between NDVI and LUCESI in consecutive years. Therefore, the efficiency of LUCESI, in representing ecological sustainability, has been established from the scatterplots. It shows the development of low to very

2 Investigation of Ecological Sustainability Through the Landscape …

51

Table 2.8 Cross-tabulation of NDVI and LUCESI with calculated Chi-square and p-value of consecutive years LUCESI (1988)

NDVI (1988) Very low

Low

Moderate

High

Very high

Total

Very low

103

141

128

204

127

703

Low

383

260

151

125

127

1046

Moderate

70

134

159

220

124

707

High

51

171

235

320

257

1034

8

28

103

213

111

463

615

734

776

1082

746

3953

High

Very high

Total

Very high Total

Pearson Chi square = 747.9555 p-value = 0.000 LUCESI (2000)

NDVI (2000) Very low

Very low

87

Low 220

Moderate 402

273

95

1077

Low

57

236

222

149

100

764

Moderate

69

314

284

315

149

1131

High

47

152

168

225

118

710

Very high Total

4

55

85

88

40

272

264

977

1161

1050

502

3954

Moderate

High

Very high

Total

Pearson Chi square = 131.9554 p-value = 0.000 LUCESI (2010)

NDVI (2010) Very low

Low

Very low

153

227

403

477

284

1544

Low

109

214

307

391

241

1262

Moderate

24

70

154

169

157

574

High

11

60

81

72

81

305

Very high

12

56

47

63

91

269

309

627

992

1172

854

3954

High

Very high

Total

153

1472 1023

Total

Pearson Chi square = 103.1789 p-value = 0.000 LUCESI (2020)

NDVI (2020) Very low

Very low

48

Low 459

Moderate 425

387

Low

24

154

239

416

190

Moderate

13

100

206

219

116

654

High

21

76

148

122

79

446 (continued)

52

A. Kundu and Sk. Mafizul Haque

Table 2.8 (continued) LUCESI (1988)

NDVI (1988) Very low

Very high Total

Low

Moderate

High

Very high

Total

22

31

58

110

127

348

128

820

1076

1254

665

3943

Pearson Chi square = 340.0398 p-value = 0.000

low sustainability of land utilization practices over time. For a more detailed and statistically significant association between the two, we have applied the Chi-square test of association. In this test, the null hypothesis (H 0 ) signifies the lack of a significant association, and the alternate hypothesis (H a ) describes the existence of a significant association between NDVI and LUCESI. The results of the Chi-square test described in Table 2.8 certainly prove that the LUCESI and NDVI have a significant association at a confidence level of 99% for all the years, and hence the alternate hypothesis has been accepted for all the years (p < 0.01). Thus, it has been affirmed that the LUCESI is indeed a robust and exhaustive model that can depict ecological sustainability and its spatio-temporal dynamics from a landscape realm, and it can be used as a generalized model for peri-urban sustainability depiction, especially in third-world countries (Fig. 2.12).

2.4 Discussion The NTR is distinctively important due to its capacity for the perspective of selfreliance and India’s first-ever zero-energy township project. Despite this, the NTR project is not devoid of controversy in terms of land acquisition, institutional negligence and ecological sustainability. Dey et al. (2016) in their book entitled “Beyond Kolkata” have adroitly mentioned the political disputes and incompetencies regarding the land acquisition process during the inception period of NTR in the 1990s and criticized the conversion of high-yielded arable lands and water bodies into impervious built-up areas. Moreover, the Master Plan of the NTR project has gone under several modifications, and the limit of the maximum residential plots in the entire NTR region has increased in consecutive Master Plans (Dey et al. 2016). The selfsufficient local habitants of the region had lost their prime livelihood of agriculture and fishing, were evicted from their ancestral lands and were kind of forced to opt for less dignified jobs of rickshaw pullers, rag pickers and household helps in the newly built residential projects. On the other hand, Kundu (2016) and Chowhan et al. (2022) have also mentioned that the NTR project becomes an exclusive living space for highincome group (HIG) people, and there lacks a sense of neighbourhood space among the HIG people living in the core residential projects in the area and the low-income group (LIG) people living in the service villages (villages kept within the residential

2 Investigation of Ecological Sustainability Through the Landscape …

53

Fig. 2.12 Spatio-temporal distribution of land use-based composite ecological sustainability index (LUCESI) for the years 1988, 2000, 2010 and 2020 in the NTR region

54

A. Kundu and Sk. Mafizul Haque

Fig. 2.13 Development of low sustainable land in relation to NDVI and LUCESI in consecutive years

projects as a habitat for the native people). Although such critical apprehensions in terms of socio-economic-political perspectives of the NTR projects had been done in previous studies, there exist only a few studies that deal with the rhetorical aspect of ecological sustainability. Therefore, it is important to look after the spatio-temporal dynamics of ecological sustainability from the landscape regime, and contemplate a comprehended idea of sustainability in the NTR region, which has been tried to be envisaged through this study. The changing dynamics of LULC classes in the NTR region reflect a typical feature of the evolution of a new township project. Before its inception, the NTR region was predominantly constituted by dense vegetative lands, water bodies and highly fertile arable lands, and such a situation is also attributed to Fig. 2.3. There was hardly any major cluster of impervious surface in the region, and the barren landscape was mostly fragmented which in turn has contributed to almost 17.84% of “Very High” and 26.30% of “High” ecologically sustainable landscapes at this time juncture. The locational advantage of the region triggered the township building process in the region, and the township was primarily planned as a residential township project keeping the environmental sustainability of the region. But the neoliberal attitude of the then left-front government soon modified the outlook of the project to an economic perspective which provoked the agglomeration of different IT sectors in this region, and different other sectors of economy also tried to emerge (Chowhan et al. 2022;

2 Investigation of Ecological Sustainability Through the Landscape …

55

Dey et al. 2016; Haque 2013). Such an economic concentration typically demanded a higher amount of land which was adequate in the region. Therefore, the process of land acquisition started, and all the major physical landscapes were converted into barren lands. In this regard, a higher amount of vegetative land and water bodies were encroached and were converted into barren lands, especially in Action Area II, also corroborated in the LULC image of 2000 in Fig. 2.14. Moreover, a large number of agricultural lands were also acquired from the landowners and were converted into barren lands in Action Area III, also reflected in Fig. 2.3. Although there exist major controversies about the land acquisition process (WBHIDCO claimed that the farmers gave their lands willingly and the arable lands which were acquired were less fertile, while the Rajarhat Jomi Bachao Committee claimed something different) (Dey et al. 2016), it is quite prominent from Fig. 2.8 that there has been an extremely low areal coverage, low cluster, low connectivity and high fragmentation of the dense vegetation, water bodies and agricultural lands in Action Areas II and III during 2000, which further had resulted to a very low degree of WBESI, VESI and AGLESI in those parts of NTR. Additionally, the conversion of lands into barren lands resulted in a higher degree of landscape expansion, cluster and connectivity with a lesser degree of landscape fragmentation of barren lands in 2000, which emanated a lesser degree of BALESI, especially in Action Area II. As the township project was almost in its inception phase in 2000, the PLAND and LPI of built-up areas were comparatively lower, except in Action Area I where a considerable amount of residential projects was started by that time. Such an immense land use transformation had critically affected the overall ecological sustainability, and there had been a 4.56% and 8.51% decrement in the “Very High” and “High” ecological sustainable classes (LUCESI) respectively by 2000, whereas an increment of 9.17% of areas under “Very Low” LUCESI class in the year 2010. The subsequent advent of the township project resulted in an enormous increment (13.68%) of built-up areas especially along Action Areas I and III in 2010. Such rapid increment in built-up areas has also been earlier corroborated in different studies (Bera and Das Chatterjee 2019; Dhali et al. 2019; Halder et al. 2021a, b; Paul 2012). This rapid increment resulted in a higher degree of PLAND and LPI of built-up classes in those parts. Conversely, as the NTR project was not completely in its mature stage, the built-up areas were quite fragmented with a lesser degree of landscape connectivity. Due to such augmentation in built-up areas, the BUAESI became moderate–low along Action Area I and two major islands of low BUAESI class concentrated in Action Area III in 2010. As two major water reservoirs (in Eco Park and Eco Urban Village) were created during this time in Action Area II, the overall amount of water bodies remained almost the same in 2010 compared to 2000 and a small patch of higher WBESI also emerged along the Action Area II by this time. But the transformation of vegetative land (13.95% decrement) and agricultural lands (12.32% decrement) into barren lands (13.78% increment) continued in 2010 also, mostly along Action Areas II and III which led to a low degree of VESI and AGLESI in these two areas. As a whole, the LUCESI values were also decreased for most of the parts of NTR, and there has been a 0.33% and 10.14% decrement in the areal coverage of “Very High” and “High” LUCESI classes respectively with a

56

A. Kundu and Sk. Mafizul Haque

Fig. 2.14 Percentage-wise areal coverage of different LUCESI classes for the years 1988, 2000, 2010 and 2020 in the NTR region

steady increase of 11.82% and 12.83% in “Very Low” and “Low” LUCESI classes in 2010. Such extensive increment in the impervious urban surface at the cost of transformation of built landscapes certainly augments the tendency to a situation of ecological and environmental unsustainability. Studies by Halder et al. (2021a, b) and Dutta et al. (2021) have observed a rapid decrease in the Normalized Difference Vegetation Index (NDVI) (an index for determining vegetative health) and an increase in the Normalized Difference Built-up Index (NDBI) (index for determining the degree of imperviousness) along NTR region within the two decades of the 1990s and 2000s. Such a deteriorating degree of land surface quality and landscape transformation promote an increase in the Land Surface Temperature (LST), which has also been corroborated by Dhar et al. (2019) as well as a certain degree of Urban Heat Island (UHI) effect has also been perceived by Dhar et al. (2019), Halder et al. (2021a, b) and Dutta et al. (2021) in the NTR region. Such land surface environmental deterioration assuredly affects the ecological sustainability of the NTR region, which has also been reflected in the LUCESI distribution in this study, although, from a landscape regime. The latter decade of 2010–2020 experienced several changes in the theoretical outlook of the NTR region. A report by Ghosal (2016a) stated that this project aimed at “green buildings” (buildings with less resource consumption and less waste generation), the creation of big water reservoirs and modern transport with low emissions. Moreover, the “Urban Development and Municipal Affairs Department” of West Bengal proposed a new scheme entitled “The Green City Mission” in 2016, which directed all the local urban bodies to increase the share of the landscape under green cover in each urban area by 15% within 2021 (Bengal, 2016). Such radical initiatives by the governing authority certainly trigger the township activities, and

2 Investigation of Ecological Sustainability Through the Landscape …

57

it has been well observed in Fig. 2.13. Despite such overwhelming “greening” and “blueing” activities, the increase in the built-up coverage in all the action areas surpassed every other activity, and a 43.62%, 30.95%, 32.15% and 18.11% increase in built-up areas have been observed in the action areas respectively, which has further degraded the BUAESI, mostly along Action Areas I and II. Although there has been a 31.97% of increment in the built-up areas in 2020, 0.75% and 16.39% of the area have been increased for water bodies and dense vegetative lands, respectively. A 0.95%, 0.56%, 0.84% and 0.90% of the increase in water bodies have been observed for Action Areas I, II, III and IV respectively whereas this increasing amount is almost 9.51%, 24.95%, 10.97% and 11.37% respectively for vegetative areas. Such an increase in the cityscape has certainly bolstered the low ecological sustainability of the region which has resulted in a certain increase in the WBESI and VESI values in 2010 and 2020. But during the field observation, it is noticed that the major share of vegetation are foreign species. These have further affected the comprehended quality of ecological sustainability, and the LUCESI values along the eastern parts of Action Areas II and III have certainly been enhanced. This study, therefore, analysed the impacts of the radical initiatives taken by public institutions and thereby urges for more such sustainable initiatives by public sectors in the future.

2.5 Conclusion The New Township Rajarhat (NTR) is indeed an important township project due to its distinctive outlook of creating a sustainable and planned cityscape as an alternative model of township planning, especially in third-world countries. But the result of this project has still not clearly visualized what it is supposed to deliver. The Sustainable Development Goal (SDG) 11 by the United Nations has particularly claimed for the development of resilient and sustainable cityscape that would further buttress the SDG 13 of combatting climate change. Which has been tried to discern from a landscape regime through this particular study. Although set up in a sustainable outlook, the primary phase of the development of the NTR is somewhat extremely unsustainable due to the encroachment of impervious areas into the water environment. This encroachment has an extreme impact on the overall performance of ecological sustainability, which has continued for almost more than 20 years since its inception. Such a long-term change in the landscape parameters has certainly destroyed several native species coupled with the eradication of ecological richness. Moreover, the acquisition and transformation of arable lands have impacted the climatic livelihood of the inhabitants. Although this study does not accomplish the NTR project from such a rhetoric space, several other studies have particularly mentioned the problems of exclusivity and alienation in the NTR region. Contrarily, the public initiatives of “green city” by the state government and the “greening” and “blueing” activities have adequately changed the landscape pattern in some parts of the NTR region and a reserved trend of comparatively better ecological sustainability has been observed along those parts in the recent times.

58

A. Kundu and Sk. Mafizul Haque

This study, therefore, calls for urgent policy interventions in the NTR region to restore its ecological sustainability in the coming future. Extensive regeneration of wetland ecosystems and plantation programs of native species should be introduced along with the creation of water bodies to reinstate ecological sustainability. Already, the eastern parts of Action Areas II and III have experienced the usefulness of such actions. Along with the creation of water bodies, the recreated water bodies should have percolative beds for better infiltration. More awareness programs regarding sustainability should be introduced among the inhabitants, and the regeneration program could be done in a participatory manner. On the other hand, this study has adroitly proved that fragmentation of impervious areas will certainly improve the sustainability of the region, hence, a mixed land use pattern should be created that would fragment the highly clustered built-up and barren patches. The effectiveness of mixed land use has also been corroborated earlier by different scholars. The western part of Action Area II, the southern part of Action Area III and the northern part of Action Area IV are still left with barren lands which can be used for adopting such mixed land use patterns. Finally, the NTR has particularly experienced the role of urban governance in augmenting the sustainability of a township project. The institutional incompetencies have degraded its sustainability, on the one hand, contrarily a lucrative political decision has bolstered its ecological sustainability in the near past. Therefore, strong urban governance should be acquainted with the NTR project which would consider further cityscape planning from a sustainable perspective and thus, the integrated radical actions from public and private stakeholders could make the NTR a model for township planning in India as well as in third-world countries.

References Aithal BH, Ramachandra TV (2016) Visualization of urban growth pattern in chennai using geoinformatics and spatial metrics. J Indian Soc Remote Sens 44(4):617–633. https://doi.org/10.1007/ s12524-015-0482-0 Araya YH, Cabral P (2010) Analysis and modeling of urban land cover change in Setúbal and Sesimbra, Portugal. Remote Sens 2(6):1549–1563. https://doi.org/10.3390/rs2061549 Atasoy M (2018) Monitoring the urban green spaces and landscape fragmentation using remote sensing: a case study in Osmaniye, Turkey. Environ Monit Assess 190(12):1–8. https://doi.org/ 10.1007/s10661-018-7109-1 Bera S, Das Chatterjee N (2019) Mapping and monitoring of land use dynamics with their change hotspot in North 24-Parganas district, India: a geospatial- and statistical-based approach. Model Earth Syst Environ 5(4):1529–1551. https://doi.org/10.1007/s40808-019-00601-2 Cai Y, Zhang H, Zheng P, Pan W (2016) Quantifying the impact of land use/land cover changes on the urban heat island: a case study of the natural wetlands distribution area of Fuzhou City, China. Wetlands 36(2):285–298. https://doi.org/10.1007/s13157-016-0738-7 Chadchan J, Shankar R (2012) An analysis of urban growth trends in the post-economic reforms period in India. Int J Sustain Built Environ 1(1):36–49. https://doi.org/10.1016/j.ijsbe.2012. 05.001 Chanu CS, Elango L, Shankar GR (2021) A geospatial approach for assessing the relation between changing land use/land cover and environmental parameters including land surface temperature

2 Investigation of Ecological Sustainability Through the Landscape …

59

of Chennai metropolitan city, India. Arabian J Geosci 14(2). https://doi.org/10.1007/s12517020-06409-0 Chowhan G, Sen A, Mukherjee J (2022) Sustainable and ‘smart’ restructuring around the making of mega and world-class cities in India: a critical review. GeoJournal (Baviskar) 2011. https:// doi.org/10.1007/s10708-022-10644-1 Dey I, Samaddar R, Sen SK (2016) Beyond Kolkata. In: Beyond Kolkata. https://doi.org/10.4324/ 9781315539928 Dhali MK, Chakraborty M, Sahana M (2019) Assessing spatio-temporal growth of urban subcentre using Shannon’s entropy model and principle component analysis: a case from North 24 Parganas, lower Ganga River Basin, India. Egypt J Remote Sens Space Sci 22(1):25–35. https:/ /doi.org/10.1016/j.ejrs.2018.02.002 Dhar RB, Chakraborty S, Chattopadhyay R, Sikdar PK (2019) Impact of land-use/land-cover change on land surface temperature using satellite data: a case study of Rajarhat Block, North 24Parganas District, West Bengal. J Indian Soc Remote Sens 47(2):331–348. https://doi.org/10. 1007/s12524-019-00939-1 Dutta K, Basu D, Agrawal S (2021) Synergetic interaction between spatial land cover dynamics and expanding urban heat islands. Environ Monit Assess 193(4):1–22. https://doi.org/10.1007/ s10661-021-08969-4 Fei W, Zhao S (2019) Urban land expansion in China’s six megacities from 1978 to 2015. Sci Total Environ 664:60–71. https://doi.org/10.1016/j.scitotenv.2019.02.008 Ghosal A (2016a) Green city mission: Rs 50 lakh each for all 125 municipalities in West Bengal. The Indian Express Ghosal A (2016b) West Bengal: Mamata dumps Centre’s Smart City project, to develop cities on its own. The Indian Express. https://indianexpress.com/article/india/india-news-india/west-ben gal-mamata-dumps-centres-smart-city-project-to-develop-cities-on-its-own-2983994/ Government of West Bengal (2016) Green city, clean city—a new mantra for West Bengal. https:/ /www.wburbanservices.gov.in/page/cms/green_city_mission#:~:text=TheMissionconsistsoft he,themedianoftheroads Habitat U (2022) Envisaging the future of cities Halder B, Bandyopadhyay J, Banik P (2021a) Monitoring the effect of urban development on urban heat island based on remote sensing and geo-spatial approach in Kolkata and adjacent areas, India. Sustain Cities Soc 74(July):103186. https://doi.org/10.1016/j.scs.2021.103186 Halder B, Banik P, Bandyopadhyay J (2021b) Mapping and monitoring land dynamic due to urban expansion using geospatial techniques on South Kolkata. Saf Extreme Environ 3(1):27–42. https://doi.org/10.1007/s42797-021-00032-2 Haque SM (2013) Urban expansion around Kolkata Metropolitan core and its impact on land use changes-a geo-spatial analysis. University of Calcutta. https://shodhganga.inflibnet.ac.in:8443/ jspui/handle/10603/163904 Haque SM (2020) Assessing infrastructural encroachment and fragmentation in the East Kolkata Wetlands. In: Bandyopadhyay S, Magsi H, Sen S, Dentinho TP (eds) Water management in South Asia socio-economic, infrastructural, environmental and institutional aspects, pp 233– 257. https://doi.org/10.1007/978-3-030-35237-0_13 Haque SM, Das A, Rukhsana (2020) Carbon footprint reduction instrument. In: Encyclopedia of renewable and sustainable materials, vol 2009, pp 300–311. https://doi.org/10.1016/b978-0-12803581-8.11042-2 Jain M (2018) Contemporary urbanization as unregulated growth in India: the story of census towns. Cities 73(May):117–127. https://doi.org/10.1016/j.cities.2017.10.017 Jain M, Dimri AP, Niyogi D (2016) Urban sprawl patterns and processes in delhi from 1977 to 2014 based on remote sensing and spatial metrics approaches. Earth Interact 20(14):1–29. https://doi. org/10.1175/EI-D-15-0040.1 Jha CS, Goparaju L, Tripathi A, Gharai B, Raghubanshi AS, Singh JS (2005) Forest fragmentation and its impact on species diversity: an analysis using remote sensing and GIS. Biodivers Conserv 14(7):1681–1698. https://doi.org/10.1007/s10531-004-0695-y

60

A. Kundu and Sk. Mafizul Haque

Kamusoko C, Aniya M (2007) Land use/cover change and landscape fragmentation analysis in the Bindura District, Zimbabwe. Land Degrad Dev 18(2):221–233. https://doi.org/10.1002/ldr.761 Kong F, Yin H, Nakagoshi N (2007) Using GIS and landscape metrics in the hedonic price modeling of the amenity value of urban green space: a case study in Jinan City, China. Landsc Urban Plan 79(3–4):240–252. https://doi.org/10.1016/j.landurbplan.2006.02.013 Kumar NM, Goel S, Mallick PK (2018) Smart cities in India: features, policies, current status, and challenges. In: International conference on technologies for smart city energy security and power: smart solutions for smart cities, ICSESP 2018—Proceedings, 1–4 January 2018. https:/ /doi.org/10.1109/ICSESP.2018.8376669 Kundu R (2016) Making sense of place in Rajarhat New Town: the village in the urban and the urban in the village. Econ Polit Wkl 51:93–101. https://www.jstor.org/stable/44003412?readnow=1&seq=9 Li Y, Cao Z, Long H, Liu Y, Li W (2017) Dynamic analysis of ecological environment combined with land cover and NDVI changes and implications for sustainable urban–rural development: the case of Mu Us Sandy Land, China. J Clean Prod 142:697–715. https://doi.org/10.1016/j.jcl epro.2016.09.011 Lillesand TM, Kiefer RW (1979) Remote sensing and image interpretation. In: Remote sensing and image interpretation. https://doi.org/10.2307/634969 Ma J, Zhang C, Guo H, Chen W, Yun W, Gao L, Wang H (2020) Analyzing ecological vulnerability and vegetation phenology response using NDVI time series data and the BFAST algorithm. Remote Sens 12(20):1–21. https://doi.org/10.3390/rs12203371 Magle SB, Theobald DM, Crooks KR (2009) A comparison of metrics predicting landscape connectivity for a highly interactive species along an urban gradient in Colorado, USA. Landsc Ecol 24(2):267–280. https://doi.org/10.1007/s10980-008-9304-x Majumdar S, Sivaramakrishnan L (2020) Patterns of land use in and around Kolkata City: a spatiopatterns of land use in and around Kolkata City. Indian Cartographer 35(April):218–223 Marulli J, Mallarach JM (2005) A GIS methodology for assessing ecological connectivity: application to the Barcelona metropolitan area. 71:243–262. https://doi.org/10.1016/j.landurbplan. 2004.03.007 McGarigal K, Marks JB (1995) FRAGSTATS: spatial pattern analysis program for quantifying landscape structure Milleniumpost (2019) HIDCO to come up with “vertical city” in Fintech hub. http://www.millen niumpost.in/kolkata/hidco-to-come-up-with-vertical-city-in-fintech-hub-373414 Mitra D, Banerji S (2018) Urbanisation and changing waterscapes: a case study of New Town, Kolkata, West Bengal, India. Appl Geogr 97(April):109–118. https://doi.org/10.1016/j.apgeog. 2018.04.012 Montgomery MR (2008) The urban transformation of the developing world. Science 319(761– 764):86 Mu B, Tian G, Xin G, Hu M, Yang P, Wang Y, Xie H, Mayer AL, Zhang Y (2021) Measuring dynamic changes in the spatial pattern and connectivity of surface waters based on landscape and graph metrics: a case study of Henan Province in Central China. Land 10(5). https://doi. org/10.3390/land10050471 O’Neill RV, Krummel JR, Gardner RH, Sugihara G, Jackson B, DeAngelis DL, Milne BT, Turner MG, Zygmunt B, Christensen SW, Dale VH, Graham RL (1988) Indices of landscape pattern. Landsc Ecol 1(3):153–162 Padmanaban R, Bhowmik AK, Cabral P, Zamyatin A, Almegdadi O, Wang S (2017) Modelling urban sprawl using remotely sensed data: a case study of Chennai City, Tamilnadu. Entropy. https://doi.org/10.3390/e19040163 Paul S (2012) Research article monitoring urban growth pattern and its impact on neighbouring villages: a case study on Rajarhat Region, West Bengal. 4:221–226 Paul S, Saxena KG, Nagendra H, Lele N (2021) Tracing land use and land cover change in periurban Delhi, India, over 1973–2017 period. Environ Monit Assess 193(2). https://doi.org/10. 1007/s10661-020-08841-x

2 Investigation of Ecological Sustainability Through the Landscape …

61

Pettorelli N, Vik JO, Mysterud A, Gaillard JM, Tucker CJ, Stenseth NC (2005) Using the satellitederived NDVI to assess ecological responses to environmental change. Trends Ecol Evol 20(9):503–510. https://doi.org/10.1016/j.tree.2005.05.011 Rahaman M, Dutta S, Sahana M, Das DN (2018) Analysing urban sprawl and spatial expansion of Kolkata urban agglomeration using geospatial approach. In: Applications and challenges of geospatial technology: potential and future trends, pp 205–221. https://doi.org/10.1007/978-3319-99882-4_12 Ringuest L (1986) A chi-square statistic for validating generated responses. Comput Oper Res 13(4):379–385 Rouse JW, Hass RH, Schell JA, Deering DW (1973) Monitoring vegetation systems in the great plains with ERTS. In: Third earth resources technology satellite (ERTS) symposium, vol 1, pp 309–317. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19740022614.pdf Saaty RW (1987) The analytic hierarchy process-what it is and how it is used. Math Model 9(3– 5):161–176. https://doi.org/10.1016/0270-0255(87)90473-8 Shafizadeh Moghadam H, Helbich M (2013) Spatiotemporal urbanization processes in the megacity of Mumbai, India: a Markov chains-cellular automata urban growth model. Appl Geogr 40:140– 149. https://doi.org/10.1016/j.apgeog.2013.01.009 Shahfahad RM, Naikoo MW, Ali MA, Usmani TM, Rahman A (2021) Urban heat island dynamics in response to land-use/land-cover change in the coastal city of Mumbai. J Indian Soc Remote Sens 49(9):2227–2247.https://doi.org/10.1007/s12524-021-01394-7 Sharma R, Joshi PK (2013) Monitoring urban landscape dynamics over Delhi (India) using remote sensing (1998–2011) inputs. J Indian Soc Remote Sens 41(3):641–650. https://doi.org/10.1007/ s12524-012-0248-x United Nations (2016) The 2030 agenda for sustainable development. https://doi.org/10.1201/b20 466-7 WBHIDCO (1999) Rapid enviroment impact assessment and environment plan management, pp 1–230 Weng YC (2007) Spatiotemporal changes of landscape pattern in response to urbanization. Landsc Urban Plan 81(4):341–353. https://doi.org/10.1016/j.landurbplan.2007.01.009 Zadeh LA (2013) Fuzzy logic. Comput Complex: Theory Tech Appl 9781461418:1177–1200. https://doi.org/10.1007/978-1-4614-1800-9_73 Zavadskas EK, Turskis Z (2010) A new additive ratio assessment (ARAS) method in multicriteria decision-making. Technol Econ Dev Econ 16(2):159–172. https://doi.org/10.3846/tede.2010.10 Zhang G, Dong J, Xiao X, Hu Z, Sheldon S (2012) Effectiveness of ecological restoration projects in Horqin Sandy Land, China based on SPOT-VGT NDVI data. Ecol Eng 38(1):20–29. https:// doi.org/10.1016/j.ecoleng.2011.09.005

Chapter 3

Advanced Remote Sensing for Sustainable Decent Housing for the Economically Challenged Urban Households F. N. Karanja

and P. W. Mwangi

Abstract Future cities are viewed as environments that will provide opportunities for all populations (gender, age and persons with disability) with access to basic services, energy, housing and transportation. The Sustainable Development Goal (SDG) 11 is on sustainable cities and communities with a total of 10 targets and 15 indicators; with the first target concerned with urban households living in inadequate housing. This is aligned with the Africa Agenda 2063 whose first goal is on ensuring a sustainable environment for all citizens. UN-Habitat estimates that by 2030, 40% of the world’s population will need access to adequate housing with 100 million people globally being homeless and one in four people living in harmful conditions. Interventions aimed at reversing these trends should be supported by evidence. In this regard, this chapter explores the role of advanced remote sensing in providing relevant spatial information on economically challenged urban households using Kibera, Nairobi Kenya as a case study. Google Earth Engine (GEE) environment with two machine learning algorithms namely Random Forest (RF) and Support Vector Machines (SVM) were investigated with Sentinel 2A image data. From the results obtained, RF performed better with 86% accuracy compared to 74% with SVM in the classification process. Keywords Economically challenged urban households · Sustainable housing · GEE · RF · SVM

F. N. Karanja (B) Department of Geospatial Engineering & Space Technology, University of Nairobi, Nairobi, Kenya e-mail: [email protected] P. W. Mwangi Department of Spatial and Environmental Planning, Kenyatta University, Nairobi, Kenya e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_3

63

64

F. N. Karanja and P. W. Mwangi

3.1 Introduction One of the fundamental human rights is access to a safe, secure habitable and affordable housing. However, according to UN-Habitat, 1.8 billion people globally live in inadequate housing characterized by overcrowding and deplorable conditions. This problem is compounded by a lack of access to hand washing facilities for a 3 billion population (UNICEF, WHO), with another 4 billion not having any form of social protection (ILO). Further, it is reported that every year 2 million people globally are forcibly evicted and many more living in fear of eviction. Tragically, 80% of cities globally do not have affordable housing options for half of their population. In urban areas, these people live in slums or informal settlements giving rise to the economically challenged urban households. They are vulnerable to natural and man-made shocks like floods, earthquakes, epidemics, landslides, etc. The global and regional frameworks namely Sustainable Development Goals (SDG) and Africa Agenda 2063, respectively, recognize the importance of housing by having specific goals and indicators. Goal number 11 in the SDG framework focuses on the issue of sustainable cities and communities with a total of 10 targets and 15 indicators. The first target of the SDG 11 is on safe and affordable housing whose indicator is used to measure the proportion of urban households living in slums, informal settlements or inadequate housing. These households are characterized by a lack of access to improved water and sanitation, insufficient living area and unsustainable housing (UN-HABITAT 2006). The 2030 goal is to ensure access for all, to adequate, safe and affordable housing and basic services and upgrade slums. This resonates well with the Africa Agenda 2063 whose first goal is on ensuring a high standard of living, quality of life and well-being for all citizens, with the priority area being modern, affordable and liveable habitats and quality basic service. Different terminologies are associated with economically challenged household for instance informal settlements, slums, shanty towns, squats, homelessness, backyard housing and pavement dwellers.

3.2 Background 3.2.1 Attributes to Adequate Housing The right to adequate housing is provided for in three areas as shown in Table 3.1. Adequate housing guarantees freedom and entitlement as well as ensures certain conditions are fulfilled. Further, access to adequate housing for all demands that the housing sector of the urbanization process embraces the three key pillars namely affordability, sustainability and inclusiveness (UN Habitat). Indeed the concept of housing is defined by health, dignity, safety, well-being and inclusion variables. These variables are elaborated in Table 3.2.

3 Advanced Remote Sensing for Sustainable Decent Housing …

65

Table 3.1 Rights to adequate housing Freedom

Entitlement

Conditionality

Protection

Security of tenure

Security of tenure

Accessibility

Interference

Housing, land and property restitution

Availability of services

Location

Choice

Equal and non-discriminatory Affordability access Participation

Cultural adequacy

Habitability

It is therefore imperative that in order to access other essential services like education, health, employment and social services among others, housing is very key. Sustainable urban development demands deliberate efforts by governments and all relevant actors to have housing as a core in their urban policies. This is the only way to ensure that by 2030 the estimated 3 billion people which translate to approximately 40% will have access to adequate housing. Translating policy into action means that 96,000 new affordable and accessible housing units are built every day. This challenge is exacerbated by the population that is homeless and those living in deplorable conditions. It is estimated that about one-quarter of the world’s urban population live in informal settlements, and they are considered to be the most economically and socially vulnerable people (Helber et al. 2018). With renewed interest in improving and eradicating informal settlements in the last 15 years, remote sensing (RS) image analysis has played a major role in providing information on the dynamics and geography of slums (Kadhim et al. 2015). To improve our understanding of changing cities, it is critical to monitor the evolution of urban spaces at temporal and spatial scales. Therefore, information that is detailed on the location and boundaries of informal settlements is important for urban development and sustainable urban planning (Fan et al. 2022). Poorly planned and unplanned areas are as a result of increased urban migration, leading to the expansion of cities. Unplanned growth can negatively impact the environment, society and economy at local, national, regional and global scales (Kadhim et al. 2015). Advances in satellite remote sensing provide cost-effective technologies that enable the monitoring of urban systems and their impacts on the environment in support of informed decisions, process and programmes.

3.2.2 Economically Challenged Urban Households Urban areas are increasingly facing challenges in providing appropriate socioeconomic and environmental living conditions due to the rise in urban populations (Gaube and Remesch 2013). It is estimated that about 490 million people in Africa lived below the poverty line of less than 1.90 dollars per day by 2021, which was an

66

F. N. Karanja and P. W. Mwangi

Table 3.2 Housing variables Variable

Description

(i) Health

Housing within the context of Health is crucial given the dynamics of urban growth, changes in the demographics for instance ageing population and the shocks occasioned by climate change (WHO 2018) Poor housing conditions associated with overcrowding, poor sanitation and lack of adequate ventilation lead to a variety of health conditions including respiratory infections Housing is also considered a social determinant of good health and well-being

(ii) Dignity

Housing ensures everyone regardless of their status has a roof over their head Most countries have provided the right to housing in their legal instruments with the aim of ensuring the dignity of the citizens is upheld The right to adequate housing should be seen through the lens of security, peace and dignity

(iii) Well being

Peace of mind and belonging Over one billion people live in sub-standard housing and informal settlements with frequent threats of displacement due to various reasons for instance constructions, natural disasters like flooding or landslides as well as forced evictions. This, therefore, means these people live in constant fear of the unknown and are not able to have long-term plans for their lives

(iv) Safety

Security Slum or informal settlements lack emergency services which expose the communities to natural and man-made shocks The most common natural shocks that threaten communities in the slum areas include flooding, earthquakes as well as cyclones. Man-made shocks include fires, human conflicts as well as motorized and non-motorized accidents There is therefore need to build resilient communities against external shocks occasioned by climate variability and change

(v) Inclusion

Social and economic inclusivity Urban areas are centres of economic activity contributing about 60% of the GDP Urbanization and population growth are not aligned to the provision of adequate housing Lack of adequate housing results in negative effects with regard to equity and inclusion, health and safety and livelihood opportunities (UNSTATS 2019)

increase of about 37 million people after the COVID-19 pandemic started (Schoch and Lakner 2020; UNCTAD 2021). Eradication of extreme poverty for all people everywhere by 2030 is an underlining theme in the seventeen (17) global sustainable

3 Advanced Remote Sensing for Sustainable Decent Housing …

67

development goals. The availability of data and timely analysis of these data are critical in determining and ending poverty in economically challenged areas (Pokhriyal and Jacques 2017). Informal settlements have been categorized into three types by UN-HABITAT (2008): (1) squatter: areas of poor housing quality built on land that is occupied illegally; (2) Irregular subdivision: in which the legal owner subdivides the land into sub-standard plots and sells or rents them out without following all relevant building bylaws; (3) slum: traditionally described as a neighbourhood of housing that was once in good condition but has since deteriorated or been subdivided into a state of high crowding and rented out to low-income groups. Further, a slum is considered an urban area where people live under the same roof and lack one or more of the following: sufficient living area, access to proper sanitation, clean water, durable housing and secure tenure (UN-HABITAT 2008). The growth of informal settlements can be classified into three phases: infancy, consolidation and saturation. Analysis of these phases is determined by the spatial characteristics of the informal settlements (Maula et al. 2019). Many of the urban poor work in the informal sector. Rural–urban migration has been touted as one of the greatest causes of an increase in urban populations. Provision of services and decent housing, which for many urban poor is beyond their economic means, has been a challenge many economies cannot guarantee (Chaudhuri 2015; Cherunya et al. 2020). For instance, India’s urban population is facing challenges of poor health, pollution, scarcity of services, housing and space. The high population growth rate of cities in India is a major concern, with approximately 26% of urban dwellers in four megacities namely Mumbai, Delhi, Kolkata and Chennai living in spontaneous settlements or slums (Chaudhuri 2015).

3.2.3 Percentage of Urban Population Living in the Economically Challenged Urban Areas In order to appreciate the gravity of the problem with regard to people living in economically challenged urban households, the indicator on the percentage of the urban population living in these environments has been employed. Three continents have been used to demonstrate the geographical spread of people living in the slums namely Africa, South America and Asia in 1990 and 2010 on a scale of 0–100% as shown in Fig. 3.1. At this scale, it is evident that the population living in slums varies from one continent to another and indeed from one country to another. Africa seems to bear the biggest burden of economically challenged population living in slums or informal settlements. Quantitative analysis of the number of people living in slums in three countries namely Kenya, Brazil and India between 1990 and 2018 shows variations and trends as depicted in Fig. 3.2.

68

F. N. Karanja and P. W. Mwangi

1990 Data (Source UN-Habitat, https://sdg-tracker.org/cities)

2010 Data (Source UN-Habitat, https://sdg-tracker.org/cities)

Fig. 3.1 Percentage of people living in slums in Africa, South America and Asia in 1990 and 2010

Kenya

Brazil

India

Fig. 3.2 Percentage of people living in slums in Kenya, Brazil and India between 1990 and 2018 (Source UN-Habitat, https://sdg-tracker.org/cities)

From Fig. 3.2, it is evident that Kenya has the highest of its urban population living in slums a situation that remained stable until the year 2014 when the number reduced. The situation in Brazil and India shows a decline in the urban population living in slums from 1990 up to around the year 2016 and 2014, respectively. In the case of India, a notable increase in the population from 2014 is observed.

3 Advanced Remote Sensing for Sustainable Decent Housing …

An Example of a Slum-Kenya

An example of a Slum-Brazil

69

An Example of a Slum-India

Fig. 3.3 Satellite images of sample slum areas in Kenya, Brazil and India

3.2.4 Spatial Characteristics for Economically Challenged Urban Areas Figure 3.3 shows samples of slum areas from satellite images for Kibera, Rocinha and Dharavi in Kenya, Brazil and India, respectively. From the visual interpretation of the images, similar characteristics of these slum areas are observed namely i. ii. iii. iv. v. vi. vii.

Overcrowded structures Different shapes of buildings Variations in building sizes Different rooftops designs Inconsistent pattern of the buildings Poor access Lack of infrastructure.

In order to inform policy on urban upgrading programmes with the goal of providing decent housing to the economically challenged urban population data is very key. Mapping urban areas from satellite imagery is a complex process given their heterogeneous nature. This is even more complex within a slum area due to the unplanned structures that vary in shape, size and pattern. This therefore requires an innovative approach to the extraction of buildings in order to quantify the number of households. The granularity of the data in terms of resolution should be such that the individual units can be identified and mapped.

3.2.5 Remote Sensing for Economically Challenged Urban Areas Urbanization process is considered as the change from rural to urban context. Urbanization can be caused by three key factors including rural–urban migration, annexation and increase in population. Further, it is related to the economy and demography of a city, the land use, topography and transportation in a city (Taubenbock and Esch 2011). Drivers of urban development are very diverse and include micro-economic

70

F. N. Karanja and P. W. Mwangi

factors (prices of land, accessibility of cheap agricultural land, increasing living standards, etc.); macro-economic factors (globalization, growth of the economy, etc.); housing preferences; transportation factors (low fuel costs, poor public transport, ownership of private cars, road availability, etc.); demography; regulatory frameworks (poor administration of existing plans, poor land use planning, etc.); inner city dynamics (insecure environments, absence of open green spaces, poor quality schools, poor air quality, noise, social issues, etc.) (Taubenbock and Esch 2011). Remote sensing (RS) is an important tool in modelling urban areas in the determination of urban morphology and the interaction of urban spaces and the builtenvironment. Its strength lies in its synoptic view and ability to provide up-to-date, area-wide, independent and relatively cost-effective transformation of data into information (Mboga et al. 2017). Remote sensing makes use of various techniques for automatic extraction of features including statistics, fury and neural classifiers, for application-driven products. Remotely sensed data provides information on the characteristics of land cover and their changes in various temporal and spatial scales (Fan et al. 2022; Taubenbock and Esch 2011). Due to its cost-effectiveness in monitoring socio-technical systems and urban changes, stakeholders can make informed decisions that reduce the negative environmental impacts of urban expansion. Conventional methods of observation over large expanses are constraining as they need more time, effort and time (Kadhim et al. 2015). Mapping of populations at fine scales is increasingly being done using remotely sensed data such as location-based service data, building features, density of road and mobile phone data (Jing et al. 2020). The built-environment morphology, particularly of informal settlements, has patterns that are unique across settlements and they include shape, size, scale and distribution (Samper et al. 2020) (Fig. 3.4a, b). Mapping of informal settlements is crucial for the management of utilities, resources and guiding policies for planning. In supporting the reporting of SDGs, UN-Habitat developed processes in 2003 for mapping and monitoring inequality in urban areas (Fallatah et al. 2019). Mapping urban forms in informal settlements using new RS technologies and digital mapping has provided an opportunity for comparative analysis in cities (Samper et al. 2020). Indicators for informal settlements are crucial in developing frameworks to undertake image analysis and they include road

(a)

(b)

Fig. 3.4 Different morphologies of slums in Kenya: a Kibera slum and b Majengo slum (Source Google Earth)

3 Advanced Remote Sensing for Sustainable Decent Housing …

71

segment type, materials, texture measures of built-up areas, lacunarity of structures of housing/vacant land, extent of vegetation, dwelling size and extent of roofing of built-up areas (Fallatah et al. 2019). The piece-meal additive construction is a key characteristic of informal settlements due to their constant changing state which complicates the study of these urban forms. Much focus and development have been put into remote sensing studies in exploring morphological features of informal settlements and identifying them by automatic classification of optical imagery (Maula et al. 2019; Samper et al. 2020). With the expansion of informal settlements in urban areas, there is a lack of a comprehensive database that explains the global situation. Informal urban developments have been mapped scholarly into two scales: national and international. There is a creation of blindness of informality with these two scales of measurement at the urban level and its relationship with global trends (Samper et al. 2020). With advances in RS methodologies and rates of growth in the quality of imagery available, mapping of urban forms would therefore be automated in future (Samper et al. 2020). Monitoring of unstructured (informal) human settlements using spatial technology needs to be low-cost in terms of acquisition and processing. It would also need to be as automated as possible to ensure results are obtained faster and are reliable, which are then founded on tested algorithms and routines (Dell’acqua et al. 2006).

3.2.6 Advanced Remote Sensing Techniques for Mapping Economically Challenged Urban Areas 3.2.6.1

Remote Sensing Data

Big data-driven technologies are crucial in sustainable development and addressing challenges dealing with urban management. Urban management and smart city approaches are compatible with big data technologies in improving urban sustainability (Wu et al. 2022). Due to the limitation of geolocation datasets, RS is integrated with geolocation datasets at finer scales or city levels for urban mapping (Xia et al. 2019). Geolocation GIS datasets can be used in mapping the dynamics of human movement as it contains information on daily human life. This dataset, together with the development of the Internet, spatial positioning technologies and open-access GIS datasets, can enable researchers in extracting urban areas to gather more precise information especially in informal settlements (Xia et al. 2019). Information on human activities and socioeconomic functions is critical for conserving biodiversity, controlling disasters, environmental management, designing landscapes and urban planning (Chen et al. 2021). Various thematic products have been developed to determine impervious areas and they include Human Built-up and Settlement Extent derived from Landsat (Wang et al. 2017), Global Human Settlement Layer (Gong and Howarth 1990), Global Man-made Impervious Surface

72

F. N. Karanja and P. W. Mwangi

(Brown de Colstoun et al. 2017), Global Artificial Impervious Area (Gong et al. 2020), Normalized Urban Areas Composite Index (Liu et al. 2018) and multi-source, multi-temporal random forest classification (Zhang et al. 2020). The volume of earth observation (EO) data has increased with the developments of large and small satellite missions such as CubeSat technology, which are used in monitoring and mapping urban areas (Kamusoko 2022). Multiple remote sensing products, numerous geospatial big data and increased power of geographic information systems (GIS) have also enabled improved estimations of the population at global and regional scales.

3.2.6.2

Optical Remote Sensing

Qualities of urban form can be analysed using high-resolution satellite imagery (Samper et al. 2020). Very high-resolution (VHR) (e.g., Quickbird, Plaeides, WorldView) and high-resolution (HR) sensors produce remotely sensed imagery that enables detailed mapping and analysis of urban areas. With the development of these new sensors, ground objects can now be easily identified at finer scales with features such as geometry, size, location and adjacent information easily identifiable (Chen et al. 2021). Obtaining up-to-date thematic maps over informal settlements is a challenge in VHR satellite imagery, but Unmanned Aerial Vehicles (UAVs) have assisted in breaching this time gap (Samper et al. 2020), hence their increased use in urban studies (Zhai et al. 2020). The Kenya Informal Settlement Improvement Project (KISIP) acquired imagery for three informal settlements using UAVs to generate base maps to aid in planning for informal settlements (Onyango 2019).

3.2.6.3

LiDAR

Light detection and ranging (LiDAR) emits a laser beam that is reflected back to the receiver which determines distances from objects. The advantage of this technology is capturing high-accuracy structural information (Zhai et al. 2020) which enables the extraction of feature information in built-up areas. In Brazil, informal settlements, known as favelas, have been mapped using LiDAR thus providing insight into their urban and architectural logic. Due to the complexity of the structure of informal settlements, satellite imagery mainly provides a birds-eye view of the organization and shape of structures. Spatial metrics derived from LiDAR such as building heights and street width can provide indications of hotspot areas where certain risks of events are likely to occur such as landslides and fire risks (Gibson et al. 2021).

3.2.6.4

Microwave Remote Sensing

In characterizing urban features and structures, micro-wave-based imaging is proving a valuable technique in imaging especially in the concept of multi-dimensional urban

3 Advanced Remote Sensing for Sustainable Decent Housing …

73

studies. Discriminating between structured and unstructured settlements is possible using coarse spatial resolution SAR data (Dell’acqua et al. 2006). There have been an increasing number of high-resolution (HR) and very high-resolution (VHR) synthetic aperture radar (SAR) sensors such as Sentinel-1, ALOS PALSAR, Envisat ASAR, TerraSAR-X, RADARSAT-1 and COSMO-Skymed (Zhai et al. 2020). Its application in urban studies was accelerated by its ability to capture height and texture information (Zhai et al. 2020), in addition to the advantages of operating in any weather condition and penetrating cloud cover. Karim and Van Zyl (2021) fused Sentinel-2 imagery with Sentinel-1 DInSAR processed imagery for urban change detection. They used as baselines two-layer convolutional neural network (ConvNet2) with machine learning models using non-deep learning classifiers and feature descriptors. Moya et al. (2022) used time series analysis of Sentinel-1 SAR images to map informal settlements located in vulnerable areas of Lima, Peru using the SNAP toolbox to process the images.

3.2.7 Approaches to Information Extraction on Economically Challenged Urban Areas Huge volumes of EO data are becoming more complex to analyse and development of open source software and use of Google Earth Engine (GEE), a platform that is cloud based with massive computational capabilities (Kamusoko 2022) without needing to download or upload massive amounts of data (Goldblatt 2018). Advances in statistical machine learning algorithms have been used in various applications such as object recognition, classification and image segmentation. These algorithms include random forest (RF), support vector machine (SVM), neural network (NN) and deep learning-based convolution neural networks (CNNs) (Chen et al. 2021; Kamusoko 2022). Using VHR imagery, these methods have higher accuracies in detecting informal settlements (Samper et al. 2020). Algorithms such as contour (snakes), object-oriented and radial casting models with VHR imagery enable the identification and extraction of informal settlements (Samper et al. 2020) (Fig. 3.5). The most common approach in classifying informal settlements is the incorporation of spatial-contextual information through Object-Based Image Analysis (OBIA). Despite some of its limitations, Level Co-occurrence Matrix (GLCM) with very highresolution imagery can enable the extraction of informal settlements (Samper et al. 2020). Fallatah et al. (2019) used the object-based image analysis (OBIA) technique in informal settlement areas in Jeddah, in the Middle East. These techniques operate in two stages with the first stage being object segmentation where pixels are grouped into homogenous objects. The second is object-based land use classification where the individual ‘land-cover objects’ are classified accurately such as road, building, tree and water body (Chen et al. 2021). Helber et al. (2018) used two different approaches and satellite datasets for detecting informal settlements. The first method used Sentinel-2 spectral satellite

74

F. N. Karanja and P. W. Mwangi

Fig. 3.5 Object-based image analysis with machine learning of informal settlements (Source Fallatah et al. 2022)

data which was analysed using Pixel-wise Classification with Canonical Correlation Forest (CCF), while the second method used semantic segmentation deep neural network on VHR satellite at 30 m.

3.3 Study Area The increase in population in urban areas continues to pose serious challenges to existing services namely housing, health and education. In Nairobi, Kenya, an estimated two million people lived in informal settlements in 2019, almost half the city population. (Hang et al. 2020). This is despite the fact that informal settlements area coverage is approximately 5% of the total residential land area. Kibera is the largest informal settlement in Kenya with more than 15 villages and it is approximately 6.6 kms from Nairobi City centre. Figure 3.6 is a map of the Kibera informal settlement which was the area of focus. Kibera, just like many other informal settlements, is characterized by a lack of security of tenure, inadequate housing, physical and social infrastructure as well as sustainable income generation opportunities. The Government of Kenya introduced two programmes namely the Kenya Slum Upgrading Programme (KENSUP) in 2004 and the Kenya Informal Settlement Improvement Project (KISIP) in 2011 with a view of improving the livelihoods of the vulnerable community. These efforts are complemented by many programmes initiated by Non-Governmental Organizations. Evidence-based impact assessment of the various interventions is critical in flagging off the successes and failures, thus informing areas that require improvement with expected outcome. Remote sensing is an indispensable technology due to the wealth

3 Advanced Remote Sensing for Sustainable Decent Housing …

75

Fig. 3.6 Study area

of data and information that can be used to support inclusive programmes for the economically challenged urban households.

3.4 Materials and Methods 3.4.1 Conceptual Framework for Advanced Remote Sensing for Economically Challenged Urban Areas According to Peter Drucker, Management Theorist, “What gets measured, gets done and managed”. In order, therefore, to facilitate the targeted implementation of programmes and interventions for sustainable decent housing within the context of urban national plans, it is important to map the spatial extent of slums or informal settlements and associated attributes. The integrated workflow of information extraction of the spatial extent of slums/informal settlements is shown in Fig. 3.7. Given the heterogeneity and dynamics of the housing structures, a number of metrics should be incorporated including indigenous knowledge (Crowd sourced data); the geometric

76

F. N. Karanja and P. W. Mwangi Informal Settlement/Slums from Space •Kibera

Integrated Image Processing Metrics

•Size •Shape •Pattern •Roof Types • Colour •Indigneous Knowledge

Spatial Distribution of Informal Settlement/Slums

• Bdgs

Fig. 3.7 Integrated image processing of informal/slum areas workflow

variables namely shape, size and pattern as well as qualitative data for instance roof types and colour.

3.4.2 Datasets The study explored cloud-based techniques within Google Earth Engine (GEE) with regard to the advantages and potential application of advanced remote sensing techniques in mapping informal settlements using machine learning. The aim was to deploy two machine learning algorithms namely Random Forest (RF) and Support Vector Machines (SVM) in mapping informal settlements. Sentinel 2A image collection (COPERNICUS/S2_SR surface reflectance dataset) was used for the analysis. It is comprised of four bands with a spatial resolution of 10 m, five bands at 20 m and six bands at 60 m. The period 1st February to 27th February, with a cloud cover of less than 10%, was selected as filtering criteria to yield multiple imageries suitable for analysis. These were combined in the GEE environment. A single image object was obtained which represented a composite image where each pixel represents the median value in each band for the filtered images.

3.4.3 Methods The image analysis process involved loading the imagery, carrying out preprocessing by filtering, creating image composites, selecting training sites, image classification and accuracy assessment using test data. The summary of the methodology workflow is depicted in Fig. 3.8. Two machine learning techniques, random forests (RF) and support vector machines (SVM), were used and a comparison of their overall accuracies was done. A bounding box around Kibera informal settlements was used as the area of interest. A total of four land-cover classes were identified within the area. These included

3 Advanced Remote Sensing for Sustainable Decent Housing …

77

Fig. 3.8 Methodology flowchart

settlements, roads, bare ground and vegetation. The bands used in the classification were B2 (blue), B3 (green), B4 (red) and B8 (NIR) bands, as they each have a spatial resolution of 10 m. Pixel-based classifiers were used in the GEE environment. RF machine learning algorithm has successfully been used to map out informal settlements. It is considered robust in mapping out complex environments where high-dimensional feature spaces are concerned (Matarira et al. 2022). SVM classifiers separate training classes based on an optimal hyperplane. By computing the margin of the hyperplane and determining the maximum value, the best hyperplane separating the two classes can then be determined (Kamal et al. 2019). RF and SVM have been used in various applications such as in agriculture, wetlands and urban areas. The classification models were built in the GEE environment and tuning of the model was done using the ‘ee.’ Classifier package. Training samples were collected as points across the area of interest, and these were areas having homogenous pixels. The model was then designed to randomly sample and create approximately 60% of the original dataset to be training data and 40% to be validation data for cross-validating the accuracy

78

F. N. Karanja and P. W. Mwangi

of RF and SVM. To ensure classifier stability, 30 replications of bootstrap sampling were carried out where the same iterative classification trials were undertaken for each model.

3.5 Results and Discussions The results obtained from the two approaches are shown in Fig. 3.9. Specifically, Fig. 3.9a are the results from the SVM model whereas Fig. 3.9b shows the output from the RF model. A visual comparison of the results does indeed show some distinct differences in the classification of some parts of the area of interest. SVM appears to be more sensitive particularly in the densely built-up areas compared to the RF model which tends to generalize and hence less noisy. The sample areas of the information classes (settlement, bare ground and vegetation) highlighted in both the classified images are a demonstration of how each of the models performed. The overall accuracy assessment of both models was calculated and determined. RF accuracy was 86% while SVM had an overall accuracy of 74%. From the results obtained, the RF model performed better compared to SVM in classifying images within informal settlements. These results are a demonstration of the robust nature of GEE in extracting varying morphological features in a heterogeneous environment.

Fig. 3.9 Kibera informal settlement with SVM (a) and RF (b) classification

3 Advanced Remote Sensing for Sustainable Decent Housing …

79

In general, Remote Sensing techniques are useful in detecting and mapping informal settlements, despite the challenge of determining the mechanisms of their genesis. It is worth noting that even with great technological advancements made in developing algorithms for detecting informal settlements, historical in-depth knowledge of each settlement is necessary and valuable in the mapping process. Various algorithms such as object-based algorithms are able to take advantage of morphological differences in informal settlements (Hofmann et al. 2015). Informal settlements are naturally dynamic and unplanned. For the creation of various scenarios of informal settlement development, the various models would need genesis steering mechanisms. The availability of historical VHR imagery is a challenge in many developing countries making it difficult to map out their genesis. Indicators to be considered and detected for informal settlements include the 2D and 3D appearance and orientation, density, street network layout and terrain. Given the unique nature of the informal settlements, the data and models, the temporal and spatial resolution must be suitable (Hofmann et al. 2015).

3.6 Conclusions and Recommendations Adequate housing is a fundamental human right. This is more critical in urban areas where many countries are struggling to ensure the population is accorded decent and sustainable housing as a result of urbanization. Indeed, from 2007, about 50% of the world’s population was estimated to be living in the urban areas a number expected to rise to 60% by 2030. Population growth together with rapid urbanization has contributed to an increase in economically challenged urban households. These households are characterized by inadequate infrastructure and services and a population vulnerable to man-made and natural shocks. In order to promote inclusivity within the framework of sustainable urban housing, there is a need for evidence-based urban plans at the national level. Remote Sensing is a useful tool for mapping and quantifying the spatial extent of the informal/slum areas. Given the complex nature of the informal/slum areas in terms of variations in size, shape, pattern, roof types and colour among many other attributes, it is imperative that an integrated image processing approach would be desirable. Further, additional information at the local level (crowdsourcing) would also be valuable in the identification, extraction and labelling of informal/slum features extracted. Advanced remote sensing is therefore a useful approach because it takes into account not only the spectral characteristics but also the context of the economically challenged urban household. The information emanating from this process would inform the urban planning process leading to focussed and targeted interventions. The end result would be an all-inclusive and sustainable urban environment.

80

F. N. Karanja and P. W. Mwangi

References Brown de Colstoun EC, Huang C, Wang P, Tilton JC, Tan B, Wolfe RE, et al (2017) Global manmade impervious surface (GMIS) dataset from landsat, v1: Global high resolution urban data from landsat. Retrieved July 11, 2022, from https://sedac.ciesin.columbia.edu/data/set/ulandsatgmis-v1 Chaudhuri S (2015) Urban poor, economic opportunities and sustainable development through traditional knowledge and practices. Glob Bioeth 26(2):86–93. https://doi.org/10.1080/11287462. 2015.1037141 Chen B, Xu B, Gong P (2021) Mapping essential urban land use categories (EULUC) using geospatial big data: progress, challenges, and opportunities. Big Earth Data 5(3):410–441. https://doi. org/10.1080/20964471.2021.1939243 Cherunya PC, Truffer B, Samuel EM, Lü C (2020) The challenges of livelihoods reconstruction in the context of informal settlement upgrading. Econ Space 53(1):168–190. https://doi.org/10. 1177/0308518X20926514 Dell’acqua F, Gamba P, Stasolla M (2006) Unstructured human settlement mapping with SAR sensors moon mapping view project snowpack monitoring-SNOWAVE View project unstructured human settlement mapping with SAR sensors. In IGARSS, pp 3602–3605. https://doi.org/ 10.1109/IGARSS.2006.927 Fallatah A, Jones S, Mitchell D, Kohli D (2019) Mapping informal settlement indicators using object-oriented analysis in the Middle East. Int J Digital Earth 12(7):802–824. https://doi.org/ 10.1080/17538947.2018.1485753 Fallatah A, Jones S, Wallace L, Mitchell D (2022). Combining Object-Based Machine Learning with Long-Term Time-Series Analysis for Informal Settlement Identification. https://doi.org/ 10.3390/rs14051226 Fan R, Li J, Song W, Han W, Yan J, Wang L (2022) Urban informal settlements classification via a transformer-based spatial-temporal fusion network using multimodal remote sensing and time-series human activity data. Int J Appl Earth Obs Geoinf 111:102831. https://doi.org/10. 1016/j.jag.2022.102831 Gaube V, Remesch A (2013) Impact of urban planning on household’s residential decisions: an agent-based simulation model for Vienna. Environ Model Softw 45:92–103. https://doi.org/10. 1016/j.envsoft.2012.11.012 Gibson L, Adeleke A, Hadden R, Rush D (2021) Spatial metrics from LiDAR roof mapping for fire spread risk assessment of informal settlements in Cape Town, South Africa. Fire Saf J 120. https://doi.org/10.1016/j.firesaf.2020.103053 Goldblatt R (2018) Artificial intelligence for smart cities and sustainable urban systems. Retrieved July 11, 2022, from https://www.gim-international.com/content/article/artificial-intelligencefor-smart-cities-and-sustainable-urban-systems Gong P, Howarth PJ (1990) The use of structural information for improving land-cover classification accuracies at-the rural-urban fringe. Photogram Eng Remote Sens 56(1):67–73 Gong P, Li X, Wang J, Bai Y, Chen B, Hu T, Zhou Y, et al (2020) Annual maps of global artificial impervious area (GAIA) between 1985 and 2018. Remote Sens Environ 236. https://doi.org/10. 1016/j.rse.2019.111510 Hang R, Wei G, Zhenke Z, Musyoka KL, Priyanko D (2020) Population density and spatial patterns of informal settlements in Nairobi, Kenya. Sustainability 12(18):7717. https://doi.org/10.3390/ su12187717 Helber P, Kaiserlautern T, Gram-Hansen B, Kopackova V, Bilí P (2018) Generating material maps to map informal settlements. In: Neural information processing systems (NeurlPS 2018), Machine learning for the developing world (ML4DW) Workshop. Retrieved from https://frontierdevelop mentlab.github.io/informal-settlements/ Hofmann P, Taubenböck H, Werthmann C (2015) Monitoring and modelling of informal settlements—A review on recent developments and challenges. In: Joint urban remote sensing event,

3 Advanced Remote Sensing for Sustainable Decent Housing …

81

JURSE 2015. Institute of Electrical and Electronics Engineers Inc. https://doi.org/10.1109/ JURSE.2015.7120513 Jing C, Zhou W, Qian Y, Yan J (2020) Mapping the urban population in residential neighborhoods by integrating remote sensing and crowdsourcing data. Remote Sens 12(3235):16. https://doi. org/10.3390/rs12193235 Kadhim N, Mourshed M, Bray M (2015) Advances in remote sensing applications for urban sustainability. Euro-Mediterr J Environ Integr 1. https://doi.org/10.1007/s41207-016-0007-4 Kamal M, Jamaluddin I, Parela A, Farda NM (2019) Comparison of Google earth engine (GEE)based machine learning classifiers for mangrove mapping. In: The 40th Asian conference on remote sensing (ACRS 2019). Daejeon. Retrieved from https://www.researchgate.net/public ation/357620992_Comparison_of_Google_Earth_Engine_GEE-based_Machine_Learning_C lassifiers_for_Mangrove_Mapping Kamusoko C (2022) Geospatial machine learning in urban environments: challenges and prospects. In: Springer geography. Springer Science and Business Media Deutschland GmbH, pp 1–24. https://doi.org/10.1007/978-981-16-5149-6_1 Karim Z, Zyl Van TL (2021) Deep/Transfer learning with feature space ensemble networks (FeatSpaceEnsNets) and average ensemble networks (AvgEnsNets) for change detection using DInSAR Sentinel-1 and Optical Sentinel-2 satellite data fusion. Remote Sens 4394.https://doi. org/10.3390/rs13214394 Liu X, Hu G, Chen Y, Li X, Xu X, Li S, Wang S, et al (2018). High-resolution multi-temporal mapping of global urban land using Landsat images based on the Google Earth Engine Platform. Remote Sens Environ 209:227–239. https://doi.org/10.1016/j.rse.2018.02.055. Massachusetts Institute of Technology. (2021, May). Making informal urban spaces visible. Retrieved July 14, 2022, from https://dusp.mit.edu/news/making-informal-urban-spaces-visible Matarira D, Mutanga O, Naidu M (2022) Google Earth Engine for informal settlement mapping: a random forest classification using spectral and textural information. Remote Sens 14(5130). Retrieved from moz-extension://7673353a-629e-463d94d6-b19836eb3e6a/enhanced-reader.html?openApp&pdf=https%3A%2F%2Fmdpires.com%2Fd_attachment%2Fremotesensing%2Fremotesensing-14-05130%2Farticle_ deploy%2Fremotesensing-14-05130.pdf%3Fversion%3D1665735135 Maula FK, Choerunnisa DN, Akbar R (2019) Mapping informal settlements using geospatial method. In: Earth and environmental science, p 385. Retrieved from moz-extension://7673353a-629e-463d-94d6-b19836eb3e6a/enhancedreader.html?openApp&pdf=https%3A%2F%2Fiopscience.iop.org%2Farticle%2F10.1088%2F17551315%2F385%2F1%2F012035%2Fpdf Mboga N, Persello C, Ray Bergado J, Stein AI (2017) Detection of informal settlements from VHR images using convolutional neural networks. Remote Sens 9(1106). https://doi.org/10.3390/rs9 111106 Moya L, Garcia F, Gonzales C, Diaz M, Zavala C, Estrada M, Adriano B, et al (2022) Brief communication: radar images for monitoring informal urban settlements in vulnerable zones in Lima, Peru. Nat Hazards Earth Syst Sci 22:65–70.https://doi.org/10.5194/nhess-22-65-2022 Onyango F (2019) UAV mapping for informal settlement’s planning. Retrieved July 14, 2022, from http://www.afgoesdigital.com/uav-mapping-for-informal-settlements-planning/ Pokhriyal N, Jacques DC (2017) Combining disparate data sources for improved poverty prediction and mapping. Proc Natl Acad Sci USA 114(46):E9783–E9792. https://doi.org/10.1073/pnas.170 0319114 Samper J, Shelby JA, Behary D (2020) The Paradox of informal settlements revealed in an ATLAS of informality: findings from mapping growth in the most common yet unmapped forms of urbanization. Sustainability 12(9510):1–25. https://doi.org/10.3390/su12229510 Schoch M, Lakner C (2020) The number of poor people continues to rise in Sub-Saharan Africa, despite a slow decline in the poverty rate. UNDP

82

F. N. Karanja and P. W. Mwangi

Taubenbock H, Esch T (2011) Remote sensing—An effective data source for urban monitoring. Retrieved June 19, 2022, from https://earthzine.org/remote-sensing-an-effective-data-sourcefor-urban-monitoring/ UNCTAD (2021) Facts and figures. Retrieved from https://unctad.org/press-material/facts-and-fig ures-7 UNESCAP & UN-HABITAT (2008) Quick guides for policy makers 2. Low-Income Housing, Thailand UN-HABITAT (2006) Slums: some definitions UNSTATS (2019) Make cities and human settlements inclusive, safe, resilient and sustainable. https://unstats.un.org/sdgs/report/2019/goal-11/ Wang P, Huang C, Brown de Colstoun EC, Tilton JC, Tan B (2017) Global human built-up and settlement extent (HBASE) dataset from landsat, v1: Global high resolution urban data from landsat. Retrieved July 11, 2022, from https://sedac.ciesin.columbia.edu/data/set/ulandsat-hba se-v1 WHO (2018) WHO housing and health guidelines Wu M, Yan B, Huang Y, Nazirul M, Sarker I (2022) Big data-driven urban management: potential for urban sustainability. Land 11(680). https://doi.org/10.3390/land11050680 Xia N, Cheng L, Li M (2019) Mapping urban areas using a combination of remote sensing and geolocation data. Remote Sens 11(1470):23. https://doi.org/10.3390/rs11121470 Zhai W, Han B, Cheng C (2020) Evaluation of Luojia 1–01 nighttime light imagery for builtup urban area extraction: a case study of 16 cities in China. IEEE Geosci Remote Sens Lett 17(10):1802–1806. https://doi.org/10.1109/LGRS.2019.2955496 Zhang X, Liu L, Wu C, Chen X, Gao Y, Xie S, Zhang B (2020) Development of a Global 30-m impervious surface map using multi-source and multi-temporal remote sensing datasets with the Google earth engine platform. Earth Syst Sci Data 1–27.https://doi.org/10.5281/zenodo.350 5079

Chapter 4

Impact of Uncontrolled Tourism Development on Landscape Ecology of Purba Medinipur Coastal Region, West Bengal: A 4-C Framework and SWOC Analysis Manishree Mondal, Rabin Das, Chayon Chakraborty, Puja Karmakar, and Sk. Mustak

Abstract Digha-Khadalgobra Census Town (Census of India, https://censusindia. gov.in, 2011) area of Purba Medinipur district is one of the most popular beach destinations of West Bengal as well as India which has a low gradient and a shallow finest sand beach with gentle waves. Although this coastal tourism is the only major driving economy of this area, throughout time, it has been faced with the newer trend in urbanization transforming the rural coastal landscape. Consequently, the natural landscape gives way to concrete constructions in the name of upgraded coastal tourism planning and rurbanization (conversion from rural to urban). Destruction of dune ecology, the quick decline of the forest cover, deterioration of coastal wetlands, infrastructural sprawling, the decline of the coastal biodiversity, coastal erosion, illegal land use and encroachment, decline of sweet water resources, land pollution, soil degradation, plastic pollution, overfishing, increase in crime, traffic congestion, resort congestion, shops and vendor congestion, tourists harassment, declining of the M. Mondal (B) · C. Chakraborty · P. Karmakar Department of Geography (UG & PG), Midnapore College (Autonomous), Midnapore, West Bengal, India e-mail: [email protected]; [email protected] C. Chakraborty e-mail: [email protected] P. Karmakar e-mail: [email protected] R. Das Department of Geography (UG & PG), Bajkul Milani Mahavidyalaya, Midnapore, West Bengal, India Sk. Mustak Department of Geography, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_4

83

84

M. Mondal et al.

coastal beauty, etc. are the few identified problems which ultimately violate the CRZ policies. Although this place has been enlightened in many research and project works, it is not emphasized as the rurban landscape featured by twin functional processes of man-nature interaction like tourism development and coastal urbanization till date. Under this backdrop, this research project has been done through the extensive literature survey, in-depth field observation, crisscross interviewing, extensive data collection, broadly viewed data compilation, and statistical and mapping analysis by proper GIS and statistical software, with 4-C Framework and SWOC analysis to provide a blueprint for the smart tourism and sustainable coastal urban development planning maintaining the coastal ecological footprint of this area. Keywords Coastal landscape · Rurbanization · 4-C Framework · SWOC analysis · Smart tourism · Sustainable

4.1 Introduction Coastal areas are commonly defined as the interface or ecotone zones between land and sea, including large inland lakes (Biswas and Jana 2013). As any coast is an ecotone area so that the resources such as marine and wildlife both in developed and undeveloped conditions form a unique coastal ecosystem where both terrestrial and estuarine characteristics are merged. Coastal areas are extremely important for the social and economic welfare of present and future generations, as coastal resources support key subsistence economic activities (Annon 2005). The economy of any country, particularly a developing country has the great necessity for these primarylevel resources for the development of agro-based industries, infrastructure, natural oil extraction, and most important coastal tourism. Sustainable tourism means the eco-friendly co-existence between travellers, tour operators, and the local community (Allen et al. 1988; Yazdi 2012). These are the three major components of eco-tourism development. The interaction between these three components should be thoroughly investigated for the sustenance of any ecofriendly tourism (UNWTO 2004; UNWTO/UNEP 2005). The development of coastal truism enhances economic growth, infrastructural development, spending leisure time, and overall public awareness for better mass communications (EEA 2001, 2006). The studies in coastal tourism mainly give importance to tourism and its related economic activities rather than the related ecological problems (Wong 1993). West Bengal has a 157.5 km long coastline (Ministry of Home Affairs, Government of India 2013), with rich flora and fauna diversity. The whole area is under the norms of Coastal Regulation Zones (CRZ) (Ministry of Environment and Forests, Govt of India 2001). Digha-Shankarpur-Mondermoni area is one of the most popular beach destinations along the northern end of the Bay of Bengal in the state of West Bengal, India, which is featured by low height sand dunes, low vegetation, longshore currents, medium turbidity, and high salinity level (Paul 2002a, b). In 2011, more than twenty

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

85

lakh people visited this region showing the extended pulling capacity. Digha is situated 187 km away from Kolkata. The most attraction of Digha sea beach is its gentle slope with gentle waves. The length of this beach is 7 km. This shallow sandy beach with a casuarina plantation makes Digha more charming. It is ideal for bathing. This sector is in no way related to tourism, but the fact remains that this industry provides ample job opportunities (Mason 2003) for the local residents. Another important attraction of Digha is its variety sea foods, mostly fish. The loading and unloading of seafood are operated at Sankarpur and auctions of sea foods are operated from Mohona. Huge depletion of the natural marine as well as terrestrial resources can generate conflict between local people and industrial sectors such as fishing (Gossling 2003; Mc Laren 2003). It can also hamper the ecological setup through soil erosion, largescale deforestation of mangroves, etc. (Neto 2003). These activities will also invite major cyclonic hazards (Fritz 1961; Burton and Kates 1964; Cutter 1996; Quarantelli 1998; Kaiser 2006) which are now the regular climatic catastrophes like severe flooding, storm surges (Harris 1963), etc. in this study area. These disasters can badly damage the fragile ecosystem as well as coastal tourism (CI 2003).

4.2 Background The natural beauty of Digha has been paving its momentum since the twentieth century and now this finest sea beach destination has flourished tremendously, almost in its climax stage. But, the journey of Digha is not as satisfactory as eco-friendly development. Although coastal tourism (Ganguly and Sharma 2015) is the driving economy of this area, throughout time, it has faced the newer trend in urbanization transforming the rural landscape. The red crabs and their habitats are crushing under the wheels of development. Robbed of its natural constituents, the beach is already showing signs of narrowing down and erosion (Paul 2006). One of the major threats of this area is that the heavy poisonous litters and solid wastes particularly from the hotels find their way to the sea. About 16 Mouzas have been affected by coastal tourism cum urbanization enhancing huge environmental problems. Environmental degradation from different aspects has been a regular feature seen here. Destruction of dune ecology, the quick decline of the forest cover, deterioration of coastal wetlands, violation of CRZ policies, infrastructural sprawling, declining coastal biodiversity, coastal erosion, illegal land use and encroachment, water resource decline, land pollution, soil degradation, plastic pollution, over and illegal fishing, increase in crime, traffic congestion, resort congestion, shops and vendor congestion, tourists harassment, declining of the coastal beauty, etc. are the problematic scenario found here. All these factors have a great impact on the ecology, economy, society, and culture (Klein and Nicholls 1999). Although Digha is now a beach town and rural in character but more or less all types of urban characteristics are present here. Hence, Digha is now a tourism hotspot having newer urban experience with a remarkable number of problems and also

86

M. Mondal et al.

prospects. Against this backdrop, Digha, along with the Khadalgobra census town, has been selected as the study area because of its tremendous tourism development along with uncontrolled urbanization and landscape transformation. Digha rurban entity over the Bengal coast is undoubtedly a new dimension of coastal urbanization to the academicians, students, scholars, researchers, environmentalists, social workers, planners, administrators, local people, and well-wishers cum nature lovers and it has also emerged as an essential issue for research. Keeping all these above discussions, the objectives set for this study are as follows: 1. To estimate the site suitability of Khadalgobra-Digha tourism cum rurban entity and reality; 2. To enlighten the demographic and spatio-temporal changes in LULC in terms of landscape transformation; 3. To investigate the major causes for the development of the tourism industry and assess the role of tourism in landscape transformation; 4. To assess the major human and environmental costs due to tourism development; 5. To analyse the SWOC of the study area in case of its newer journey with tourism and urbanization and to provide suggestions towards the new momentum of the processes of smart tourism and sustainable journey.

4.3 Study Area Any study area indicates its geographical nature and characters with respect to its geology, geomorphology, climate, pedology, administration, and socio-economic background. Our study area also reflects a distinct character of geo-environmental scenario. Digha is a beautiful small beach resort in the East Midnapore district of West Bengal and situated at 21° 38, 18,, N latitude and 87° 30, 35,, E longitude and has a potential coastline of about 12 km of its own (Udaypur to Digha Mohana) (Figs. 4.1 and 4.2). Digha police station has an area of 3153 km2 and with a total population of 35,054 (Census of India 2011). Originally Digha is known as which is designated by Warren Hastings (1780) as ‘Brighton of the East (National Informatics Centre 2006) in a letter to his wife (O’Malley 1911). English tourist John Frank Smith came to Digha in 1923 and was charmed by its beauty. He lived here, and after independence, he proposed to Dr. Bidhan Chandra Roy, the first Chief Minister of West Bengal, to make it a tourist resort (DighaShankarpur Development Authority, Retrieved 20th Dec. 2020). Small Digha town is now crowded with hotels which is the main business at Digha. Although it is crowded with Beerkul and tourists throughout the year but winter is the peak season in this lovely sea holiday inn. Khadalgobra (Figs. 4.1 and 4.2) is a census town located between 21.62766° N latitude and 87.522178° E longitude with an area of 2.1858 km2 and a total population of 5344 (Census of India 2011). The urban population of this town is only 6.45% (District Census Handbook, Purba Medinipur District 2014; Hunter 1876).

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

87

Fig. 4.1 Location of the study area (Source Google Earth Image). Source Prepared by authors from Google Earth Image and ARC GIS 10.7 and administrative maps of various units

Environmentally this area consists of Coastal Rurban Landscape resulting from the ecological transformation of the rural environment to the urban one. Coastal tourism is the major driving factor for the urbanization of this region (Ghosh 1996). Geologically, it is the recent Quaternary Formation having coastal sedimentary and alluvial formation (6000–8000 BP) (Chakraborty et al. 2012; Dey et al. 2002;

88

M. Mondal et al.

Fig. 4.2 Major important spots dignifying tourism and urbanization in the study area, 2018 (Source Google Earth Image)

Niyogi 1970). Geomorphologically, it is the westernmost tip of Midnapore as well as the Bengal Coast which is included of Rasulpur-Pichhabani Sub-basin over the South Bengal Basin having the finest sedimentological character of beach formation (Das and Debnath 2014; Jana and Bhattacharya 2012; Paul 2002a, b). This area is under a tropically hot and humid monsoon maritime climate. It is frequently experienced by tropical cyclones in the monsoon and nor-wester in the pre-monsoon season. The coastal regions are influenced by land and sea breezes regularly. The soil is salty and sandy type under-covered by clayey soils. The climate, soil, and topography altogether make vegetation diversified in nature at Digha. The permanent and temporary salt and fresh water logging area carry a tremendous floral diversity. At Digha 81 plant species under 76 genera and 42 families of angiosperms are present. The five dominant families, as per the relative abundance (RA), studied are Asteraceae with 13.7, Fabaceae with 9.7, Scrophulariaceae with 5.4, Solanaceae with 8.6, and Verbenaceae with 9.7 value (Paul 1996; Das 2014; Jana 2016).

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

89

4.4 Materials and Methods Digha town and Khadalgobra census town have been selected as study areas for intensive field survey. Digha is an already setup town with urban characteristics whereas Khadalgobra is a rural area with urban prominences. These two urban and rural areas were selected to measure the acuteness of land use transformations and also assess the impact of these transformations on the livelihood pattern whatever it is—positive or negative from the end of local residents. The most important part of this study is to collect the information from the resident’s end. For this personal interviews were taken with the help of an open-ended semi-structured questionnaire schedule among 289 respondents selected through a random sampling method. Various interactive group discussion sessions were also held. Land use and land cover maps had been prepared by using GIS software ARC GIS 10.7 and satellite images to show the land use transformations. Primary and secondary data had been presented through various statistical and cartographic mapping techniques. A SWOC analysis had also been carried out to assess the strength and weaknesses of these land use changes and also to predict the opportunities and challenges of future coping strategies. A 4-C Framework had been come out for the future planning of sustainable development of smart tourism without hampering the natural environment. The Land Use and Land Cover (LULC) of this area has been changed drastically due to the huge mushrooming of tourism amenities such as hotels, goods, and parks shops with the total ignoring of rural ecology which were quantified using statistical techniques to understand change. All tourismcentric developmental activities have been done to make this place an economic giant (Mukherjee 2009). The tourism industry obviously invites rapid urbanization which indiscriminately encroaches on the natural environment of this area quantified using statistical techniques. Like other regions, urbanization builds its empire on the sacrifice of environmental quality. In this study, the thrust has been developed on whether the interaction between the tourism industry and the environment of this area will be symbiotic or not.

4.5 Result and Discussion 4.5.1 Changing Population Scenario The above diagrams (Figs. 4.3 and 4.4) reflect the changing scenario of the rurban population of Khadalgobra Census Town and Khadalgobra-Digha Rurban Entity. In both cases, tourism is the prime cause of this crowding. On one hand, the immigration of the rural people of surrounding regions occurs for the opportunities of residence and employment, and on the other, the invasion of outsiders is also found for economic ventures. All these activities are the major causes of overcrowding in this small tourism destination. In the meantime, the State Government took various initiatives

90

M. Mondal et al.

Figs. 4.3 and 4.4 Changing population scenario in Khadalgobra Census Town and KhadalgobraDigha urban region (Source Census of India 2011; DSDA 2020)

for the development of this region and the population intake capacity now crossing the limit.

4.5.2 Changing LULC Scenario Due to Tourism in the Study Area To study the spatio-temporal changes in Land Use and Land Cover (LULC), Google Images from 2002 to 2018 and LANDSAT-8 specifically have been taken as the base images (Fig. 4.5). From those databases, comprehensive LULC along with DEM variation, drainage and water bodies’ variation, variation in vegetation cover, settlement, resort, road network, etc. have been prepared to compare different LULCs throughout time. The year 2002 has been selected as the base year to show the nature and status of the pre-urban scenario of the study area. The 2010 year has been selected because this is the year when the precondition of the study area has been set for being declared as the census town, and finally, the 2018 year has been selected for comparative study from the viewpoint of LULC in the study area (Figs. 4.6, 4.7 and 4.8). The above diagrams show the LULC variations in the Study Area from 2002 to 2018. The compiled data reflects the variation of different land uses and land covers here. It has been seen that different anthropogenic features and land use have increased over time but different physical features like water bodies including drainage and wetlands (Figs. 4.15 and 4.16) beach sectors, dune sectors, (Figs. 4.17 and 4.18), settlement (Fig. 4.10), etc. have been declined with time. The most interesting fact has been derived from figure no 9 that the area under vegetation has been increasing since 2000 but the species diversity has been gradually decreasing. The decreasing trend of agricultural land has also been found (Fig. 4.14) whereas the tremendous growth of built-up areas like settlements, hotels (Fig. 4.12), market facilities (Fig. 4.11), transportation (Fig. 4.16), and other institutional facilities has

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

91

Fig. 4.5 Google Earth Images: 2002, 2010 and 2018 (Source Google Earth Image)

been observed. All these above factors show the fabulous growth of urban tourism in this area. The following CRZ-III LULC analysis (map) and longitudinal profile (Figs. 4.19 and 4.20) prepared based on GPS survey, landscape survey, and Google image analysis reflects the landscape scenario in the study area. Most of the survey and derived

92

M. Mondal et al.

Fig. 4.6 LULC in 2002 (Source Google Earth Image)

profiles show the detailed LULC scenario relating to their location and the elevationbased situation in the study area. Interestingly, CRZ-I and III have been tremendously affected by tourism-related activities and architecture in the study area. Not only that the most sensitive coastal forest, wetlands, dune sector, and beach zone have been acutely changed. These critical natural resources of this fragile ecosystem of the study area are threatened due to the development of the tourism industry and its associated urban processes.

4.5.3 Major Causes for Tourism Development and Tourist Flow The above diagram shows the major causes of tourism and its development through the landscape transformation in the study area (Fig. 4.21). A perception survey has been conducted through a structured questionnaire schedule to assess the major driving indicators and responsible causes for the flourishing of the tourism industry. The size of respondents was 289, selected by random sampling method from different backgrounds of the tourism sector like market, resorts, hotels, residents, tourists, official/institutional, refreshments, amusements, recreational, etc. But, in most of

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

93

Fig. 4.7 LULC in 2010 (Source Google Earth Image)

the cases, the respondents put their perception on the caption of very high and high magnitude of different roles/indicators. This scenario reflects the acute tourism development in the study area. A field survey has been executed by observation method for the identification of the major or minor tourist spots in this region. These tourist spots are shown in Figs. 4.22 and 4.23. About fifteen such spots have been identified like science city, particularly for students, marine aquarium, Amarabati park, etc. Figure 4.24 depicts the number of people who visited from 2007 to 2019. Digha is now becoming the first choice of people’s weekend destination.

4.5.4 Roles of Tourism for Landscape Transformation in the Study Area The above diagram (Fig. 4.25) shows the role of tourism in landscape transformation in the study area. But, in most of the cases, the respondents put their perception on the caption of very high and high magnitude of different roles/indicators. This scenario reflects the acute influence of tourism and its development in the study area. This is one of the most critical parts of the study. The ecosystem of this area is not yet

94

M. Mondal et al.

Fig. 4.8 LULC in 2018 (Source Google Earth Image)

Figs. 4.9 and 4.10 detection maps)

Trend of vegetation change and settlement change (Source LULC change

stable situation. It is very fragile. Any kind of disturbance can generate a devastating situation. Those areas which are not occupied by any anthropogenic activity have encroached in a very unscientific and unplanned way.

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

95

Figs. 4.11 and 4.12 Trend of market area change and resort area change (Source LULC change detection maps)

Figs. 4.13 and 4.14 Trend of transport area change and agricultural and vegetable land change (Source LULC change detection maps)

Figs. 4.15 and 4.16 Trend of drainage and wetland change and institutional area change (Source LULC change detection maps)

4.5.5 Problem and Management Scenario for Tourism in Study Area Major physical environmental problems due to tourism and urbanization The above diagram (Fig. 4.26) shows the major physical environmental problems observed due to tourism development and rurbanization in the study area. Decrease in

96

M. Mondal et al.

Figs. 4.17 and 4.18 Trend of dune area change and beach area change (Source LULC change detection map)

Figs. 4.19 and 4.20 LULC along the CRZ-III & longitudinal landscape profile showing the tourism and urban development in the study area, 2018 (Source Google Earth Image)

local environmental resources, a huge decline in coastal forest, a huge decline in dune segment, a decrease in local species and regional biodiversity, a decrease in wetlands, drainage interruption, soil and land degradation, water pollution and degradation, and decrease in fish/aquatic resource, etc. are the major physical problems developed in the study area. Interestingly, in most of the cases, the responses on very high and high magnitude of the qualitative scale reflect the gravity of problems which seriously hits the physical environment directly in the study area. Major anthropogenic problems created due to tourism and urbanization: The following diagram (Fig. 4.27) depicts the major anthropogenic problems generated due to tourism development and rurbanization in the study area. Drastically

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

97

Fig. 4.21 Major causes for tourism development in the study area (Source Field survey)

Figs. 4.22 and 4.23 Major important tourist spots and average tourists at those places throughout the year (Source Google Earth Image and administrative maps)

changes in LULC, health and education-related problems, drinking water problems, lack of employment opportunities for local people rather than outsiders, emphasis on the only tourism industry rather than other promising sectors of development by govt., the invasion and dominancy of outside entrepreneurs in the region, lacking proper implementations of different govt. schemes, lacking proper sanitation facilities in the region, deficiency of proper waste dumping sites in the region, promoter dominancy in housing, construction, land business, illegal and frequent collection by protractor, lack of proper development of transport network towards remote areas, increase in different kinds of pollution and degradation, etc. are the major physical problems developed by aforesaid two processes. Interestingly, in most of the cases,

98

M. Mondal et al.

Fig. 4.24 Daily tourist flow in the study area (from 2007 to 2019) (Source Ministry of Tourism, Govt. of West Bengal 2020)

Fig. 4.25 Roles of tourism for landscape transformation in the study area (Source Field survey)

the responses on very high and high magnitude of the qualitative scale have been reflected as the gravity of problems faced by the people in the study area. Management efforts from government, administration, and NGO’s end: Now it is time to see the management efforts to mitigate these serious problems taken by the government and non-government levels. It is a general tendency that the sudden growth of any kind of industry invites greediness and illegal activities which can hamper first the natural environment and later on the socio-economic and the entire cultural environment. This very fact is also observed in the study area (Figs. 4.28 and 4.29). The tourism industry is now in a growing and promising stage in the DighaShankarpur area. This place offers tremendous prospects in this industry. Government

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

99

Fig. 4.26 Major physical environmental problems existed in the study area (Source Field survey)

Fig. 4.27 Major anthropogenic problems in the study area (Source Field survey)

100

M. Mondal et al.

administrations of higher to lower order are heartily trying to develop this area as one of the major tourist destinations in India but side by side they are not trying to save and manage the sustainability of nature and natural recourses. It does not mean that they are doing nothing. They are trying to mitigate and recover the issues like disaster management, coastal erosion management, proper infrastructural development, etc. But the vulnerability issues like building construction, road construction, unplanned land use transformation, waste disposal, and beach recreational activities are the prime concerns to be managed. Management gaps between problems and efforts: The survey shows (Fig. 4.30) the management gaps between the problems and efforts to manage those in the study area. Due to unethical and unplanned urbanization and tourism, different kinds of physical and anthropogenic environmental problems, issues, and changes are observed in the

Fig. 4.28 Different roles of govt. and higher level administration for management of the problem. Source Computed by authors from field survey

Fig. 4.29 Different roles of local-level administration, NGO, and other institutions for the management of the problems and issues (Source Field survey)

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

101

Fig. 4.30 Gap between existing problem and management efforts (Source Field survey)

study area. The survey depicts that the management procedures are traditional and insufficient. Tourism cum commercial atmosphere and urbanized environment has been flourished tremendously at the cost of mining of the bases of local environment and ecology. The local people are not capable to compete with resourceful outsiders who enjoy the benefits of such a growing industry as tourism. Govt. has played an important role as the stimulator of tourism development by the acceleration of infrastructural and service-related developments. But, these types of efforts are not found in the case of protecting the environment and ecology. Other non-governmental efforts are also not activated to save these potential landscape resources. Few gaps in managemental levels have been identified through the survey. These are the gap between government and local administration; the gap between planner and policy maker; policy maker and public; the gap between government project and local demands/ expectations; the gap between local need and outsiders’ greed; the gap between society, development, planning, and environment; the gap between knowledge about the place, problem, and potentiality; the gap between reality and sustainability; etc. which are reflected as the weakness from the viewpoints of tourism development and smart urbanization. Interestingly, the perception study shows that in most of the cases, the responses on the very high and high magnitude of the qualitative scale have been reflected as the gravity of problems in the study area which refers to the huge gap in managemental efforts of the problems, issues, and challenges in the study area.

4.5.6 SWOC Analysis (Strength-Weakness-Opportunity-Challenges) SWOC analysis is a very important tool to crosscut the merits and demerits of a system. Strength and opportunity are the positive aspects whereas weakness and

102

M. Mondal et al.

challenges are the negative aspects. This technique is used for strategy formulation and decision-making. This method is also very suitable for environmental analysis. The SWOC has been justified based on the intensive field observation and feedback of 289 respondents from households, markets, resorts, hotels, tourists, academicians, experts, and researchers. Major Strengths: The survey shows the major strengths that existed in the study area for the well development of the tourism industry and urbanization in this region. There are observed different types of strengths that are very much responsible for tourism development and rurbanization in the study area. Scenic natural beauty having seasand-sun scenario, fluvial-coastal site suitability, huge local natural resources, govt. and administrative schemes and facilities, self-governance for development, local to global interest, initiatives from local and outsider entrepreneurs, available and sufficient buffer and hinterland, historical background and initiatives, recent research and project works towards its sustainable development (UNEP 2009), etc. are the major strengths for developing the aforesaid two processes. Interestingly, in most of the cases, the responses on the very high and high magnitude of the qualitative scale have been reflected as the gravity of strengths in the study area which refers to the developing scenario of tourism cum urbanization. Major Weaknesses: In the same way, the major weakness is depicted in the perception study. The study reflects the major weaknesses that existed in the study area for uncontrolled tourism development and urbanization. There are different types of weaknesses which are very much responsible for tourism development and rurbanization in the study area. Geological and environmental sensitiveness, bio-geomorphic ecotone situation and its natural and anthropogenic stress, tourism affected by rurbanization and sprawling in terms of quick change in LULC, decreasing the quantity and quality of resources by its abuse, misuse, and overuse, need to greed: competition in growth and development, lacking the knowledge about cost–benefit of the processresponse, etc. are the major weakness here. Interestingly, in most of the cases, the responses on the very high and high magnitude of the qualitative scale have been reflected as the gravity of weakness in the study area which refers to the developing scenario of tourism cum urbanization. Major Opportunities: Through the field investigation and study, there are observed different types of opportunities that are very much responsible for tourism development and rurbanization in the study area. Resource potentiality having the coastal interfluves of River Subarnarekha and Pichhabani, scope to optimum and sustainable uses of local resources, childhood journey of the urban cum growth centre, leading character to eco-tourism and smart tourism under DSDA, govt. initiatives over time, recent efforts from different levels for its development and protection, awareness improvement and advancement of local people, etc. are the major opportunities for developing the aforesaid two processes. Interestingly, in most of the cases, the responses on the very high and high magnitude of the qualitative scale have been reflected as the gravity of opportunities in the study area which refers to the developing scenario of tourism cum urbanization.

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

103

Major Challenges: From the survey, the major challenges are shown as very much responsible factors to future tourism development and rurbanization in the study area. Assimilation and synchronization of people, policy, plan, and programme, balancing the tourism and urbanization, these twin processes, covering whole landscape by ICT facility for making the smart entity, developing smart infrastructure, smart governance, smart business, smart citizen, smart society and smart environment, expensiveness for smart tourism and smart urbanization, establishing the sustainability by communicating the public and private sector both, providing politics and promoter free egalitarian development, etc. are the major opportunities for developing the aforesaid two processes (TERI 2011). Interestingly, in most of the cases, the responses on the very high and high magnitude of the qualitative scale have been reflected as the gravity of opportunities in the study area which refers to the developing scenario of tourism cum urbanization. SWOC Index (SWOCI) Analysis Table 4.1 reflects the SWOC Index analysis to assess the tourism potentiality cum prospect against the resistance and challenges here. The SSI and OSI show the indices as 3.8 and 3.4 dignifying the large potentiality of tourism whereas WSI and CSI having the indices as 4.0 each drawback the mass momentum of tourism development. Not only that weakness and challenges decelerate the strength and opportunity due to unplanned, unscientific, and haphazard development of tourism cum urban landscape. SWOC Index of 40% shows the significant average variation between resistance and potentiality where weakness and challenges are significantly more dominant than strength and opportunity which depicts the challenging journey of tourism townscape.

4.5.7 Essential Dimensions for Smart Tourism and Challenges in the Study Area Smart tourism is the result of the integrated collaboration of ICT, innovation, accessibility, and sustainability (Junior et al. 2017). The ICT sector has to be the string to facilitate the national as well as international tourists for providing accurate information at the correct time. As Digha is growing as the fastest affordable tourism destination, therefore, it should be grown. For this, ICT will play a crucial role in developing Digha tourism as a smart tourism destination not only in West Bengal but also in India and abroad. The help of efficient ICT discoveries, ideas, and effective accessibility can make Digha a sustainable smart tourism destination. Smart Tourism Technological Index (STTI) analysis See Table 4.2.

4 3 4

Fluvio-coastal site suitability and huge local natural resources

Govt. and administrative schemes and facilities

Self-governance by DSDA for planning and development

4 4 4 4

4

Geological and environmental sensitiveness to anthropogenic development

Tourism affected by rurbanization and sprawling in terms of quick change in LULC

Decreasing the quantity and quality of resources by its abuse, misuse, and overuse

Need to greed: unplanned efforts, haphazard growth and unscientific development in terms of quick and smart tourism

Lacking the knowledge of the cost–benefit of the process-response to tourism development

Weakness

Local to global interest for capital investment and 3 development

5

Grade on 5-point scale

Scenic natural beauty having sea-sand-sun scenario

Strength

Dimension-based indicators/aspects

Table 4.1 Determination of the SWOC Index

20

19

Total on scale

Weakness Specific Index (WSI) = 4.0

Strength Specific Index (SSI) = 3.8

Dimension Specific Index (DSI) 72

Tourism Development and Potentiality Index (TDPI) in % 80 = 40

× 100

(continued)

{(SSI∼WSI)+(OSI∼CSI)} 2

Tourism SWOC Index (SWOCI) in % Resistance Index (TRI) in %

104 M. Mondal et al.

4 3 3

Childhood journey of the urban cum growth centre

Leading efforts towards eco-tourism and smart tourism under DSDA and Govt. initiatives

Recently protecting efforts from different levels of well-wishers

4 4

4

Balancing tourism and urbanization, these twin processes

Covering the whole landscape by ICT facility developing the smart infrastructure, smart governance, smart business, smart citizen, smart society, and smart environment

Expensiveness for smart tourism and smart urbanization

Source Perception survey

Establishing the sustainability by communicating 4 the public and private sectors both

4

Assimilation of people, policy, plan, and programme

Challenges

Improvement and advancement of local people to 4 understand its development and impacts

3

Grade on 5-point scale

Optimum and sustainable uses of local resources

Opportunities

Dimension-based indicators/aspects

Table 4.1 (continued)

20

17

Total on scale

Challenge Specific Index (CSI) = 4.0

Opportunity Specific Index (OSI) = 3.4

Dimension Specific Index (DSI)

Tourism Development and Potentiality Index (TDPI) in %

Tourism SWOC Index (SWOCI) in % Resistance Index (TRI) in %

4 Impact of Uncontrolled Tourism Development on Landscape Ecology … 105

106

M. Mondal et al.

Table 4.2 Technological foundations of smart tourism Major technological foundations

Details in smart tourism

Status in the study area

Grade on a 5-point scale

Smart Tourism Technological Index (STTI) in %

Sphere

Bridging digital and physical spheres

Partly

2.5

40

Core technology

Sensors and smartphones

Partly

2.0

Travel phase

During trip

Partly

2.5

Lifeblood

Big data

A little bit

1.0

Paradigm

Technology-mediated co-creation

A little bit

1.5

Structure

Ecosystem

Not considerable

1.5

Exchange

Public–private-consumer collaboration

Partly

3.0

Source Perception survey

Smart Tourism Index (STI) analysis The following indices indicate the status of technological foundations and essential agenda for smart tourism development in the study area. From the compilation of different primary and secondary data from observation, field measurement, perception study, research base, and official sources, it is seen that STTI is at poor level (0.4/ 40%) showing the initial state of technological foundations for smart tourism (Table 4.2). On the other hand, agenda analysis for smart tourism development assessment in the study area reflects poor progress also. STI is only 38% which assesses the juvenile situation of the smart tourism journey here (Table 4.3).

4.5.8 Main Challenges that Tourist Destinations Faced On 1. Adoption of new technologies to attract new tourists to the destination. 2. Improve efficiency in managing the destination mitigating the pressure of tourism on city services. 3. Promote co-existence between citizens and tourists and boost trade through segmentation and communication strategies. 4. Influence the behaviour of visitors and promote tourism flows throughout the year avoiding seasonality. 5. Encourage tourists to repeat the destination and act as prescribers. 6. Recommendations of Coping Strategy for Development of a Sustainability Approach towards Smart Tourism cum Smart Urbanization: A 4-C Framework The following table (Table 4.4) represents a comprehensive coping strategy

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

107

Table 4.3 Smart tourism agenda analysis Smart tourism dimension

Smart tourism indicators

Status

Grade on a 5-point Scale

Smart Tourism Index (STI) in %

Utilization

Privacy

Partly

2.0

38

View towards co-development

A little bit

1.0

Utility

A little bit

1.0

Service provision

Facilitation

Connectivity and network Not considerable

0.5

Aspiration to be free from A little bit technology

1.5

Technological accessibility

Partly

2.0

Information quality

A little bit

1.5

Usable technology and market facilities

A little bit

1.0

Appropriate business facilities

Partly

2.5

Innovation capabilities

Partly

2.0

Skilled human resources

Partly

2.5

Collaboration and coordination

Partly

2.0

Market dynamics

Partly

2.5

E-governance

Partly

2.0

Infrastructure

Partly

2.5

Socio-environmental cost

Partly

2.0

Mechanical intelligence

Partly

2.0

Public–private-consumer collaboration

Partly

3.0

Structural–functional interlinkage and facility

Partly

2.5

Cost–benefit tourism and valuation of ecosystem

Partly

2.0

8/30

14/35

16/35

Source Perception survey

for the development of smart tourism and smart urbanization maintaining the sustainability.

Causal investigation of common and root causes, process-response parameter, and variables

Consequence assessment through causality analysis, human and environmental cost assessment, and findings of the managemental efforts along with managemental gaps

Coping strategy through quantity and quality mapping, thinking towards recovery reminding reality, and planning the programmes for the public to prime characters Making the blue print linking the people, politician, planners, and plan implementer

Project and plan: manipulation → method fixation → measurement → manipulation → modelling → monitoring → making the decision and plan

Change detection and analysis of morphology, topology, LULC, ecology, livelihood—lifestyle, and man-nature relationship with interactions

Coping strategy for the development of a sustainability approach towards smart tourism

Table 4.4 Coping strategy for the development of a sustainability approach towards smart tourism

108 M. Mondal et al.

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

109

4.5.9 Ten Steps on the Way Forward Against Tourism Cum Urban Sprawling • Formulation of new tourism townscape with full consideration of local vies and demands. • Tourism will be sustainable only with the inclusion of local residents in the decision-making process. • Urban mobility should be for everyone. • Sustainable construction processes, buildings, and maintenance. • Emphasis should be given to the use of more renewable energy sources to prevent pollution levels as well as to enhance energy security. • Valuing local skills and non-market-based solutions. • Financing should be adequate and on time. • Measuring success and sharing data and knowledge. • Cities proactive in a globalized world. • Towards a culture of sustainability.

4.6 Conclusion The study areas Khadalgobra Census Town and Digha urban area are famous tourist destinations in South Bengal Based in West Bengal. Under DSDA this is the hotspot of tourism development influencing other coastal tourist spots on the Midnapore coastal stretch. The natural scenic beauty of all those spots/tourist pockets supports the tourism industry along with the socio-economic development of the region. The human cost of tourism flourishment is mostly positive in sense. But, far-sighted negative costs from the human and environmental background are very much more important than short-term socio-economic gain. Economic growth cum development is undoubtedly important for regional economy and development through tourism and urban processes. Hence, undermining and dwindling scenario of the potential environment through urbanization and advanced tourism affects local society and environment directly and indirectly. Urban morphology, urban economy, urban ecology, urban typology, and dynamics of urban landscape are not developed and maintained properly. The growth of this census town and the urban area has been progressing breaking down the urban policy, and the thought of smart urbanization is far away from its reality. On the other hand, tourism, this base economy, and the process are being advanced by the acceptance and implementation of different government schemes since 2011 here. Recent govt. has also given new momentum to its journey over time. Side by side, due to this, the regional environment and ecology have deteriorated consequently. Coastal, forest, dune, wetland, beach, and other ecologically and geomorphologically sensitive features of this coastal landscape have been affected quantitatively and qualitatively throughout time. The thoughts of eco-tourism and smart tourism have been far sighted dreams for the area from the

110

M. Mondal et al.

viewpoint of well development of the tourism industry here. Recently, privatization, liberalization, politicization, and capitalism in tourism development deeply affect the local resources and socio-economic development of the region. In this perspective, tourism and urbanization are the ways to coastal development, but a huge challenge to local society, development and the environment. So, this problem regarding tourism and urbanization is the backbone to conduct the study in the region. The unplanned and unscientific growth of tourism in Digha welcomes various, severe, ecological threats although tourism is the backbone of the economy of the region. For the sustenance of Digha tourism, all the stakeholders should keep a balance between the economy and ecology. Another major important problem for the development of Digha as a smart tourism destination is the quality of ICT infrastructure, particularly the Internet speed. If the loopholes can be rectified and the gap between ecology and economy conflict may be minimized, then Digha will become a major smart tourism destination in India.

References Allen LR, Long PT, Perdue RR, Keselbach S (1988) The impact of tourism development on residents’ perceptions of community life. J Travel Res 27:16–21 Annon (2005) Studies on bio-resource assessment and management of degraded mangrove ecosystem of Midnapore coast, West Bengal. Research Project Report, Ministry of Environment & Forest, Govt. of India (Sanction No. 3/6/2001-CSC-M, Dated-5th Nov 2000) pp 1–99 Biswas A, Jana A (2013) Shoreline changes in response to sea level rise along Digha coast, eastern India: an analytical approach of remote sensing, GIS and statistical techniques. J Coast Conserv Burton I, Kates RW (1964) The perception of natural hazards in resource management. Nat Resour J 3(3):412–441 Census of India (2011) https://censusindia.gov.in Chakraborty K, Chatterjee D, Das A (2012) Provisions and perceptions of the hosts in business at Digha. Antrocom Online J Anthropol 8(1) CI (2003) Tourism & biodiversity, mapping tourism global footprint: the driver of coastal development. U.S. Federal Agencies with ocean related programmes, Washington, DC Cutter SL (1996) Vulnerability to environmental hazards. Prog Hum Geogr 20(4):529–539 Das R (2014) An analytical study on the phyto resources and vegetation ecology of coastal Medinipur of West Bengal in India. IJSR 3(10):240–249 Das R, Dandapath P (2014) Existence and experience of Purba Medinipur coastal belt on its morpho-dynamic journey with the distinctive geology and geomorphology. Int J Sci Res (IJSR) 3(6):1242–1251 Dey S, Dutta S, Adak SB (2002) Holocene sea level change of West Bengal coast. Indian Geogr J 77:7–20 Digha-Sankarpur Development Authority (DSDA). Annual Report-2008–2009 Digha-Sankarpur Development Authority (DSDA) (1997) Land use and development control plan 1995–2011. Urban Development Department, Govt. of West Bengal District Census Handbook (2014) Purba Medinipur District Digha Sankarpur Development Authority (2020) Available at https://www.dsda.org.in/ EEA (2001) Environmental signals 2001. EEA, Copenhagen EEA (2006) The changing faces of Europe’s Coastal areas, EEA 6/2006. European Environment Agency, Copenhagen

4 Impact of Uncontrolled Tourism Development on Landscape Ecology …

111

Fritz CE (1961) Disasters. In: Merton RK, Nisbit RA (eds) Contemporary social problems: an introduction to the sociology of deviant behaviour and social disorganization. University of California Press, Riverside, pp 651–694 Ganguly P, Sharma S (2015) Sustainable development of coastal tourism in Digha, West Bengal: an investigation of local residents attitude. IJLTEMAS 4(III):34–44 Ghosh SK (1996) Kanthir Purabritta (Regional History of Kanthi). midnapore.in Gossling S (2003) Market integration and ecosystem degradation: is sustainable tourism development in rural communities a contradiction in terms? Environ Dev Sustain 5:383–400 Harris LD (1963) Characteristics of the Hurricane storm surge. U.S. Weather Bureau. Technical paper No. 48 Hunter WW (1876) A statistical account of Bengal (District-Midnapore) I-WIN Advisory Service Limited & DSDA (2013–14) Digha-Sankarpur Integrated Beachfront Development Plan, Report No.: I WIN/13-14/FR/RO/006 Jana B (2016) Study on sand dune vegetation in East Midnapore district, West Bengal, India. IJSRD 4(05):1662–1666 Jana A, Bhattacharya AK (2012) Assessment of coastal erosion vulnerability around MidnapurBalasore Coast, Eastern India using integrated remote sensing and GIS techniques. J Indian Soc Rem Sens 41(3):675–686. https://doi.org/10.1007/s12524-012-0251-2 Junior AS, Filho LM, GarcIa FA, Simoes JM (2017) Smart tourism destinations: a study based on the view of the stakeholders. Revista Turismo Em Analise 28(3):358–379 Kaiser G (2006) Risk & vulnerability analysis to coastal hazards—an approach to integrated assessment. PhD thesis, Christian Albrecht University, Kiel, Germany Klein RJT, Nicholls RJ (1999) Assessment of coastal vulnerability to climate change. Ambio 2:182– 187 Mason P (2003) Tourism impacts, planning and management. Butterworth Heinemann, London McLaren D (2003) Rethinking tourism and ecotravel, 2nd edn. Kumarian Press Inc., CT Ministry of Environment and Forests, Govt of India (2001) Coastal Regulation Zones (CRZ). https:/ /parivesh.nic.in Ministry of Home Affairs, Government of India (2013) Loksabha Starred Question No. 498, 30th April. https://www.mha.gov.in Ministry of Tourism, Govt of West Bengal (2020). https://wbtourism.gov.in Mukherjee R (2009) The changing face of Bengal: a study in riverine economy. University of Calcutta, 2008–09 National Informatics Centre Archived, 17th Feb, 2006. Accessed 2 April 2020 Neto F (2003) A new approach to sustainable tourism development: moving beyond environmental protection. Nat Resour Forum, 3(27):212–222 Niyogi D (1970) Geological background of beach erosion at Digha, West Bengal, India. Bull Geol Min Metall Soc 43:1–36 O’Malley LSS (1911) Bengal District Gazetteers: Midnapore Paul AK (1996) Degradation of coastal vegetation in West Bengal. Indian J Landsc Ecol Stud 19(1) Paul AK (2002a) Coastal geomorphology and environment: Sunderban coastal plain, Kanthi coastal plain and Subarnarekha delta plain. ACB Publication, Kolkata Paul AK (2002b) Environmental management and prospects of coastal tourism in Purba Medinipur district. Indian J Geogr Environ. Vidyasagar University, Medinipur Paul SK (2006) Issues in coastal zone management of Digha-Shankarpur coastal area. ISRO-RSAM project report, RRSSC, Kharagpur Quarantelli EL (1998) What is a disaster? Perspectives on the question. Routledge, London TERI (2011) Measuring, monitoring and managing sustainability in Indian coastal areas: the socioeconomic dimensions UNEP (2009) Sustainable coastal tourism: an integrated planning & management approach UNWTO (2004) Making tourism more sustainable: a guide for policy makers, pp 09–12 UNWTO/UNEP (2005) Policies and tools for sustainable tourism—“a guide for policy makers”

112

M. Mondal et al.

Wong PP (1993) Tourism versus environment: the case for coastal areas. Kluwer Publisher, Dordrecht, Boston Yazdi SK (2012) Sustainable tourism. Am Int J Soc Sci 1(1):50–56

Chapter 5

Impact of Urban Heat Island: A Local-Level Urban Climate Phenomenon on Urban Ecology and Human Health Sangita Singh, Priya Priyadarshni, and Puneeta Pandey

Abstract With the rapid rate of urbanization in the last few decades, tremendous changes in the land use and land cover pattern of urban areas have occurred. These include the conversion of pervious surfaces to impervious surfaces, a decrease in the open and green spaces, and decreased sky-view factor that has led to the formation of Urban Heat Islands. These urban heat islands have implications for urban climate as it leads to elevated temperatures in the city centers and lower temperatures on the outskirts of urban areas. This further leads to increased energy consumption in cities besides having adverse effects on human health that get manifested in the form of distress as well as stroke due to high intensity of heat, fatigue, sapping of energy, irritation, and suicidal tendencies. With the ever-increasing size of urban areas by means of urban sprawl due to a greater influx of people from rural to urban areas in search of better amenities and opportunities, the problem of urban heat islands is bound to exacerbate in the coming years. Therefore, the present paper aims at critically assessing the spatio-temporal domain of Urban Heat Island in relation to urban ecology and human health using geospatial technology. A case study of Bathinda city has been described to assess the Impervious Surfaces and their impacts on Land Surface Temperature in Bathinda City of Punjab. Further, an attempt would be made to study the adaptation and mitigation measures for urban heat islands. Keywords Urban heat island · Urban climate · Urban ecology · Human health · Urban sprawl

S. Singh · P. Priyadarshni · P. Pandey (B) Department of Environmental Science and Technology, School of Environment and Earth Sciences, Central University of Punjab, Badal Road, VPO Ghudda, Bathinda 151401, Punjab, India e-mail: [email protected]; [email protected]; [email protected] P. Pandey Centre of Environmental Studies, University of Allahabad, Uttar Pradesh 211002 Prayagraj, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_5

113

114

S. Singh et al.

5.1 Introduction Statistically, the world urban population has grown vividly from 750 million to 4.46 billion in the past 60 years (WUP 2018). Urbanization has a significant influence on land use and land cover (LULC) pattern of the urban area. As of now, 56% of the world population is living in the urban settlement, and the proportion is expected to be increased to 68% by the year 2050 (www.un.org). North America, Latin America and the Caribbean, Europe, and Oceania have the most urbanized regions of the world. Asia and Africa are contributing fast to the scenario. In fact, 90% of the projected growth is expected to come from these continents (www.un.org). The most notable effect of urban sprawl on the biophysical environment is the replacement of the soil and vegetation by impervious building material which affects the reflectivity and the attributes of runoff from land based on topography (Weng 2001). This has a bearing on the exchange and transformation of energy between land and the atmosphere (Ketterer and Matzarakis 2014). Although the urbanization has a positive effect on the global gross value, it has some negative effects too. The urban heat island (UHI) is one such concern. Urban heat island is a microclimatic condition in which the urban area experiences elevated temperature as compared to the nearby non-urban settlement. Heat is generated by energy from all the people, cars, buses, and trains in big cities. UHI is the cumulative effect of the activities performed by the people living in an area as they burn energy while jogging, breathing, or driving. It has been contributing actively in an unpleasant manner to global warming, heat stroke, unpredictable climate change, urban air quality, and energy expenditure pattern in addition to threat to the residents’ well-being (Kim 1992). Thus, in current times, UHI has been a central focal point for the climate change research. A report by the Intergovernmental Panel on Climate Change has stated that there is an increase in global mean temperature of 0.74 °C during 1906–2005 (IPCC 2007). This has implications for urban residents as UHIs increase the temperature of urban areas; which are already facing the brunt of climate change by frequent and intense heat waves and erratic rainfall. Figure 5.1 depicts these changes in temperature in an urban structure and the occurrence of UHIs.

5.2 Classification of UHI Several heat islands are formed due to the relative warmth of the atmosphere, surface, and substrate material. Figure 5.2 describes the various types of UHI. Surface UHI: The surface UHI represents the relative warmth of the dry and uncovered surfaces in cities compared to that in rural areas. Land surface temperature (LST) variation of urban and rural areas is determined in order to study SUHI. The magnitude of SUHI relies on the intensity of the sun’s radiations and the availability of sunlight during the daytime. Surface UHIs are mostly seen during summertime, as in summer there is calm low wind speed and clear weather which in turn results in

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

115

Fig. 5.1 Urban heat island (Source usgs.gov)

Fig. 5.2 Types of urban heat island (UHI)

less mixing of air and reduced sun rays’ dispersal (Oke 1997; Voogt and Oke 2003). Remote sensing techniques help in determining the surface temperature by providing images from the satellite or aircraft-mounted instruments (Ngie et al. 2014; Liu et al. 2015). Atmospheric UHI: Temperature difference in the air of the urban and rural parts is the key reason for the formation of atmospheric urban heat island. In urban areas, the air is warmer compared to cooler air in nearby rural surroundings, thus forming

116

S. Singh et al.

UHIs (Oke 1997). This phenomenon is usually observed at night-time (Roth et al. 1989). Canopy layer UHI: Canopy layer UHI is the well-known microscale-level UHI. It is formed where the communities and constructions are present and extend from ground level to the average maximum height of the nearby rooftops or vegetation. Hence, it has a direct effect on human activity (Smoliak et al. 2015). It can be easily accessed by ground-based instruments (stationary/mobile) (Fabrizi et al. 2010). It is observed in calm winds and clear skies, mostly seen at nights. It grows rapidly near the late afternoons and early evenings. It shows temporal variability under calm and clear weather conditions as it may show positive magnitude during daytime. Its spatial structure shows an isotherm (isolines of the corresponding temperature). The spatial temperature gradient, the change of temperature over space, is typically large near the edge of the island. Boundary-layer UHI: The urban boundary-layer heat island represents an urbanscale warming through the depth of the urban boundary layer (up to 1–2 km during daytime with clear skies and a few 10–100 s of meters at night). Its magnitude decreases with height by night and remains constant by day that has a smaller magnitude and is much less spatially and temporally variable than that of the underlying UCL heat island (Voogt 2020). Measurements of the urban boundary-layer heat island are relatively rare as access to this layer is difficult. Thermometers must be mounted on tall towers, balloons, or aircraft; or temperatures can be observed by remote sensing techniques using ground-based instruments (Fabrizi et al. 2010). Subsurface layer UHI It is formed due to the contribution of heat from the basement of buildings and subsurface infrastructure. The infrastructure (asphalt and concrete buildings) alters the energy balance of the earth surface and results in the local warming of the ground and air temperature (Previati and Crosta 2021). It is affected by the relative warmth of the ground and atmosphere. It can be measured by ground or borehole temperature measurement. Spatio-temporal variation mainly occurs due to variations in surface characteristics and subsurface heat sources from urban infrastructure and depth, respectively, in order to show past conditions at the surface (Sharma and Pandey 2015; Kesavan et al. 2021). If we analyze the methods by which urban heat islands can be studied, it is observed that they fall into three broad categories: Thermal remote sensing, Smallscale models, and Transect studies (Fig. 5.3). These are discussed in detail below: Thermal remote sensing: Surface heat islands are primarily measured by the remote sensing in the thermal infrared (TIR) region of the electromagnetic (EM) spectrum. TIR wavelength between 8 and 15 µm is commonly used for LST estimation. It helps in determining the hotspot in the urban area. It helps in studying the urban thermal environment at various spatial and temporal scales (Weng 2009). Small-scale models: A blueprint of the study area is made and compared with a standardized mathematical model. On the basis of that resemblance and variations are estimated to get the precise results (Mirzaei and Haghighat 2010).

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

117

Fig. 5.3 Methods to study urban heat island

Transect studies: It is an urban planning model which consists of a series of transition zone. It includes uninterrupted air data monitoring by using weather stations on the mobile vehicles. Hence, it is good for superior temporal studies (Dijoo 2021).

5.3 Factors Contributing to Urban Heat Island Formation Urbanization: Urbanization is primarily the process by which towns and cities are created and grow as more people move into the center of the city for their livelihood as a result of the migration of the population to urban areas. According to the United Nations Population Fund, 61% of the population is expected to house the urban settlement by the end of 2030 (Stempihar et al. 2012). It brings with it more population density and increased hard and modern surfaces with less permeability and absorption capacity (USEPA 2022a, b). Surface Geometry: It is concerned with the properties and relations of the surfaces. With the decrease in the ratio of building height to the street width, the UHI magnitude increases (Wang et al. 2016). Anthropogenic Heat Input: Simpler activities like respiration and combustion also add on to the increased effect of the urban heat island. For instance, the air we breathe becomes warmer when we exhale it into the environment. It is the same case that happens with the vehicular exhaust which releases toxic and heated air into the environment. Transport: A higher volume of transport is required to serve the purpose of connectivity. This means more heat-generating engines and more contribution to the UHI effect. Urban Canyon effect: Urban canyon refers to the deep gorge or cleft of street and alleyways formed between the tall buildings of the cities which give an impression of a canyon. This kind of arrangement hinders the free flow of air. Trapping the heat in between the buildings increases the temperature of the area. It plays a significant role in nocturnal heat island formation and human activity-induced heating (Kusaka and Kimura 2004).

118

S. Singh et al.

Fig. 5.4 Factors affecting UHI formation

The factors affecting UHI formation have been represented in Fig. 5.4. Reduced vegetation: Vegetation or tree cover helps the environment to cool off. But in order to suffice the area for the urban construction and infrastructure development, the tree cover is reduced first. It also affects the natural process of evapotranspiration which helps in the cooling as the moisture that helps in cooling down the temperature becomes unavailable due to less greenery. Concrete surfaces contribute to three-fourths of impervious areas in cities while in vegetated areas, impervious surfaces amount to about 10%. This clearly depicts that the presence of green spaces provides the adequate moisture to reduce the temperature near the surface (Dijoo 2021). Energy Needs: Electricity needs increase by 2–4% with rise in 1 °C rise in temperature (Akbari et al. 2001). Air conditioning systems have added comfort to our lives at the cost of anthropogenic heat generation (Grimmond 2007).

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

119

5.4 Effects of UHI on Urban Ecology and Human Health 5.4.1 Urban Ecology Urban Heat Islands are known to significantly affect the urban ecology. Although UHI has positive effects in colder regions by minimizing hazards due to ice or snow in the cold cities (Voogt 2002) or aiding in the blooming or growth of plants in colder regions; however, the negative consequences far outweigh the positive consequences. UHIs are a consequence of changing urban landscapes from pervious vegetated surfaces to impervious concrete surfaces. Urbanization has caused tremendous change in the land use and land cover pattern resulting in a decrease in the open space and green cover all over the world (Singh and Singh 2014). Urbanization has caused a rapid change in the surface type from pervious to impervious. Pervious surface refers to those surfaces that are natural and through which air and water can pass through; on the other hand, Impervious surfaces are those man-made concrete structures that replace the natural surface into which air or water cannot pass through like highways, buildings, parking area, pavements, footpaths, and other urban concrete structures (Yang et al. 2003). This can also serve as an indicator of spatio-temporal LULC change, urban environment quality, and water quality (Kaur and Pandey 2020). They not only absorb heat but also disturb the percolation of surface water necessary for recharging groundwater (Yang et al. 2003; Liu et al. 2015). As a result, the reduced vegetation cover and increasing concrete cover by means of settlements, roads, and other urban infrastructure exacerbate the thermal environment of an urban area that not only affects the quality of life, but also the residents’ health besides energy and water consumption. UHIs increase the demand for electricity consumption due to greater cooling requirements by refrigeration, air conditioners, etc., which further increase the anthropogenic heat. The electricity supply majorly is from fossil fuelbased power plants, which in turn leads to an increase in air pollutants and greenhouse gas emissions. These pollutants contribute to complex air quality problems such as the smog formation and emission of gases such as sulfur oxides, carbon monoxides, ground-level ozone, and other greenhouse gases, besides fine particulate matter and acid rain. The relationship between UHI and aerosol load and fluctuation in air temperature has been studied in Delhi by Pandey et al. (2009, 2012a, b) for summer and winter months. The study revealed distinct UHI during night-time in both the summer and winter months and aerosols were found to play dual behavior during daytime and night-time. As regards impairment of water quality, higher temperatures due to UHIs often result in hot storm-water runoff from hot roofs and hot pavement, which then gathers in the sewers raising the temperature of sewers’ water further. The rivers and other adjacent water sources get heated by the hot sewer water that is discharged into them. This has repercussions for aquatic biota such as fishes that are sensitive to changes in temperature. This causes stress, disease, and even death of fishes, which disturbs the entire food chain in the water body and causes ecological imbalance. Somers et al.

120

S. Singh et al.

(2016) reported that urban streams are hotter by 7°F than streams in forested areas on an average, due to heated runoff from urban materials. High built-up structure results in reduced movement of air which decreases the moisture in the ambient air and adding of waste heat to the urban atmosphere (Pandey et al. 2014). Higher urban and sub-urban temperature of ambient air and an increase in storm-water runoff lead to an impact on environmental health quality. These perturbations of the local heat energy budget of the area cause local climate change (Roth 2002). This altered energy budget has been one of the visible impacts of urbanization in the formation of urban heat island (UHI) (Li et al. 2010). The microclimatic environment in the urban area is the attribute of the urban heat island phenomenon affecting energy demand, and human and environmental health (Kaur and Pandey 2021, 2022).

5.4.2 Human Health One of the main consequences of urban heat islands is higher day and night temperatures coupled with increased pollution levels. As a result, the elevated temperatures result in heat stress and heat-related discomfort such as heat stroke, heat cramps, heat exhaustion, and heat-related respiratory difficulties (Kotharkar et al., 2018). Studies by various researchers (Lo and Quattrochi 2003; Harlan et al. 2006; Grass and Crane 2008; Heaviside et al. 2017; Singh et al. 2020) have reported that direct and incidental temperature alterations coupled with air pollution in cities worsen the discomfort and aggravate human health. Not only this, heightened temperatures augment the ground-level ozone concentrations, which influence human health, and are amplified by the effects of higher daily maximum temperatures (Oke 1997). Further, UHI also aggravates the impact of naturally occurring heat waves that affect both older adults and young children. These are the vulnerable population that face the brunt of high heat and suffer from physiological and psychological factors. People working in the sun are more susceptible to heat exhaustion and heat stroke due to increased ozone air pollution and heat stress exposures (Heaviside et al. 2017; Pavithra 2022). Further, people in low-income groups are more likely to get sick from the heatrelated illnesses due to poor housing conditions, lack of proper ventilation, and crowded dwelling areas (U.S. Climate Change Science Program 2008). Also, people who already have poor health, such as those with diabetes, physical disabilities, chronic illnesses, mobility issues, or who are under medications, are especially susceptible to high heat (Phelan et al. 2015). Besides health issues that are non-fatal, many a times excessive temperatures cause heat-related mortality (Wong et al. 2013). Center for Disease Control and Prevention recorded 10,527 heat-related deaths in the United States from 2004 to 2018 (Vaidyanathan et al. 2020). According to WMO, more than 25,000 deaths occurred due to heatwaves in India in 28 years between 1992 and 2020.

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

121

Bathinda city

Fig. 5.5 Study area-Bathinda city

5.5 Case Study 5.5.1 Background of Study A study was carried out by the authors for assessing the Impervious Surfaces and their impacts on Land Surface Temperature in Bathinda City of Punjab at the Department of Environmental Science and Technology, Central University of Punjab, Bathinda; where land surface temperature (LST) of the city of Bathinda, Punjab, India, was estimated using Landsat-8 satellite images. Not only this, impervious and pervious surfaces were classified in Bathinda city, and various indices such as Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), and Normalized Difference Impervious Index (NDII) were examined. The study area for the present case study has been shown in Fig. 5.5.

5.5.2 Materials and Methods Landsat Satellite data was procured from USGS and after layer stacking in the ERDAS Imagine 2014 software. The stacked bands were resampled to 30 m spatial resolution (for thermal band). The next step was carrying out image classification to

122

S. Singh et al.

assign a LULC class to each pixel in the image, depending on its spectral characteristics. In the present study, Supervised classification was carried out using a maximum likelihood classification (MLC) algorithm followed by accuracy assessment.

5.5.3 Results and Discussion The Land Use and Land Cover (LULC) map of the region was prepared by supervised classification using a maximum likelihood classifier (MLC) with an overall classification accuracy of 83.98%. The supervised classification image represents four LULC classes viz. water- comprising all canals and distributaries; vegetation— including agriculture and vegetation areas; settlements and bare soil as represented in Fig. 5.6. The total area of the Bathinda city was found to be 14,678.1 ha. As per the classification, the highest area in the city falls under the settlement (built-up) area (4655.43 ha), an indication of increased urbanization. Further, it was followed by vegetation (3882.51 ha), bare soil 3763.8 ha, and then water bodies (2376.36 ha). In the next step, Land surface temperature (LST) map of the study area was prepared using OLI and TIRS band data of Landsat-8 that were converted to Top of Atmosphere (TOA) spectral radiance, which was converted again to brightness temperature (Fig. 5.7). Later, NDVI, NDBI, and NDII were calculated using the following equations: NDVI = (NIR − RED)/(NIR + RED)

(5.1)

NDBI = (NIR − SWIR)/(NIR + SWIR)

(5.2)

NDII = (VIS − TIR)/(VIS + TIR)

(5.3)

where NIR stands for Near Infrared band, RED for red band, SWIR for short wave infrared, VIS for visible, and TI for thermal Infrared wavelength. Finally, the built-up area was extracted by the following index: Based on Eqs. 5.1, 5.2, and 5.3, NDVI (Fig. 5.8a), NDBI (Fig. 5.8b), and NDII (Fig. 5.8c) were estimated. The maximum value for NDVI was found to be 0.35, indicating that the vegetation is scarce (Fig. 5.8a). Ground-based studies revealed this vegetation area to be belonging to agricultural fields. In the present study, NDBI values range from −0.627 to 0.189 (Fig. 5.8b) with the highest value representing the built-up structures in red color; while the Normalized Difference Impervious Index (NDII) ranges from 0 to 1. In Fig. 5.8c, the whiteshaded region represents fallow or barren land, which appears very bright in the image due to very low vegetation cover and high surface temperature. Hence, these fallow lands exhibited the highest values of NDII. These were followed by grayishpurple area belonging to settlements and built-up area that exhibited high NDII values, but lower than fallow land. Next in decreasing order of NDII value lay the

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

123

Fig. 5.6 Land use and land cover (LULC) map of Bathinda city

vegetation and agricultural area; while the least NDII value was exhibited by water bodies (represented by green color).

5.5.4 Conclusion and Recommendations Based on the above study, the following conclusions were drawn: . NDVI for Bathinda city was 0.35 which was very low (sparse vegetation). . NDBI for Bathinda city was 0.189 which was quite high for a small city. . NDII was extracted solely for the study of impervious surfaces which was reported as highest for fallow lands followed by settlements and lowest for water bodies.

124

Fig. 5.7 Land surface temperature map of Bathinda city

S. Singh et al.

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

125

Fig. 5.8 Indices for Bathinda city: a NDVI, b NDBI, and c NDII

. The maximum temperature was reported in the central part of the city having higher settlement and built-up area (impervious surfaces) and comparatively low toward the outskirts. . Change in temperature pattern revealed that low vegetation cover and dense builtup accelerate the waste heat in the atmosphere, to overcome which circulation of air and pervious surfaces (open areas with dense or moderate vegetation) would be helpful. . The prominent change in temperature could be due to the conversion of natural cover (land or vegetation) to agricultural or impervious surfaces during the last few years.

5.6 Recommendations Various strategies have been devised by various researchers for minimizing the urban heat island effect: . Building green infrastructure (USEPA 2022a) . Planting trees and other vegetation in fallow or barren areas of city, and vacant land along side roads and streets (Wang and Akbari 2016; Wang et al. 2016). . Building green roofs to provide both direct and ambient cooling effects by reducing the heat island effect; besides improving the air quality (Khare et al. 2021).

126

S. Singh et al.

Initiatives by the Government: The Heat Action Plan was put into action by Government of India in 2013 in collaboration with National Disaster Management Authority (NDMA) to develop complete early warning system and emergency response strategy for excessive heat events in various Indian states.

5.7 Conclusions Urban Heat Island (UHI) is a local micro-climate modification of urban areas resulting from altered land use and land cover as a result of urbanization. These are regions of elevated temperatures, which owe their origin to emission of excess anthropogenic heat, replacement of pervious soil cover by impervious concrete/ asphalt surfaces; as well as reduced sky-view factor. These have implications for the air and water quality of urban areas and causes discomfort to urban residents. With the increasing climate change and role of greenhouse gases and aerosols in climate change, urban heat islands serve as a very good example in learning about these phenomena at local level and utilizing the knowledge to solve environmental issues at global level. Acknowledgements The authors would like to thank the Central University of Punjab, Bathinda, for providing the necessary facilities under Research Seed Money Grant (GP-25). The authors declare no conflict of interest.

References Akbari H, Pomerantz M, Taha H (2001) Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Sol Energy 70(3):295–310 Dijoo ZK (2021) Urban heat island effect concept and its assessment using satellite-based remote sensing data. Geograph Inf Sci Land Resour Manage 81–98 Fabrizi R, Bonafoni s, Biondi R (2010) Satellite and ground-based sensors for the urban heat island analysis in the city of Rome. Remote Sens 2:1400–1415. https://doi.org/10.3390/rs2051400 Grass D, Crane M (2008) The effects of weather and air pollution on cardiovascular and respiratory mortality in Santiago, Chile, during the winters of 1988–1996. Int J Climatol 28:1113–1126 Grimmond S (2007) Urbanization and global environmental change: local effects of urban warming. Geograph J 173:83–88. https://doi.org/10.1111/j.1475-4959.2007.232_3.x Harlan SH, Brazel AJ, Prashad L, Stefanov WL, Larsen L (2006) Neighborhood microclimates and vulnerability to heat stress. Soc Sci Med 63:2847–2863 Heaviside C, Macintyre H, Vardoulakis S (2017) The urban heat island: implications for health in a changing environment. Current Environ Health Rep 4(3):296–305 IPCC (2007) In: Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt K, Tignor M, Miller HL (eds) Climate change 2007: The physical science basis. Contribution of Working Group I to the Fourth Assessment Report of the IPCC. Cambridge University Press, New York Kaur R, Pandey P (2020) Monitoring and spatio-temporal analysis of UHI effect for Mansa district of Punjab, India. Adv Environ Res 9(1):19–39

5 Impact of Urban Heat Island: A Local-Level Urban Climate …

127

Kaur R, Pandey P (2021) Spatial trends of surface urban heat island in Bathinda: a semiarid city of northwestern India. Int J Environ Sci Technol 1–22 Kaur R, Pandey P (2022) Air pollution, climate change, and human health in Indian Cities: a brief. In: The role of climate and air pollution in human health and urban chemistry in Asian cities Kesavan R, Muthian M, Sudalaimuthu K, Sundarsingh S, Krishnan S (2021) ARIMA modeling for forecasting land surface temperature and determination of urban heat island using remote sensing techniques for Chennai city, India. Arab J Geosci 14(11):1–14 Ketterer C, Matzarakis A (2014) Human-biometeorological assessment of the urban heat island in a city with complex topography—The case of Stuttgart, Germany. Urban Climate 10:573–584 Khare VR, Vajpai A, Gupta D (2021) A big picture of urban heat island mitigation strategies and recommendation for India. Urban Climate 37:100845 Kim HH (1992) Urban heat island. Int J Remote Sens 13(12):2319–2336 Kotharkar R, Ramesh A, Bagade A (2018) Urban heat island studies in South Asia: a critical review. Urban Climate 24:1011–1026 Kusaka H, Kimura F (2004) Thermal effects of urban canyon structure on the nocturnal heat island: numerical experiment using a mesoscale model coupled with an urban canopy model. J Appl Meteorol 43(12):1899–1910 Li S, Zhao Z, Miaomiao X, Wang Y (2010) Investigating spatial non-stationary and scale-dependent relationships between urban surface temperature and environmental factors using geographically weighted regression. Environ Model Softw 25(12):1789–1800 Liu K, Su H, Zhang L, Yang H, Zhang R, Li X (2015) Analysis of the urban heat island effect in Shijiazhuang, China using satellite and airborne data. Remote Sens 7(4):4804–4833 Lo CP, Quattrochi DA (2003) Land-use and land-cover change, urban heat island phenomenon, and health implications. Photogramm Eng Remote Sens 69(9):1053–1063 Mirzaei PA, Haghighat F (2010) Approaches to study urban heat island–abilities and limitations. Build Environ 45(10):2192–2201 Ngie A, Abutaleb K, Ahmed F, Darwish A, Ahmed M (2014) Assessment of urban heat island using satellite remotely sensed imagery: a review. South African Geograph J=SuidAfrikaanseGeografieseTydskrif, 96(2):198–214 Oke TR (1997) Urban climates and global environmental change. In: Thompson RD, Perry A (eds) Applied climatology: principles and practice. Routledge, London, pp 273–287 Pavithra KM (2022) Data: as temperatures soar across India, different agencies of the government provide varied data on heat wave mortality. https://factly.in/data-as-temperatures-soar-acrossindia-different-agencies-of-the-government-provide-varied-data-on-mortality/. Accessed 2nd July 2022 Pandey P, Kumar D, Prakash A, Kumar K, Jain VK (2009) A study of the summertime urban heat island over Delhi. Int J Sustain Sci Stud 1:27–34 Pandey P, Kumar D, Prakash A, Lodhi N, Singh M, Singh S, Jain VK, Kumar K (2012a) Paper no 500: temporal dynamics of urban heat island formation over Delhi. In: ICUC8 8th international conference on urban climates, 6th–10th August, UCD, Dublin Ireland Pandey P, Kumar D, Prakash A, Masih J, Singh M, Kumar S, Jain VK, Kumar K (2012b) A study of urban heat island and its association with particulate matter during winter months over Delhi. Sci Total Environ 414:494–507 Pandey AK, Singh S, Berwal S, Kumar D, Pandey P, Prakash A, Kumar K (2014) Spatio–temporal variations of urban heat island over Delhi. Urban Climate 10:119–133 Phelan PE, Kaloush K, Miner M, Golden J, Phelan B, Silva H III, Taylor RA (2015) Urban heat island: mechanisms, implications, and possible remedies. Annu Rev Environ Resour 40:285–307 Previati A, Crosta GB (2021) Characterization of the subsurface urban heat island and its sources in the Milan city area, Italy. Hydrogeol J 29(7):2487–2500 Roth M, Oke TR, Emery WJ (1989) Urban heat islands from NOAA satellite data. Rem Sens Environ 10:1699–1720

128

S. Singh et al.

Roth M (2002) Effects of cities on local climates. In: Proceedings of workshop of IGES/APN megacity project. Kitakyushu, Japan. 23–25 Jan. 2002. Inst. for Global Environmental Strategies, Arlington, VA, pp 1–13 Sharma N, Pandey P (2015) Study of urban heat island in Bathinda city, Punjab. In: 16th ESRI India user conference, pp 1–14 Singh B, Singh J (2014) Land use/land cover change in a fast growing urban centre in National Capital Region (NCR): a case study. Int Res J Earth Sci 2(1):15–20 Singh N, Singh S, Mall RK (2020) Urban ecology and human health: implications of urban heat island, air pollution and climate change nexus. In: Urban ecology. Elsevier, pp 317–334 Smoliak BV, Snyder PK, Twine TE, Mykleby PM, Hertel WF (2015) Dense network observations of the Twin Cities canopy-layer urban heat island. J Appl Meteorol Climatol 54(9):1899–1917 Somers KA, Berhnardt ES, McGlynn BL, Urban DL (2016) Downstream dissipation of storm flow heat pulses: a case study and its landscape-level implications. J Am Water Resour Assoc 52(2):281–297 Stempihar JJ, Pourshams-Manzouri T, Kaloush KE, Rodezno MC (2012) Porous asphalt pavement temperature effects for urban heat island analysis. Transp Res Rec 2293(1):123–130 USEPA (2022a) Reduce urban heat island effect. https://www.epa.gov/green-infrastructure/reduceurban-heat-island-effect. Accessed 28 June 2022a USEPA (2022b) Learn about heat islands. https://www.epa.gov/heatislands/learn-about-heat-isl ands. Accessed 28th June 2022b U.S. Climate Change Science Program (2008) Analyses of the effects of global change on human health and welfare and human systems. A Report by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research [Gamble, J.L. (ed.), K.L. Ebi, F.G. Sussman, T.J. Wilbanks (Authors)]. U.S. Environmental Protection Agency, Washington, DC, USA US Geological Survey Urban Heat Islands. https://www.usgs.gov/media/images/urban-heat-isl ands. Accessed 1 Mar 2020 Vaidyanathan A, Malilay J, Schramm P, Saha S (2020) Heat-related deaths—United States, 2004– 2018. MMWR Morb Mortal Wkly Rep 2020, 69:729–734. 10…15585/ Voogt JA (2002) Urban heat island. In: Douglas I (ed) Causes and consequences of global environmental change, vol 3. Wiley, Chichester, UK, pp 660–666 Voogt JA (2020) Urban heat islands. In: Atmosphere and climate. CRC Press, pp 265–272 Voogt JA, Oke TR (2003) Thermal remote sensing of urban climates. Remote Sens Environ 86(3):370–384 Wang Y, Akbari H (2016) The effects of street tree planting on urban heat island mitigation in Montreal. Sustain Cities Soc 27:122–128 Wang Y, Berardi U, Akbari H (2016) Comparing the effects of urban heat island mitigation strategies for Toronto, Canada. Energy Build 114:2–19 Weng Q (2001) A remote sensing? GIS evaluation of urban expansion and its impact on surface temperature in the Zhujiang Delta, China. Int J Remote Sens 22(10):1999–2014 Weng Q (2009) Thermal infrared remote sensing for urban climate and environmental studies: Methods, applications, and trends. ISPRS J Photogram Remote Sens 64(4):335–344. https:// doi.org/10.1016/j.isprsjprs.2009.03.007 WUP (2018) United Nations, Department of Economic and Social Affairs, Population Division (2019). World Urbanization Prospects 2018: Highlights (ST/ESA/SER.A/421) Wong KV, Paddon A, Jimenez A (2013) Review of world urban heat islands: many linked to increased mortality. J Energy Resour Technol 135(2). https://doi.org/10.1115/1.4023176 Yang L, Huang C, Homer CG, Wylie BK, Coan MJ (2003) An approach for mapping large-area impervious surfaces: synergistic use of Landsat-7 ETM+ and high spatial resolution imagery. Can J Remote Sens 29(2):230–240

Chapter 6

Identification of Environmental Epidemiology Through Advanced Remote Sensing Based on NDVI Vibhanshu Kumar, Birendra Bharti, Harendra Prasad Singh, Himanshu Kumar, and Sanjay Paul Kujur

Abstract Advanced Remote Sensing (ARS) is a new scientific subject that combines high-performance computers and innovative spatial science techniques to extract knowledge from geographical large data to analyze Environmental Epidemiology (EE). Phenology of ecological landscapes (EL) is a cast-off approach to conduct EE to evaluate the hydrological scenarios based on the assumptions that Normalized Difference Vegetation Index (NDVI) time series follow annual cycles of growth and decline of vegetation, and that clouds or poor atmospheric conditions usually depress NDVI values. The application of ARS in the planning, implementation, and management of different water resources projects, is primarily concerned with the evaluation of ecological landscape response. When it comes to EE, ARS provides significant benefits, including the capacity to include massive volumes of vast geographical and temporal data in variability of forms; computational efficiency; workflows to accommodate crucial properties of spatial (environmental) processes, such as spatial nonstationarity; and scalability to EL with additional environmental exposures across multiple geographies. Key principles in geographic data science, data mining in EL with recent accessible spatial data applications in research, and future possibilities for phenology will be presented in this article. Keywords Advanced remote sensing · Environmental epidemiology · Phenology · Normalized difference vegetation index

V. Kumar (B) · B. Bharti · H. P. Singh · H. Kumar Department of Water Engineering & Management, Central University of Jharkhand, Ranchi, India e-mail: [email protected] S. P. Kujur Department of Geoinformatics, Central University of Jharkhand, Ranchi, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_6

129

130

V. Kumar et al.

6.1 Introduction Vegetation is a key component of the Earth’s biological system and a defining link between soil, climate, and water. The role of vegetation in the biological system is significant, and it can serve as a “marker” in the study of global change. Acknowledging the behaviour of terrestrial ecosystems and its relationships with climate variability or sapient natural ecosystem deterioration requires a detailed review and explanation of vegetation dynamics at regional and global scales (Chen et al. 2014; Maclean and Wilson 2011; Thomas et al. 2004; Zheng et al. 2015). When we talk about the ecological environment, we’re talking about the sum total of human society and all the numerous natural variables that surround it (Galford et al. 2008). For this reason, keeping track of the state of the ecosystem and how it’s changing is crucial. Long-term observation is needed to get a good understanding of how the ecological environment and processes change and work in a landscape with plants. Hence, in this respect, the assessment of EE plays a crucial role. The ability to observe changes in vegetated areas over time from a distance via ARS is a valuable tool for researchers. Predicting the occurrence of destructive natural catastrophes, assessing their impact, and developing preventative measures all benefit from vegetation analysis. Vegetation surveillance from the ground is limited in its ability to track changes in vegetation’s geographical and temporal distribution in real time, but a satellite’s synoptic perspective can keep a close watch on the entire planet. Due to the prevalence of drought in India, where many farmers use precipitation-based agriculture, agricultural output rates and economic losses have increased. Therefore, it is crucial to keep an eye on the spatial and temporal variation of vegetation risk for agronomy and refuge, which is calculated using a seminal drought index like the Normalized Difference Vegetation Index (NDVI) drafted by Kriegler et al. (1969) and presented by Rouse et al. (1974) is extensively castoff by various researchers at a different time scale to monitor vegetation (Kasoro et al. 2021; Luo et al. 2020; Rouse et al. 1974). In order to assess climate’s impact, it is necessary to have a firm grasp on the phenomenon itself, as well as its qualitative and quantitative development. As part of a comprehensive drought tackle strategy, it is essential to identify the unique characteristics of vegetation in order to create a more accurate and advanced monitoring system. A drought is a period of unusually little precipitation over an extended period of time, causing disruptions in water supply and water interest. A drought is a typical cataclysmic disaster that can affect a large area and last for a much extended time. Therefore, a successful hypothetical basis for the management of vegetation as an EE mitigation method may be obtained by focusing on the link between the dry season and vegetation. People can gain an understanding of drought through traditional methods, but they will need hydrological and meteorological data to do so. But in the present decades, spatial data are also used for drought analysis. Therefore, it is crucial to correctly comprehend the connection between territorial vegetation cover and environmental change by learning how different types of normal vegetation respond differently in terms of NDVI to EE.

6 Identification of Environmental Epidemiology Through Advanced …

131

The vulnerability of different regions to drought varies according to physioclimatic factors (Sehgal and Dhaker 2016). Reducing the likely hazard in the agriculture sector requires a convincing assessment of the vulnerability of agricultural land to drought. Rainfall is the primary risk factor that can establish water scarcity due to precipitation shortfall. Vegetation, soil moisture, and temperature are especially important for determining how sensitive a certain area is to drought (Bhuiyan et al. 2006, 2017; Kundu et al. 2016; Singh et al. 2003). This paradigm still requires scientific proof that vegetation gets hampered due to agricultural drought. In this study, we will be exploring the NDVI as a land use classification in the context of drought analysis.

6.2 NDVI There are a number of vegetation indices available for use in conjunction with ARS data to provide an estimate of plant cover. It is possible to calculate a vegetation index using several combinations of remote sensing bands. In order to assess vegetation cover, the NDVI is the most widely employed as a spectral index. Algebraically stated, a spectral index is a ratio between the red and near-infrared spectral bands (NIR). The value of this index is found by dividing the difference between the Red and NIR reflectances by the total of the two (Kriegler et al. 1969; Weier and Herring 2000). NIR (near-infrared) light is reflected by leaf tissue, and the sensor measures this reflectance in vegetated regions. NDVI can be extracted by NDVI = (NIR − RED)/(NIR + RED)

(6.1)

The NDVI index can have a value between minus one and plus one. Values are typically positive for vegetated areas, neutral for bare soil, and negative for open water. Our next step is to select different values as the limits for the water required in NDVI based on the output. The NDVI is the measure of the estimate of the amount of radiation being absorbed by plants. When compared to the near-infrared, the amount of visible light reflected by green and healthy vegetation is significantly lower. High NDVI values indicate that the vegetation is both healthy and dense. In contrast, negative index returns are produced by objects like clouds, water, and snow because their visible reflectance is higher than their NIR reflectance. Values close to zero are produced by areas covered in rock or bare soil due to their comparable reflection in the visible/nearinfrared range. As a result of these features, NDVI has emerged as the go-to method for documenting shifts in plant cover and assessing the effects of environmental phenomena. In addition to accurately describing regional land coverings, classifying vegetation, and determining the phenology of vegetation, the NDVI is helpful in evaluating rainfall and drought, quantifying the net primary production of vegetation, monitoring crop productivity and yields, recognizing weather impacts, and more.

132

V. Kumar et al.

The available NDVI data sets contain some noise, Pettorelli et al. (2005) recommend smoothing the time series before using it. The use of NDVI is intended to enhance the study of remotely sensed data pertaining to vegetation. Pettorelli et al. (2005) found that NDVI accurately distinguished between evergreen and seasonal forest types, and other studies have shown that it can accurately estimate a wide range of vegetation properties, such as leaf area index, biomass, chlorophyll concentration in leaves, and plant productivity (Chavez et al. 2016; Pastor-Guzman et al. 2015; Zhu and Liu 2015; Tian et al. 2017). Estimates like these are frequently produced by comparing NDVI readings from satellites with readings taken on the ground. Models based on NDVI depend critically on their accuracy (Butt 2018).

6.3 Acceptance of Use of NDVI Globally The NDVI is appealing in commercial agriculture and land use studies because of its speed in distinguishing between healthy and stressed plants. In fact, all indices developed to simplify data assemblages that are otherwise complex exhibit this characteristic. This index’s utility was soon acknowledged by scientists in the early 1970s, and since then, all earth observation satellite remote sensing data have included the capability to generate it at a variety of geographical and temporal resolutions. Observing drought circumstances based on the NDVI has been carried out at the global, continental, regional, and catchment scales by utilizing a wide variety of sensors (NicolaiShaw et al. 2017; Rojas et al. 2011). When determining how susceptible a region is to drought, the NDVI’s detailed information on plant growth is invaluable (Mohmmed et al. 2018; Murthy et al. 2017). In North America (Hwang et al. 2017), South America (Sayago et al. 2017), Europe (Zribi et al. 2016), the Middle East (Pervez et al. 2014), Australia (Chen et al. 2014), Asia (Yu et al. 2003), and Africa, analysts have utilized NDVI to evaluate and screen drought conditions (Funk and Brown 2006). Plenty of studies have attempted to employ these sensors to evaluate vegetation at finer, more local scales, especially when vegetation is highly diverse or sparse (Assal et al. 2016). One example is how the Pathfinder AVHRR Land Science Working Group prioritized, at the planetary scale, the creation of global NDVI data sets (James and Kalluri 1994). The Normalized Difference Vegetation Index (NDVI) is currently the most widely used index for evaluating plant health because of its long history, ease of use, and reliance on widely available multi-spectral bands. The NDVI has lately been applied to the study of agriculture during the offseason by concentrating on hyper-spatial/spectral symbols (collected by highly high spatial and otherworldly target satellites, aeroplanes, or ground/cable car sensors). The Quickbird and Rapid Eye satellite sensors, for instance, have shown impressive promise for high spatial aim (1 m) appraisal of dry period influences on vegetation (Garrity et al. 2013; Krofcheck et al. 2014). It appears from these investigations that vegetation components faithfully reflect rainfall deficits at the landscape level (Laliberte et al. 2004). Since the expense of gaining such pictures or the establishment of ground-based sensor frameworks isn’t effectively met in less evolved areas, most

6 Identification of Environmental Epidemiology Through Advanced …

133

of this high/hyper-goal research has been directed in North America and Europe (Coates et al. 2015; Mänd et al. 2010). NDVI-based approaches in all actuality do anyway have constraints. NDVI, for example, does not show conditions over long periods of time due to its inability to compare conditions based on a single explicit date, which is greatly influenced by soil brilliance in areas with low biomass (Huete 1988; Jasinski 1990). However, on the opposite end of the range, its awareness is limited in high biomass conditions (Galidaki et al. 2016; Mutanga et al. 2012). The analysis of remote sensing papers on farming dry season observation confirmed what we already knew: NDVI is the most widely used method for observing the farming environment. This is crucial since many other, more accurate proxies may now exist since then, such as purely remote sensing techniques and integrated remote sensing and field-based indices. Regarding the type of sensor used, the examined literature favours passive methods over active methods. Several factors, including data accessibility, resolution, timeliness, interpretability, familiarity with methodologies, and consensus on “best practises,” are likely responsible for this (Bachmair et al. 2016). It’s well-known that the Indian monsoon is influenced by the dynamic weather over the Indian subcontinent. Changes in precipitation, temperature, and other climatic conditions can have far-reaching effects on the region’s vegetation and agricultural productivity. Vegetation distribution in the Indian subcontinent is strongly influenced by monsoon rainfall and land surface temperature (Sarkar and Kafatos 2004). India is a vast land having a number of different climatic conditions and agro-climatic zone. This leads to variations in the vegetation condition and quality of the vegetation. The NDVI has been used by many researchers to assess the state of the vegetation and evaluate the drought scenario throughout India (Bachmair et al. 2016; Bhuiyan 2017; Dutta et al. 2015; Kundu et al. 2016; Patel et al. 2007; Singh et al. 2003).

6.4 Case Study 6.4.1 Study Area The Indian state of Jharkhand covers an area of 74,677 km2 in Southeast Asia. At an elevation of between 3 and 1359 m, it is located between latitudes 23°37, 3,, N and 24°4, N and longitudes 86°6, 30,, E and 86°50, E. About 22% of the state’s land is used for farming, making up about 1.8 million acres (Fig. 6.1). Jharkhand is divided into the central, north eastern, and south eastern plateau sub-zones, each with its own unique agro climate. The state relies largely on the monsoon and the Rabi season (October–November) for its primary cropping season (Kharif, June– September). The availability of rainfall makes farming possible during the monsoon season, but farmers in the off-season can’t rely on it for irrigation because it doesn’t rain as often (undependable rainfall). The state of Jharkhand is known for its mineral deposits of bauxite, coal, limestone, stone, and uranium. However, it is also home to

134

V. Kumar et al.

Fig. 6.1 a Map of the study area, b Stacked NDVI of the study area, and c Agro-climatic zones of the study area

unique Jurassic plant remains and a safe habitat for flora and fauna. There has been extensive mining in the study region for quite some time, as it is the mineral centre. The Jharkhand vegetation ecology is therefore under stress from major anthropogenic intercessions.

6.4.2 Methodology of Assessment Although there are some benefits to using analogue film, such as increased spatial resolution, the inevitable shift to digital imaging has been underway since around 1816 (History of the Camera 2020; Campbell and Wynne 2011). Colour images are created by recording values at various wavelengths in a digital snapshot. An RGB image, for instance, would consist of the colours blue (450–490 nm), green (520– 560 nm), and red (635–700 nm). RGB cameras are sensors based on this band combination, and their main goal is to simulate the human visual system digitally. Numerous agricultural researchers have made use of these primed cameras. Many different types of plant diseases, as well as pests and fungi that may be seen directly on the foliage, are studied with the help of RGB cameras (Padmavathi and Thangadurai 2016). Panchromatic images, obtained by cameras that are sensitive to most or all wavelengths in the visual range, typically offer higher spatial resolution than band-separated ones.

6 Identification of Environmental Epidemiology Through Advanced …

135

For this reason, smaller detectors can be used on panchromatic cameras while still maintaining a high signal-to-noise ratio because they gather cumulative energy from the whole spectral range. In general, the great spatial resolution of panchromatic and RGB visible-range images makes them ideal for studies that require the finest level of detail (Zhao et al. 2017). A methodology created by De Bie et al. (2008) was used to process the data [9]. In this study, we first performed an unsupervised classification, and then we classified the stacked Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI (MOD13Q1 data set) layer using a fixed number of categories (10–100). A maximum of 60 iterations were performed, and a convergence criterion of 1 was established. To create the most accurate separation of greenery from the background, several combinations of the R, G, and B bands are tried out. This is done in an effort to reduce the impact of environmental and lighting factors. Consequently, these indices have restricted applications because of their unpredictable thresholding. CIR or multispectral cameras outperform RGB cameras that only use the visible spectrum when it comes to segmenting vegetation (Zheng et al. 2018) and estimating canopy cover across the seasons (Ashapure et al. 2019). To determine which categorized image performed better, we calculated a divergence statistic and reported the results as separability values. High levels of minimal separability (the degree of similarity between the two most similar classes) and average separability (the degree of similarity between all classes) are desirable, but keeping the number of classes small is preferable. After considering both minimum and average divergence, the class Image and NDVI cluster signature with the lowest values were selected for future examination. Examining the temporal and spatial structure of the class images and profiles was the focus of this study. Visual comparison (IMSD 1995; Kameswara Rao 1995) and field data were used to establish crop schedule class recognition. Using crop schedule data and NDVI profiles, we were able to accurately locate sequential cropping systems across space. Low NDVI values in the average of stacked NDVI images can be used to detect a loss of vegetation as well. Information gathered through significant focuses on surveys as well as soil analysis was used to locate notes that connected with various classes. Following that, an NDVI unit map with a full legend displaying by NDVI unit was created.

6.4.3 Result and Discussion QGIS was used to stack the MODIS NDVI Image from January 2011 to December 2020. Subsequently, the stacked layers combined data into a single composite image. This was done to ensure temporal monitoring in three different agro-climatic zones over Jharkhand, India. Using a minimum spectral distance calculation, Iterative Self-Organizing Data Analysis Techniques (ISODATA) (Duda and Hart 1974) is an unsupervised clustering

136

V. Kumar et al.

method for organizing data (Campbell 2002). The stacked NDVI Layer was sorted into a fixed number of categories using these methods (10–100). The performance of the ISODATA tool can be controlled to prevent it from running indefinitely by increasing the maximum number of iterations. The best minimum divergence and average divergence statistics were selected for study to compare classification results using the divergence statistical measures of distance (class separability) between created clusters signatures of “classes.” Higher values in monthly mean composite layers of NDVI were compared to drought detection in terms of vegetation loss in corresponding NDVI classes. Drought conditions can be viewed in Fig. 6.2 in the month of June, July, and September in the year 2019, especially in the central part and south western part of the study area and it can be compared with the same months in the year 2020 of Fig. 6.3. In addition, the temporal variation that existed between the years was analysed and taken into consideration while evaluating the profiles. In 2019, the NDVI showed the smallest spatial and temporal variation in the state of Jharkhand’s south eastern plateau, and the high in its east northern part of the central eastern plateau could be seen in Figs. 6.2 and 6.3. Vegetation shifts are dynamic processes, heavily influenced by rain pattern shifts. Vegetation change in the Central Eastern Plateau’s lower and middle latitudes was severe for the region’s ecological stability, as indicated by the NDVI value. Grassland, arable land, and forest are the three most common forms of vegetation, both of which have the potential to significantly affect the Eco environmental status of the area under consideration. Comparatively (in 2019 and 2020), the south eastern and south western regions will have the highest NDVI, while the north western and central regions will have the lowest. In cases where the drought had little impact on farmers’ crop schedules and would only last for a short period of time, it was deemed insurmountable. A region’s lack of drought management is deemed to be out of its control when the severity of the drought lasts for an unusually long period of time and has a negative impact on planting schedules. If the entire area of a certain NDVI Class is afflicted by drought but it is of short duration and does not encompass the entire NDVI Class, then we may say that this drought is extensive. Hyper-temporal NDVI uses an unsupervised ISODTA clustering technique, making it a suitable tool for agricultural field mapping (De Bie et al. 2008). Number statistics can be calculated from the divergence statistic which can be assessed for different purposes. It was suggested by Swain and Davis (1978) that picking the signature with the largest average divergence was a viable technique. NDVI-derived land cover mapping can also be aided by divergence statics, as demonstrated by DeFries and Townshend (1994). One of the primary factors impacting the distribution of land use systems throughout the three distinct agro-climatic zones of the research region was shown to be drought as vegetation loss of NDVI Classes. Examining the area’s crop schedule and agricultural land usage is crucial for a holistic understanding of vegetation loss. Vegetation loss, on the other hand, could be recognized by correlating a decrease in NDVI value with the commencement of a decrease in NDVI in different layers of a

6 Identification of Environmental Epidemiology Through Advanced …

Fig. 6.2 Monthly mean of NDVI of the year 2019

137

138

Fig. 6.3 Monthly mean of NDVI of the year 2020

V. Kumar et al.

6 Identification of Environmental Epidemiology Through Advanced …

139

mean stacked NDVI image, and its coverage could be seen by superimposing NDVI land use classes. Persistent periods of low NDVI between crop cycles (between NDVI peaks) are usually related to drought. This can be seen in the fluctuating NDVI profiles over time (Figs. 6.2 and 6.3). As a result, the NDVI approach accomplishes its goal of providing spatial and temporal information on the multiyear phenomenon. To assess and study the extensive loss of vegetation ecosystems and associated services brought on by both human activity and natural disasters, continuous monitoring of land use, land cover, and vegetation dynamics is required. In Jharkhand, such surveillance is essential for gauging the extent to which mining and other challenging situations pose risks to ecosystem services.

6.5 Conclusion The assessment of agricultural drought and tracking of vegetation conditions are the two common applications of ARS. Nowadays, a variety of vegetation indices may be accessed; nevertheless, while some are more suited than others for specific applications, no single main index can claim to be universally superior. The NDVI is a powerful statistical tool for determining land use in situations when complex crop schedules are in play. In order to better distinguish crucial land use related EE like drought, NDVI has incorporated long-term historical patterns. With its easy methodology, NDVI has found widespread application in regional and worldwide studies of vegetation. Since agriculture is a very dynamic system, a time series of NDVI values is a useful way to tackle this problem and ensure that the results are correct and can be double-checked. Use of the NDVI in aggregation with other metrics is highly recommended for better output because it doesn’t take into account hierarchy from general to specific, which is critical for studying patterns and processes accountable for generating conditions for a specific land use system. It can also be used to create a long NDVI time series that can be employed in the spatial modelling of water conservation. As a result, the discipline of water conservation science will face longterm challenges. However, it has the potential to assign vegetation conditions to any place, which aids researchers and policymakers in harnessing profitable outcomes in the area of EE.

References Assal TJ, Anderson PJ, Sibold J (2016) Spatial and temporal trends of drought effects in a heterogeneous semi-arid forest ecosystem. For Ecol Manage 365:137–151 Ashapure A, Jung J, Chang A, Oh S, Maeda M, Landivar J (2019) A comparative study of RGB and multispectral sensor-based cotton canopy cover modelling using multi-temporal UAS data. Remote Sens 11(23):2757

140

V. Kumar et al.

Bachmair S, Stahl K, Collins K, Hannaford J, Acreman M, Svoboda M, Knutson C, Smith KH, Wall N, Fuchs B, Crossman ND, Overton IC (2016) Drought indicators revisited: the need for wider consideration of environment and society. Wiley Interdisciplinary Rev Water 3(4) Bhuiyan C, Singh RP, Kogan FN (2006) Monitoring drought dynamics in the Aravalli region (India) using different indices based on ground and remote sensing data. Int J Appl Earth Observat Geoinf 8:289–302 Bhuiyan C, Saha AK, Bandyopadhyay N, Kogan FN (2017) Analyzing the impact of thermal stress on vegetation health and agricultural drought: a case study from Gujarat, India. GIScience and Remote Sens 54(5). https://doi.org/10.1080/15481603.2017.1309737 Butt B (2018) Environmental indicators and governance. Curr Opin Env Sust 32:84–89 Campbell JB (2002) Introduction to remote sensing. London and New York, Taylor and Francis Campbell JB, Wynne RH (2011) Introduction to remote sensing. Guilford Press Chavez RO, Clevers JGPW, Decuyper M, De Bruin S, Herold M (2016) 50 years of water extraction in the Pampa del Tamarugal basin: can Prosopis tamarugo trees survive in the hyper-arid Atacama Desert (Northern Chile)? J Arid Environ 124:292–303 Chen T, De Jeu R, Liu YY, Van der Werf GR, Dolman AJ (2014) Using satellite based soil moisture to quantify the water driven variability in NDVI: a case study over mainland Australia. Remote Sens Environ 140:330–338 Coates AR, Dennison PE, Roberts DA, Roth KL (2015) Monitoring the impacts of severe drought on Southern California chaparral species using hyperspectral and thermal infrared imagery. Remote Sens 7(11):14276–14291 De Bie CAJM, Khan MR, Toxopenus AG, Venus V, Skidmore AK (2008) Hyper temporal image analysis for crop mapping and change detection. In: ISPRS 2008: proceeding of the XXI congress: Comm VII, WGVII/5, Beijing, 3–11 July, pp 803–812 DeFries RS, Townshend JRG (1994) NDVI-derived land cover classification at a global scale. Int J Remote Sens 15:3567–3586 Duda R, Hart PE (1974) Pattern classification and scene analysis. Wiley Dutta D, Kundu A, Patel NR, Saha SK, Siddiqui AR (2015) Assessment of agricultural drought in Rajasthan (India) using remote sensing derived Vegetation Condition Index (VCI) and Standardized Precipitation Index (SPI). Egypt J Remote Sens Space Sci 18:53–63 Funk CC, Brown ME (2006) Intra-seasonal NDVI change projections in semi-arid Africa. Remote Sens Environ 101(2):249–256 Galidaki G, Zianis D, Gitas I, Radoglou K, Karathanassi V, Tsakiri-Strati M, Woodhouse I, Mallinis G (2016) Vegetation biomass estimation with remote sensing: focus on forest and other wooded land over the Mediterranean ecosystem. Int J Remote Sens 38(7):1940–1966 Galford GL, Mustard JF, Melillo J, Gendrin A, Cerri CC, Cerri CEP (2008) Wavelet analysis of MODIS time series to detect expansion and intensification of row-crop agriculture in Brazil. Remote Sens Environ 112:576–587 Garrity SR, Allen CD, Brumby SP, Gangodagamage C, McDowell G, Cai DM (2013) Quantifying tree mortality in a mixed species woodland using multitemporal high spatial resolution satellite imagery. Remote Sens Environ 129:54–65 History of the Camera (2020) Wikipedia [Online]. https://en.wikipedia.org/w/index.php?title=His tory_of_the_camera&oldid=963373872. Accessed 8 June 2022 Huete AR (1988) A soil-adjusted vegetation index (SAVI). Remote Sens Environ 25(3):295–309 Hwang T, Gholizadeh H, Sims DA, Novick KA, Brzostek ER, Phillips RP, Roman DT, Robeson SM, Rahman AF (2017) Capturing species-level drought responses in a temperate deciduous forest using ratios of photochemical reflectance indices between sunlit and shaded canopies. Remote Sens Environ 199:350–359 IMSD (1995) Integrated Mission for Sustainable Development-Madnur Area, Nizamabad District; AP. Intergrated Survey Group, NRSA, Dept. of Space, Govt. of India, Hyderabad James ME, Kalluri SNV (1994) The Pathfinder AVHRR land data set: an improved coarse resolution data set for terrestrial monitoring. Int J Remote Sens 15(17):3347–3363

6 Identification of Environmental Epidemiology Through Advanced …

141

Jasinski MF (1990) Sensitivity of the normalised difference vegetation index to subpixel canopy cover, soil albedo, and pixel scale. Remote Sens Environ 32(2–3):169–187 Kasoro FR, Yan L, Zhang W, Asante-Badu B (2021) Spatial and temporal changes of vegetation cover in china based on modis ndvi. Appl Ecol Environ Res 19:1371–1390 Kameswara Rao SVC (1995) Cropping pattern analysis using temporal IRS-LISS II data—a case study of Madnur Watershed, Nizamabad District, AP. Agric and Soil Division, IIRS, NRSA, Department of Space, Govt. of India. Dehradun Kriegler FJ, Malila WA, Nalepka RF, Richardson W (1969) Preprocessing transformations and their effect on multispectral recognition. Remote Sens Environ VI:97–132 Krofcheck DJ, Eitel JUH, Vierling LA, Schulthess U, Hilton TM, Dettweiler-Robinson E, Pendleton R, Litvak ME (2014) Detecting mortality induced structural and functional changes in a piñonjuniper woodland using Landsat and RapidEye time series. Remote Sens Environ 151:102–113 Kundu A, Dwivedi S, Dutta D (2016) Monitoring the vegetation health over India during contrasting monsoon years using satellite remote sensing indices. Arab J Geosci 9:144 Laliberte AS, Rango A, Havstad KM, Paris JF, Beck RF, McNeely R, Gonzalez AL (2004) Objectoriented image analysis for mapping shrub encroachment from 1937 to 2003 in Southern New Mexico. Remote Sens Environ 93(1):198–210 Luo N, Mao D, Wen B, Liu X (2020) Climate change affected vegetation dynamics in the Northern Xinjiang of China: evaluation by SPEI and NDVI. Land 9:90 Maclean IM, Wilson RJ (2011) Recent ecological responses to climate change support predictions of high extinction risk. Proc Natl Acad Sci USA 2011(108):12337–12342 Mänd P, Hallik L, Peñuelas J, Nilson T, Duce P, Emmett BA, Beier C, Estiarte J, Kalapos T, Schmidt IK, Kovács-Lang E, Preto P, Tietema A, Westerveld JW, Kull O (2010) Responses of the reflectance indices PRI and NDVI to experimental warming and drought in European shrublands along a north-south climatic gradient. Remote Sens Environ 114(3):626–636 Mohmmed A, Zhang K, Kabenge M, Keesstra CA, Reuben M, Elbashier MMA, Dalson T, Ali AAS (2018) Analysis of drought and vulnerability in the North Darfur region of Sudan. Land Degrad Dev 29:4424–4438 Murthy C, Singh J, Kumar P, Sai MS (2017) A composite index for drought hazard assessment using CPC rainfall time series data. Int J Environ Sci Technol 14:1981–1988 Mutanga O, Adam E, Cho MA (2012) High density biomass estimation for wetland vegetation using WorldView-2 imagery and ramdon forest regression algorithm. Int J Appl Earth Observat Geoinf 18:399–406 Nicolai-Shaw N, Zscheischler J, Hirschi M, Gudmundsson L, Seneviratne SI (2017) A drought event composite analysis using satellite remote sensing-based soil moisture. Remote Sens Environ 203:216–225 Padmavathi K, Thangadurai K (2016) Implementation of RGB and grayscale images in plant leaves disease detection–comparative study. Indian J Sci Technol 9(6):1–6 Patel NR, Chopra P, Dadhwal VK (2007) Analyzing spatial patterns of meteorological drought using standardized precipitation index. Meteorol Appl 14:329–336 Pervez MS, Budde M, Rowland J (2014) Mapping irrigated areas in Afghanistan over the past decade using MODIS NDVI. Remote Sens Environ 149:155–165 Pastor-Guzman J, Atkinson P, Dash J, Rioja-Nieto R (2015) Spatiotemporal variation in mangrove chlorophyll concentration using Landsat 8. Remote Sens 7(11):14530–14558 Pettorelli N, Vik JO, Mysterud A, Gaillard JM, Tucker CJ, Stenseth NC (2005) Using the satellitederived NDVI to assess ecological responses to environmental change. Trends Ecol Evol 20(9):503–510 Rojas O, Vrieling A, Rembold F (2011) Assessing drought probability for agricultural areas in Africa with coarse resolution remote sensing imagery. Remote Sens Environ 115(2):343–352 Rouse JW, Haas RH, Scheel JA, Deering DW (1974) Monitoring vegetation systems in the great plains with ERTS. In: Proceedings, 3rd earth resource technology satellite (ERTS) symposium, vol 1, pp 48–62

142

V. Kumar et al.

Sarkar S, Kafatos M (2004) Interannual variability of vegetation over the Indian sub-continent and its relation to the different meteorological parameters. Remote Sens Environ 90(2):268–280 Sayago S, Ovando G, Bocco M (2017) Landsat images and crop model for evaluating water stress of rainfed soybean. Remote Sens Environ 198:30–39 Sehgal VK, Dhakar R (2016) Geospatial approach for assessment of bioEcomonic vulnerability to agricultural drought and its intra-seasonal variations. Environ Monit Assess 1 88(3):1–97 Singh RP, Roy S, Kogan F (2003) Vegetation and temperature condition indices from NOAA AVHRR data for drought monitoring over India. Int J Remote Sens 24:4393–4402 Swain PH, Davis SM (1978) Remote sensing: the quantitative approach. McGraw-Hill, USA Tian J, Wang L, Li X, Gong H, Shi C, Zhong R, Liu X (2017) Comparison of UAV and WorldView-2 imagery for mapping leaf area index of mangrove forest. Int J Appl Earth Obs Geoinf 61:22–31 Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BF, Siqueira MD, Grainger A, Hannah L (2004) Extinction risk from climate change. Nature 2004(427):145 Weier J, Herring D (2000) Measuring vegetation (NDVI & EVI). NASA Earth Observatory, 20 Yu F, Price KP, Ellis J, Shi P (2003) Response of seasonal vegetation development to climatic variations in eastern central Asia. Remote Sens Environ 87(1):42–54 Zhao T, Stark B, Chen YQ, Ray AL, Doll D (2017) Challenges in water stress quantification using small unmanned aerial system (sUAS): lessons from a growing season of almond. J Intell Robotic Syst Theory Appl 88(2–4):721–735. https://doi.org/10.1007/s10846-017-0513-x Zheng H, Shen G, He X, Yu X, Ren Z, Zhang D (2015) Spatial assessment of vegetation vulnerability to accumulated drought in Northeast China. Reg Environ Chang 15:1639–1650 Zheng H, Cheng T, Li D, Zhou X, Yao X, Tian Y, Cao W, Zhu Y (2018) Evaluation of RGB, colorinfrared and multispectral images acquired from unmanned aerial systems for the estimation of nitrogen accumulation in rice. Remote Sens 10(6):824 Zhu X, Liu D (2015) Improving forest aboveground biomass estimation using seasonal Landsat NDVI time-series. ISPRS J Photogramm Remote Sens 102:222–231 Zribi M, Dribi G, Amri R, Lili-Chananne Z (2016) Analysis of the effects of drought on vegetation cover in a Mediterranean region through the use of SPOT-VGT and TERRA-MODIS long time series. Remote Sens 8(12):992

Chapter 7

Assessment of Land Utilization Pattern and Their Relationship with Surface Temperature and Vegetation in Sikkim, India Shashi Sekhar, Nitu Singh, Sudhir Kumar Singh, Meenakshi Dhote, and Kumar Rajnish

Abstract The land use/land cover change (LULC) is prominent in the hilly states of India. Population growth and consumerism have impacted the dynamics of LULCC. In this work, we have used satellite data to understand the NDVI and LST change over the period of 1995–2005–2021. The LULC maps were prepared and analysed to find out the change in different LULC classes. In contrast to land use classes like water bodies, agricultural land, rocky or barren, and scrubland or grassland found decreased whereas dense forest cover and snow or glaciers showed overall increased slightly from 1995–2021. The maximum change areal extent was changed from 1448.91 (1995) to 765.20 km2 (2021) into other categories. The LST and NDVI showed a drastic change over the studied period in the Sikkim state, India. The statistical analysis between NDVI and LST shows the relevant positive coefficient of determination for the years 1995 (R2 = 0.47), 2005 (R2 = 0.52), and 2021 (R2 = 0.46), respectively.

S. Sekhar · M. Dhote School of Planning and Architecture, ENVIS RP, IP Estate, New Delhi 110002, India e-mail: [email protected] M. Dhote e-mail: [email protected] N. Singh Survey of India, Pushpa Bhawan, New Delhi, India e-mail: [email protected] S. K. Singh (B) K. Banerjee Centre of Atmospheric and Ocean Studies, IIDS, Nehru Science Centre, University of Allahabad, Prayagraj 211002, Uttar Pradesh, India e-mail: [email protected] K. Rajnish ENVIS Cell, Ministry of Environment, Forest and Climate Change, GoI, New Delhi, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_7

143

144

S. Sekhar et al.

Keywords NDVI · LST · Sikkim · Correlation · Bivariate analysis

7.1 Introduction The land use/land cover (LULC) change is very common throughout the globe to meet the needs of the rising human population (Pyngrope et al. 2021; Sahu et al. 2022). Land cover refers to the surface cover on the ground, whereas land use refers to the purpose the land serves and both are interchangeable. Even though the hill regions LULC are also continuously getting modified or converted from natural to the manmade system; these modifications lead to an imbalance in the services of the natural ecosystems. Many researchers have identified the causes of these changes, their impacts, and their relationships using satellite data and modelled future LULC (Kushwaha et al. 2021; Munthali et al. 2020; Singh et al. 2015, 2018). The longterm satellite data analysis aids to understand the trend of change and its associated impacts on the humans and environment (Kumar et al. 2023; Singh et al. 2022). The spectral indices-based analysis of satellite data is an easy to apply and reliable method for getting the information. There are a variety of spectral indices in practice for different analyses, namely normalized difference vegetation index (NDVI) used to study the crop and vegetation health and greenness (Kumar et al. 2018), soil adjusted value index (SAVI) used for the soil moisture estimation (Kumar et al. 2018); normalized difference water index (NDWI) for extraction of water surface (Gašparovi´c and Singh 2022; Balázs et al. 2018); many others for the estimation of soil salinity and the built environment. The NDVI index is very common and robust for the analysis of vegetation health and it is also used as a predictor for the change in the environmental conditions (Rawat and Singh 2018; Szabó et al. 2019). Land Surface Temperature (LST) is the radiative skin temperature of the land surface, as measured in the direction of the remote sensor (Kustas and Anderson 2009; Zhang et al. 2021). LST is an important factor in coupled surface-atmosphere interactions and regulates energy fluxes between ground and atmosphere (Ahmed et al. 2020; Brunsell and Gillies 2003; Kikon et al. 2016). Nowadays, land surface temperature (LST) is also quantified by many researches to understand how the modification and conversion of land use regulate the LST and understand the urban heat island effects (Gašparovi´c et al. 2021). LST estimation using satellite data for different purposes has been quantified by the researchers. Researchers have also explored and developed the linkage between NDVI and LST (Alademomi et al. 2022; Ivanova et al. 2020; Naga Rajesh et al. 2022). Sharma et al. (2022) assessed the sensitivity of soil moisture, NDVI, and LST for Gautam Buddha Nagar, India, and reported that NDVI is as more sensitive to LST. Seasonal variability of LST-NDVI correlation on different LULC was investigated through Landsat images for Raipur City, India (Guha and Govil 2022). We aimed to study the LULC change and identified the NDVI and LST relation for the north east region of India, i.e., Sikkim state using the freely accessible data. This study will help in understanding the causes of LULC change and how the NDVI and

7 Assessment of Land Utilization Pattern and Their Relationship …

145

LST can be used as criteria for identifying the problem and providing the solution to environmental problems.

7.2 Study Area Description Sikkim (Fig. 7.1) is a state in northeast India (latitudes of 27°5, N to 20°9, N and longitudes of 87°59, E to 88°56, ), bordering countries are Bhutan, Tibet, and Nepal. It has India’s highest mountain, 8586 m Kangchenjunga. The state is rich in faunal diversity, glaciers, and alpine meadows. The climate of the state ranges from sub-tropical in the south to tundra in the north. Temperature exceeding 28 °C in summer and an average annual temperature is 18 °C. The state has four seasons, i.e., winter (December–February), spring (March–May), South-West Monsoon (June– September), and retreating monsoon (October–November). The average number of rainy days spans from 100 days (Thangu in North Sikkim) to 184 days (Gangtok in East Sikkim). The maximum rainfall is observed in the South Sikkim district (2700– 2800 mm), however, East Sikkim receives the lowest annual rainfall of 2045 mm. Five forest types are reported such as Sub Tropical, Moist Mixed Deciduous, Wet Temperate, Conifer, and Sub-Alpine forest. The state has 30.77% of the forest as protected area. In the state of Sikkim, urbanization has mainly been driven by the growth of administrative and commercial activities and the per capita net income has improved significantly in the last couple of decades. The state income is largely dependent on agriculture and tourism.

7.3 Methodology 7.3.1 Data Sets Used The Landsat 4, 5, and 8 OLI sensor data sets of 1996, 2006, and 2021 were used and downloaded at no cost from the United States Geological Survey (USGS) for the analysis. Table 7.1 showed brief information about the data used for the estimation of LULC, NDVI, and LST. In this study, level-1 data of Landsat TM, ETM band OLI/TIRS sensors were used. The pre-processing of satellite data such as radiometric correction, geometric correction, etc., was performed using ERDAS Imagine software. The digital numbers were converted into reflectance and then further analysis was applied as mentioned in the methodology flowchart. LST was retrieved through the TIR bands of Landsat data sets (band 6 for TM and ETM data, whereas band 10 for OLI/ TIRS data). Near Infra-red (NIR) and Red bands were used for NDVI generation, while red, NIR, and TIR bands for Land Surface Temperature (LST). The schematic flowchart showed the applied methodological steps during this research work (Fig. 7.2).

146

S. Sekhar et al.

Fig. 7.1 The map showed the study area location of Sikkim State, India

Table 7.1 Brief information about the data sets used for the LULC, NDVI, and LST estimation Satellite data

Sensors

Date of acquisition

Path and row

Number of bands used

Spatial resolution (m)

Landsat 4

MSS and TM

1995

139 and 041

07

30 and 60

2005

08

30 and 100

OLI and TIRS

2021

08

30 and 100

Landsat 5 Landsat 8

7.3.2 Land Use/Cover Classification and Accuracy Assessment The derived LST, NDVI, and LULC images of the same month and year were taken into account. The estimation of land utilization pattern in the study area was derived by LULC classification of satellite imageries for the years 1995, 2005, and 2021 (Fig. 7.3). In addition to this, the study also assesses the changes that takes place in the land utilization pattern using the change detection method. The unsupervised classification technique was applied as the prior knowledge about the study area was limited. Hence, the unsupervised classification technique provides better results. The image was classified as per the classification scheme of the National Remote Sensing Centre, Hyderabad, India. Total of seven LULC classes were identified from the satellite data. LULC map of 2021 was re-sampled on 30 m to match the

7 Assessment of Land Utilization Pattern and Their Relationship …

147

Fig. 7.2 The depicted flow chart shows the adopted methodology in this work

spatial resolution of all the classified maps thereafter, the change detection technique was performed using ArcGIS 10.3.1 for analyzing the changes through the periods. The change matrix presents important information about the spatial distribution of changes in LULC (Shalaby and Tateishi 2007). A change matrix showing the land cover changes in each decade was generated from classified images of 1995–2005,

148

S. Sekhar et al.

2005–2021, and a change matrix was generated from 1995 to 2005 to assess the overall changes in LULC classes between 1995 and 2021.

7.3.3 NDVI Computation The NDVI is a measure of greenness. The NDVI was computed using the near infrared and red band reflectance. NDVI is calculated by using the red and infra-red bands of Landsat data based on Eq. 7.1 proposed by Rouse et al. (1974). The ratio of the NIR and red band is used for the calculation because absorption by chlorophyll of these two bands of the electromagnetic spectrum is highest (Bindi et al. 2009). N DV I = (ρ N I R − ρ R)/(ρ N I R + ρ R)

(7.1)

where, ρ N I R = Reflectance in Near Infra-red band and ρ R = Reflectance in red band.

7.3.4 LST Computation A comparative analysis has been done to assess land surface temperature (LST) using Mono Window Algorithm for 1995, 2005, and 2021 on Landsat 5 and 8. There are a series of steps and equations needed to retrieve LST from satellite imagery which is mentioned below. First, OLI and TIRS band data are converted to radiance using the radiance rescaling factors provided in the metadata file by using Eq. (7.2) Lλ = M L ∗ Qcal + AL

(7.2)

In the second step, band data are converted into reflectance using reflectance rescaling coefficients provided in the product metadata file. Equation (7.3) was used to convert DN values to TOA reflectance for OLI data. ρλ = Mρ ∗ Qcal + Aρ

(7.3)

Now, TIRS band data are converted from spectral radiance to brightness temperature using the thermal constants provided in the metadata file using Eq. (7.4) BT =

ln[

K2 ( K1) Lλ

+ 1]

(7.4)

Finally, to obtain the results in Celsius, the radiant temperature is now converted using Eq. (7.5)

7 Assessment of Land Utilization Pattern and Their Relationship …

T ◦ C = T (K ) − 273.15

149

(7.5)

7.3.5 Correlation Analysis To analyze the relationship between NDVI and LST, correlation is done in a statistical package for social science (SPSS) software and a scatter plot is prepared to assess the correlation amongst the variables. ⎡

⎤ } (Y Y )(Y − Y ) − t t p p i=1 ⎦ / r = ⎣/ Σn 2 Σn 2 i=1 (Yt − Y t ) i=1 (Y p − Y p ) Σn {

(7.6)

where, r = correlation coefficient and value r ranges from −1 to +1; if r = +1 means strongest positive relation, r = −1 discloses the strongest negative correlation and r = 0 means no correlation.

7.4 Results and Discussion 7.4.1 Accuracy Assessment Accuracy assessment is considered as the most important and crucial step for image classification to check whether the classification technique is reliable or not (Singh et al. 2018; Yinga et al. 2022). Furthermore, the selection of pixels or points for accuracy assessment over the study area should be clearly recognized on satellite imageries and other platforms such as Google Earth. In this study, a total of 706 random points (locations) were considered for the assessment using Erdas-Imagine software version 15. The accuracy assessment of the classified images of 1995, 2005, and 2021 showed the producer’s accuracy, user’s accuracy, overall accuracy, and Kappa coefficient. A Kappa analysis is a well-known discrete multivariate technique that produces a Khat statistic (an estimate of Kappa) indicates a measure of agreement or accuracy. The Kappa value lies in the range of +1.0 to −1.0, if the Kappa coefficient having positive, then it indicates the high accuracy. However, a zero value of Kappa coefficient indicates there is no correlation in the classification. These assessments usually help to identify and quantify overall classification accuracy by comparing classified map data with reference data. In this work, the results show better and more robust classification taken place with 91.87% (1995), 91.4% (2005), and 90.95% (2021), respectively. The built has the highest producer accuracy (100%) and scrub land showed the lowest (45.83%) as the cropland showed

150

S. Sekhar et al.

Table 7.2 Classification accuracy statistics Class name

2021 Producers accuracy (%)

2005 Users accuracy (%)

Producers accuracy (%)

1995 Users accuracy (%)

Producers accuracy (%)

Users accuracy (%)

Dense forest

98.95

80.34

99.33

87.65

99.09

88.62

Water bodies

75.00

100.00

36.36

80.00

36.36

80.00

Crop land/ Agriculture

59.38

86.36

76.09

81.40

78.05

86.49

Built up

100.00

100.00

100.00

75.00

100.00

66.67

Rocky barren

91.18

93.94

59.35

85.88

52.94

87.10

Glacier

97.65

97.65

93.94

89.30

95.71

86.19

Scrub land

45.83

78.57

76.67

85.19

92.50

82.22

Overall accuracy

91.87

Table 7.3 Conditional kappa statistics for each category

91.40

Class name

90.95

Kappa 2021

2005

1995

Dense forest

0.74

0.83

0.84

Water bodies

1.00

0.79

0.79

Agriculture

0.85

0.80

0.85

Built up

1.00

0.74

0.66

Rocky barren

0.93

0.84

0.85

Glacier

0.97

0.86

0.82

Scrub land

0.77

0.84

0.81

Overall Kappa

0.89

0.88

0.88

the lowest producer accuracy (Table 7.2). The conditional kappa statistics for each category are mentioned in Table 7.3.

7.4.2 LULC Change Dynamics The dense forest area has increased from 2897 to 3133.94 km2 and the average increment was around 3% over the studied period from 1995 to 2021 (Table 7.4). The water body area was slightly changed which is evident due to the loss of the surface stream. The cropland area was decreased and the built-up area was increased. The rock and barren area also showed a decreasing trend due to the construction activities and removal of vegetative cover. The class snow/glaciers showed an increase of 29.53 during the studied period (1995–2021). The scrub/grassland showed a decreasing

7 Assessment of Land Utilization Pattern and Their Relationship …

151

Table 7.4 The Sikkim land utilization pattern statistical distribution Class name

1995 km2

2005 km2

%

2021 km2

%

%

Dense forest

2897.58

40.67

2845.20

39.93

3133.94

Water bodies

52.96

0.74

51.35

0.72

49.60

0.70

636.99

8.94

687.43

9.65

463.89

6.51

Cropland

43.99

20.82

0.29

22.63

0.32

47.16

0.66

Rocky barren

1448.91

20.34

810.77

11.38

765.20

10.74

Snow/glaciers

1241.80

17.43

1985.86

27.87

2103.69

29.53

Built up

Scrub/grassland Total

825.54

11.59

721.36

10.12

561.12

7.88

7124.60

100.00

7124.60

100.00

7124.60

100.00

trend. The results showed that dense forest remained unchanged almost and other LULC classes have changed (Fig. 7.4). Population pressure and LULC policies are the main reason for LULC change in the region. An alteration in land use produces a change in land cover hence the term LULC change, this can be detected principally by the use of field survey and analysis of remote sensing imagery and it can be due to industrialization, urbanization, deforestation (Das and Angadi 2022; Gull et al. 2022; Mishra et al. 2020; Prokop and Płoskonka 2014; Singh et al. 2018).Urbanization in the Himalayas region occurs through dual pathways either directly by altering LULC to meet these requirements or indirectly in the form of infrastructure constructions to meet their communication, commuting, and energy demands (Anees et al. 2022). Infestation in the eastern Himalayan state of Sikkim due to invasive species affects the forest ecosystem (Kumar et al. 2022).

7.4.3 LST and NDVI Dynamics (1995–2005–2021) The LST (Fig. 7.5) was max (35.65 °C) for the rocky barren with a mean of 2.49 °C and min (−25.96 °C) for snow/glacier with an average of −3.07 °C for the year 1995. For the year 2005 the min LST was −36.70 °C and max for 34.46 °C. In case of 2021, the max LST was reported for 27.96 °C for cropland and min for rocky barren and snow/glaciers (−12.38 °C). The highest NDVI was reported for the dense forest (Table 7.5). Zheng et al. (2021) combined NPP-VIIRS nighttime light, NDVI, NDWI, and NDBI to monitor urban built-up areas. Chakraborty and Chanda (2022) carried out the vegetation dynamics analysis using multi-temporal satellite data and reported that forest areas are continuously decreasing from south to north in North Sikkim. However, due to strict legal regulations, different stake holders and community intervention, dense forest cover has been the least degraded forest in the last 30 years, and forest cover has increased across various Himalayan regions (Mondal and Zhang 2018). Guha and Tiwari (2022) performed a temporal change

Fig. 7.3 The LULC map of the year 1995, 2005, and 2021 of Sikkim, India

152 S. Sekhar et al.

153

Fig. 7.4 The LULC change analysis map of 1995–2005 and 2005–2021

7 Assessment of Land Utilization Pattern and Their Relationship …

154

S. Sekhar et al.

analysis of glaciers using area changes, retreat, and surface elevation changes for the entire Sikkim and reported that surface elevation shows a negative trend. Sharma and Thapa (2021) studied the forest fire in Sikkim using visual interpretation and reported that the West district of Sikkim recorded the highest number of forest fire incidences followed by the south and east districts; the north district was least affected due to low rainfall, dry winter season, and type of vegetation. Peng et al. (2020) used Landsat 8 OLI/TIRS data, and radiative transfer model to retrieve the LST and study the influence of topographic factors (elevation, slope, aspect, and shaded relief) of Hangzhou, China. They reported that elevation and slope are negatively correlated with LST. Table 7.5 The statistics of LST and NDVI, LULC categories wise of respective years (1995, 2005, and 2021) Year/ Variable

LST (°C)

Statistics

MIN

1995

2005 MAX

MEAN

MIN

2021 MAX

MEAN

MIN

MAX

MEAN

Cropland

−0.76

28.77

19.41

−0.23

33.66

21.55

6.55

27.96

21.97

Built up

13.31

30.42

19.12

8.98

27.52

18.40

9.40

23.46

19.13

Rocky barren

−21.78

35.65

2.49

−36.70

32.05

−3.04

−12.38

27.30

8.36

Water Bodies

−7.44

26.25

13.51

−31.14

31.65

9.06

−3.67

27.29

12.76

−11.56

30.01

5.08

−26.68

32.05

0.64

−10.18

27.95

10.45

−8.60

34.06

11.94

−17.80

33.26

9.25

−1.51

27.98

16.90

−25.96

24.97

−3.07

−39.23

34.46

−6.81

−12.38

17.11

4.75

MAX

MEAN

Scrub/ grassland Dense forest Snow/ glaciers

Year/Variable

NDVI

Statistics

MIN

1995

2005 MAX

MEAN MIN

2021 MAX

MEAN MIN

Cropland

−0.04 0.03

−0.01

0.03

0.39

0.18

0.53

0.56

0.54

Built up

−0.07 −0.05 −0.06

0.11

0.16

0.13

0.05

0.21

0.12

0.11

0.11

Rocky barren

−0.08 0.01

−0.03

−0.13 −0.02 −0.07

0.10

Water Bodies

−0.25 −0.17 −0.21

−0.09 −0.01 −0.06

−0.04 −0.01 −0.03

0.08

0.14

0.29

0.42

0.52

0.42

0.67

0.79

0.71

−0.06

−0.04 0.10

0.00

Scrub/grassland 0.16

0.25

Dense forest

0.16

0.65

0.50

0.33

Snow/glaciers

−0.05 0.00

0.00

−0.15 0.00

0.20

0.11

0.36

Fig. 7.5 The NDVI and LST maps of respective years (1995, 2005, and 2021)

7 Assessment of Land Utilization Pattern and Their Relationship … 155

156

S. Sekhar et al.

7.4.4 Correlation Between LST Versus NDVI The analysis of Fig. 7.6 showed the LST versus NDVI coefficient of determination for the years 1995 (R2 = 0.47), 2005 (R2 = 0.52), and 2021 (R2 = 0.46). A high NDVI value showed a higher greenness or good vegetation health. The lowest LSTs usually are found in areas with high NDVI. The study has reported the negative correlation between NDVI and LST and it can be utilized for urban climate studies (Yuan and Bauer 2007). Weng et al. (2004) outlined that the higher vegetation fraction leads to lower LST and showed a stronger negative correlation between NDVI with LST. Yue et al. (2007) explained different land use types have different NDVI and LST and hence varied the correlation. Joshi and Bhatt (2012) explained that the areas with higher water bodies and vegetation have lower temperature as compared to the built-up areas as water bodies and vegetation regulates the air temperature. Sun and Kafatos (2007) stated that the correlation between LST and NDVI was positive for winter and negative during warm seasons.

7.5 Conclusion There is a relevant change in the LULC observed in the study area during the studied period (1996–2006–2021). The environmental changes can be monitored by LST and NDVI. Hence, it is important to explore the relationship between NDVI and LST in the hilly area. The NDVI showed an increasing trend from 1996 to 2021, whereas the LST showed a decreasing trend for the same period. This is also corroborated by the inverse relationship between NDVI and LST. The LST is decreasing as the vegetation cover is increasing in the region at the state level. The final scale dynamics information is required for better visualizations and for developing a caustic effect analysis between NDVI and LST. The highest NDVI was observed in the year 2021 (0.83), however, the LST was observed for the same year as 28.97 °C. The seasonal level change in NDVI and LST need will be performed at the state level for an improved understanding of the relationship between NDVI and LST.

-40.00

-20.00

0.00

20.00

40.00

0.00

NDVI

0.20

NDVI

0.40

0.60

0.80

1.00

Fig. 7.6 The correlation plot of NDVI versus LST of 1995, 2005, and 2021

-15.00

-0.20-5.00 0.00

5.00

15.00

1.00

y = 17.093x + 6.1298 R² = 0.4599

0.50

y = 24.52x + 6.021 R² = 0.47

Correlation with LST & NDVI Sikkim March 2021

25.00

-0.50

LST

LST

-0.40

-0.20

-40.00

-20.00

0.00

20.00

40.00

0.00

NDVI

0.20

0.40

0.60

0.80

y = 50.365x - 0.4792 R² = 0.52

Correlation between LST & NDVI 2005

LST

Correlation with LST and NDVI SIkkim 1995

7 Assessment of Land Utilization Pattern and Their Relationship … 157

158

S. Sekhar et al.

Acknowledgements The authors are thankful to the editor and USGS for providing the satellite data at no cost. The authors are also thankful to the Head of their Centre for providing the necessary facility to carry out this work.

References Ahmed HA, Singh SK, Kumar M, Maina MS, Dzwairo R, Lal D (2020) Impact of urbanization and land cover change on urban climate: case study of Nigeria. Urban Clim 32:100600 Alademomi AS, Okolie CJ, Daramola OE, Akinnusi SA, Adediran E, Olanrewaju HO, Odumosu J et al (2022) The interrelationship between LST, NDVI, NDBI, and land cover change in a section of Lagos metropolis, Nigeria. Appl Geomatics 14(2):299–314 Anees MM, Sharma R, Joshi PK (2022) Urbanization in Himalaya—An interregional perspective to land use and urban growth dynamics. In: Mountain landscapes in transition. Springer, Cham, pp 517–538 Balázs B, Bíró T, Dyke G, Singh SK, Szabó S (2018) Extracting water-related features using reflectance data and principal component analysis of Landsat images. Hydrol Sci J 63(2):269– 284 Bindi M, Brandani G, Dessì A, Dibari C, Ferrise R, Moriondo M, Trombi G (2009) Impact of climate change on agricultural and natural ecosystems. Am J Environ Sci 5(5):633–638 Brunsell NA, Gillies RR (2003) Length scale analysis of surface energy fluxes derived from remote sensing. J Hydrometeorol 4(6):1212–1219 Chakraborty S, Chanda A (2022) Assessment of forest cover change, community responses, and conservation strategy: evidence from North Sikkim District, India. In: Conservation, management and monitoring of forest resources in India. Springer, Cham, pp 221–257 Das S, Angadi DP (2022) Land use land cover change detection and monitoring of urban growth using remote sensing and GIS techniques: a micro-level study. GeoJournal 87(3):2101–2123 Gašparovi´c M, Zori´c Š, Singh SK (2021) Urbanisation impact on creation of heat islands in large cities. Int Arch Photogram Remote Sens Spat Inf Sci 43:313–318 Gašparovi´c M, Singh SK (2022) Urban surface water bodies mapping using the automatic kmeans based approach and sentinel-2 imagery. Geocarto Int 2148757. https://doi.org/10.1080/ 10106049.2022.2148757 Guha S, Govil H (2022) Seasonal variability of LST-NDVI correlation on different land use/land cover using Landsat satellite sensor: a case study of Raipur City, India. Environ Dev Sustain 24(6):8823–8839 Guha S, Tiwari RK (2022) Analysis of differential glacier behaviour in Sikkim Himalayas in view of changing climate. Geocarto Int 1–23 Gull S, Shah SR, Dar AM (2022) Assessing land use/land cover change detection of north-eastern watersheds of Kashmir valley using GIS and remote sensing techniques. Water Pract Technol 17(8):1603–1614 Ivanova Y, Kovalev A, Soukhovolsky V (2020) Modeling the radial stem growth of the pine (Pinus sylvestris L.) forests using the satellite-derived NDVI and LST (MODIS/AQUA) data. Atmosphere 12(1):12 Joshi JP, Bhatt B (2012) Estimating temporal land surface temperature using remote sensing: A study of Vadodara urban area, Gujarat. Int J Geol Earth Environ Sci 2(1):123–130 Kikon N, Singh P, Singh SK, Vyas A (2016) Assessment of urban heat islands (UHI) of Noida City, India using multi-temporal satellite data. Sustain Cities Soc 22:19–28 Kumar N, Singh SK, Pandey HK (2018) Drainage morphometric analysis using open access earth observation datasets in a drought-affected part of Bundelkhand, India. Appl Geomatics 10(3):173–189

7 Assessment of Land Utilization Pattern and Their Relationship …

159

Kumar R, Singh A, Pandey U, Srivastava P, Mehra S (2022) Mapping the extent of invasive species: an assessment based on high-resolution data for selected species in parts of Eastern Himalaya in Sikkim. In: Forest dynamics and conservation. Springer, Singapore, pp 249–259 Kumar N, Singh VG, Singh SK, Behera DK, Gašparovi´c M (2023) Modeling of land use change under the recent climate projections of CMIP6: a case study of Indian river basin. Environ Sci Pollut Res. https://doi.org/10.1007/s11356-023-26960-z Kushwaha K, Singh MM, Singh SK, Patel A (2021) Urban growth modeling using earth observation datasets, Cellular Automata-Markov Chain model and urban metrics to measure urban footprints. Remote Sens Appl Soc Environ 22:100479 Kustas W, Anderson M (2009) Advances in thermal infrared remote sensing for land surface modeling. Agric for Meteorol 149(12):2071–2081 Mishra PK, Rai A, Rai SC (2020) Land use and land cover change detection using geospatial techniques in the Sikkim Himalaya, India. Egypt J Remote Sens Space Sci 23(2):133–143 Mondal PP, Zhang Y (2018) Research progress on changes in land use and land cover in the western Himalayas (India) and effects on ecosystem services. Sustainability 10(12):4504 Munthali MG, Mustak S, Adeola A, Botai J, Singh SK, Davis N (2020) Modelling land use and land cover dynamics of Dedza district of Malawi using hybrid Cellular Automata and Markov model. Remote Sens Appl Soc Environ 17:100276 Naga Rajesh A, Abinaya S, Purna Durga G, Lakshmi Kumar TV (2022) Long-term relationships of MODIS NDVI with rainfall, land surface temperature, surface soil moisture and groundwater storage over monsoon core region of India. Arid Land Res Manage 1–20 Peng X, Wu W, Zheng Y, Sun J, Hu T, Wang P (2020) Correlation analysis of land surface temperature and topographic elements in Hangzhou, China. Sci Rep 10(1):1–16 Prokop P, Płoskonka D (2014) Natural and human impact on the land use and soil properties of the Sikkim Himalayas piedmont in India. J Environ Manage 138:15–23 Pyngrope OR, Kumar M, Pebam R, Singh SK, Kundu A & Lal D (2021) Investigating forest fragmentation through earth observation datasets and metric analysis in the tropical rainforest area. Abstract SN App Sci 3(7). https://doi.org/10.1007/s42452-021-04683-5 Rawat KS, Singh SK (2018) Appraisal of soil conservation capacity using NDVI model-based C factor of RUSLE model for a semi arid ungauged watershed: a case study. Water Conserv Sci Eng 3(1):47–58 Sahu SR, Rawat KS, Singh SK, Bahuguna A( 2022, November). Land use land cover (LU/LC) change analysis using earth observation data sets over Jharsuguda districts of Odisha. In AIP Conference Proceedings (Vol. 2481, No. 1, p. 020040). AIP Publishing LLC Shalaby A, Tateishi R (2007) Remote sensing and GIS for mapping and monitoring land cover and landuse changes in the Northwestern coastal zone of Egypt. Appl Geogr 27(1):28–41. https:// doi.org/10.1016/j.apgeog.2006.09.004 Sharma K, Thapa G (2021) Analysis and interpretation of forest fire data of Sikkim. For Soc 5(2):261–276 Sharma M, Bangotra P, Gautam AS, Gautam S (2022) Sensitivity of normalized difference vegetation index (NDVI) to land surface temperature, soil moisture and precipitation over district Gautam Buddh Nagar, UP, India. Stoch Env Res Risk Assess 36(6):1779–1789 Singh SK, Mustak S, Srivastava PK, Szabó S, Islam T (2015) Predicting spatial and decadal LULC changes through cellular automata Markov chain models using earth observation datasets and geo-information. Environ Processes 2(1):61–78 Singh SK, Laari PB, Mustak SK, Srivastava PK, Szabó S (2018) Modelling of land use land cover change using earth observation data-sets of Tons River Basin, Madhya Pradesh, India. Geocarto Int 33(11):1202–1222 Singh VG, Singh SK, Kumar N, Singh RP (2022) Simulation of land use/land cover change at a basin scale using satellite data and markov chain model. Geocarto Int 1–26 Sun D, Kafatos M (2007) Note on the NDVI-LST relationship and the use of temperature-related drought indices over North America. Geophys Res Lett 34(24)

160

S. Sekhar et al.

Szabó S, Elemér L, Kovács Z, Püspöki Z, Kertész Á, Singh SK, Balázs B (2019) NDVI dynamics as reflected in climatic variables: spatial and temporal trends–a case study of Hungary. Gisci Remote Sens 56(4):624–644 Weng Q, Lu D, Schubring J (2004) Estimation of land surface temperature–vegetation abundance relationship for urban heat island studies. Remote Sens Environ 89(4):467–483 Yinga OE, Kumar KS, Chowlani M, Tripathi SK, Khanduri VP, Singh SK (2022). Influence of land-use pattern on soil quality in a steeply sloped tropical mountainous region, India. Arch Agro Soil Sci 68(6):852–872 Yuan F, Bauer ME (2007) Comparison of impervious surface area and normalized difference vegetation index as indicators of surface urban heat island effects in Landsat imagery. Remote Sens Environ 106(3):375–386 Yue W, Xu J, Tan W, Xu L (2007) The relationship between land surface temperature and NDVI with remote sensing: application to Shanghai Landsat 7 ETM+ data. Int J Remote Sens 28(15):3205– 3226 Zhang C, Long D, Zhang Y, Anderson MC, Kustas WP, Yang Y (2021) A decadal (2008–2017) daily evapotranspiration data set of 1 km spatial resolution and spatial completeness across the North China Plain using TSEB and data fusion. Remote Sens Environ 262:112519 Zheng Y, Tang L, Wang H (2021) An improved approach for monitoring urban built-up areas by combining NPP-VIIRS nighttime light, NDVI, NDWI, and NDBI. J Clean Prod 328:129488

Chapter 8

Monitoring Land Use and Land Cover Change Over Bhiwani District Using Google Earth Engine Suraj Kumar Singh, Shruti Kanga, Bhartendu Sajan, Sayali Madhukarrao Diwate, and Gaurav Tripathi

Abstract Land use and land cover classes were mapped in the Bhiwani district of Haryana, India, from 1991 to 2021. Mapping and monitoring of land use and land cover are critical for government, industry, and human purposes. Landsat imagery was used for image classification due to its high spatial and radiometric resolution, as well as its historical availability beginning in 1972 with higher quality images. In the current study, a random forest classifier was used to classify the feature using Google Earth Engine (GEE). GEE is a cloud computing platform, and its advantages include the ability to preview data without downloading it, being free, and being simple to use. Random forest classification was used in GEE software with Python coding algorithms to categorise the images into five categories: cropland, fallow land, built-up area, waterbody, and open scrub. The results showed that 56.21%, 65.48%, 71.91%, and 74.34% of the land was classified as cropland in 1991, 2001, 2011, and 2021, respectively, followed by scrubland. It can also be concluded that fallow land has decreased over the year as cropping practises have increased. The accuracy assessment is a critical parameter for determining the accuracy of classified satellite S. K. Singh Centre for Sustainable Development, Suresh Gyan Vihar University, Jaipur, Rajasthan, India e-mail: [email protected] S. Kanga (B) School of Environment and Earth Sciences, Department of Geography, Central University of Punjab, Bathinda, India e-mail: [email protected] B. Sajan · S. M. Diwate · G. Tripathi Centre for Climate Change and Water Research, Suresh Gyan Vihar University, Jaipur, Rajasthan, India e-mail: [email protected] S. M. Diwate e-mail: [email protected] G. Tripathi e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_8

161

162

S. K. Singh et al.

images. Based on the findings, the classified maps exhibit an impressive overall accuracy exceeding 85%, demonstrating the reliability of the classification process. In contrast, water resources are depleting, and the Bhiwani district relies solely on the Dohan River to meet irrigation, domestic, industrial, and household demands. The study underscores the significance of addressing the threats posed by urban sprawl and natural phenomena to the natural environment. It highlights the crucial role of land and its services in benefiting humanity, emphasizing the need for continued efforts to protect and enhance its quality. Keywords Land use/Land cover · Landscape change detection · Cloud computing · Google Earth Engine

8.1 Introduction 8.1.1 Background The purposeful and unintended impacts of nature and society on a region lead to changes in land use and land cover, which are then controlled by human activity over time and place (Chamling and Bera 2020). Land Cover measures how much of an area is covered by different land and water types, such as forests, wetlands, agriculture, and other land and water types, while Land Use shows under practise, whether for development, conservation, or a combination of the two. Diverse land cover types can be used or managed in very different ways. The majority of changes in land use and land cover are influenced by population increase, economic expansion, and physical factors such as terrain, slope condition, soil type, and climate (Anees et al. 2020). Using the land has an impact on how people move around, and altering how people use the land has an impact on how the land is utilised. It has a significant impact on the availability of many resources, including plants, soil, and water. Evapotranspiration, groundwater infiltration, and runoff levels are influenced by changes in land use (Banzhaf et al. 2017). Land Use and Land Cover change must be taken into account in future planning for our nation in light of global dynamics and their responses to environmental and socioeconomic influences (Shen et al. 2016; Zhou et al. 2017). Land Use and Land Cover changes have a negative impact on land use, land cover changes, agricultural practises, hazard intensity, and socioeconomic dynamics patterns both globally and locally (Xiong et al. 2017). Having a thorough knowledge of land use and cover as well as the potential for their best use in land resource management and land use planning is essential because they are all relevant to selection, planning, sustainable land resource management, and identifying changes in hydrological processes to meet the rising demands for basic human needs and welfare (Leta et al. 2021). Crop agriculture, urbanisation, and other LULC changes are all included in the change detection process and may all benefit from it (Schürmann et al. 2020). Effective land management and decision-making depend

8 Monitoring Land Use and Land Cover Change Over Bhiwani District …

163

on an understanding of the patterns, changes, and interactions between human activities and natural occurrences in the landscape. Strong and affordable analyses of LULC spatial and temporal change can be provided by both GIS and remote sensing (Ngondo et al. 2021). Recent years have seen the use of remote sensing data to track changes in land use and cover. Remote sensing data typically detects, quantifies, and maps LULC patterns due to its broad and frequently repeated data collection, which makes it ideal for processing and excellent georeferencing (Ramachandra et al. 2012). Land use and land cover change (LUCC) has received a lot of attention in the context of environmental change due to its significance in both global and local environmental change. The spatial pattern of land use, on a variety of scales, from local to regional, reflects human activities including urbanisation processes and development policies (Lv et al. 2018). Landscape pattern measures should be used to assess ecological processes and the effects of major urbanisation. There are several indexes for describing landscape patterns, and these have been very helpful for the structure of the landscape and its temporal/spatial dynamics (Rehana et al. 2021). Integrating data from geographic information systems (GIS) and remote sensing has lately benefited in the characterisation of land use/land cover and the dynamics of landscape pattern (RS). India, which is still developing but has had substantial growth, has seen a noticeable increase in urbanisation across several regions of the nation. India has conducted substantial research on the spatiotemporal patterns of land use and land cover patterns with reference to human activities and the dynamics of change. The land is one of the most important natural resources on the globe, and as a result, it has a big impact on how long life may persist there. The shifting land and cover are equally crucial elements, even though the main issue is global and regional (Hsin et al. 2019). Land use and land cover are severely impacted by the city’s population increase. An increase in the need to feed the expanding population is another factor encouraging the growth of agricultural land (Desta and Fetene 2020). The size of the region has decreased as a result of the conversion of the scrubland to arable land. The growth of the built-up area is caused by an increase in the demand for institutional and residential space for capital-level establishments. Despite greater encroachment by low-income inhabitants and developers, the water body’s size has not considerably changed. Continually reflecting this evolutionary process is the spatial growth pattern (Terefe et al. 2017). In recent decades, there has been a notable trend of increasing urbanization, driven by improved economic opportunities in metropolitan areas (Spruce et al. 2020). This shift has brought about a sense of unpredictability in terms of population patterns, with implications that can occasionally pose risks and hazards. Despite having a substantial negative influence on developing countries like India, urbanisation is still very severe (Bera et al. 2021). Cities must evaluate growth challenges in conjunction with those factors and set development priorities accordingly since the expansion has important implications, including the need for improved infrastructure and more amenities. Landscape systems provide serious problems since they constantly alter the species’ composition, habitats, and driving factors, which in turn affect the different patterns of the landscape. As a result. Patterns of land use and land cover are scale- and time-dependent, making them vulnerable to environmental factors (Kumar

164

S. K. Singh et al.

et al. 2021). In contrast, developed landscapes have abrupt changes in grain size, whereas undeveloped landscapes experience only mild variations. Understanding urban dynamics requires a thorough understanding of patterns and sizes. Due to the ecological, economic, and cultural aspects associated with the land, it also provides an excellent foundation for using various planning and management strategies (Tomar et al. 2021). Seasons and temporal complexity have undergone constant change, as a result of human-caused developments like industrialisation and the copper and iron eras. Patterns of human behaviour in the landscape are used to first identify and describe changes to the landscape (Joy et al. 2021). Depending on the dynamics of man-made, natural, managed, cultivable, suburban, and urban landscapes are grouped into five different major categories: natural, managed, cultivable, suburban, and urban. Change is defined on a temporal and spatial scale by a variety of factors, including the size, shape, number, and origin of patches. Peri-urban and rural environments can coexist within an ecosystem that has its own unique characteristics (Kanga et al. 2021). We can observe how these four significant changes—accessibility, urbanisation, globalisation, and natural disasters—are highlighted in this situation and take place at the same time. The mechanics of these movements have been attributed to social processes, industrial and agricultural policy, and agricultural policy. Monitoring the terrain and assessing how spatial patterns have changed over time is necessary for this dynamic phenomenon (Meraj et al. 2021). Predicting factors that drive change in the environment is vital for sustainability. How changes in the landscape’s spatial pattern and biological processes affect the components of ecosystems, the distribution of biodiversity, the complexity and structure of the landscape, and the importance of culture are caused by these interconnections. As a result, landscape changes require quantification that accounts for alterations in the arrangement and their implications. This study seeks to ascertain the changing pattern of land use and land cover in the Bhiwani district to answer this question better. In this research, remote sensing data were used using image processing software to track changes in land use. The results of this study will aid decision-making for those engaged in sustainable resource management and development.

8.2 Study Area The Bhiwani district is located in south-western Haryana, India, and it is spatially located between 28° 47, 56.5656,, N and 76° 8, 0.6504,, E. The district of Bhiwani has 3432 km2 in total. To the north of the study area lie the districts of Hisar, while the eastern region is bordered by Rohtak. In the southeastern direction, we find the district of Charkhi Dadri (Fig. 8.1). As Choti Kashi, Bhiwani in Haryana is a popular tourist destination because of its religious significance. This region of the Trans-Gangetic Plain experiences cold winters and scorching summers as a result of the seasonal variations. The northern and southern regions of the Bhiwani district have sandy and loamy soil types. The Dohan River is the sole river in the area, and

8 Monitoring Land Use and Land Cover Change Over Bhiwani District …

165

Fig. 8.1 Location of the Bhiwani District of Haryana, India

the groundwater level is low. Small ponds have been erected on either side of the settlement, although they are equally reliant on rainfall. In July and August, there is 483 mm of yearly precipitation. The southwest’s water table is dropping quickly. The average temperature in this region is 25.20 °C, with the minimum temperature in winter being 20 ° and the highest temperature in summer being 450 °. The climate in this area is extremely dry with quite large variations in temperature range.

8.3 Materials and Methods Landsat multispectral images capture a range of electromagnetic spectrum bands, such as visible, reflecting infrared, and thermal infrared light. The onboard sensor on the Landsat satellite platform can capture this multispectral imagery. Understanding sensor strengths and weaknesses is crucial for choosing the proper remotely sensed data for picture classification. The first step in sorting sensor data is choosing the relevant sensor data. The needs of the user, the scope and characteristics of the field of research, the variety of image data, their costs and time constraints, as well as the analyst’s background, must all be taken into account when choosing an image. In this study, the identification of LULC modification was done using satellite images. The basin was captured on camera by Landsat satellites during the years 1991, 2001,

166 Table 8.1 Data used in this study

S. K. Singh et al.

Dataset

Date of acquisition

Spatial resolution (m)

Landsat 5 TM

18th February 1991

30

Landsat 7 ETM+ 06th February 2001

30

Landsat 7 ETM+ 22nd February 2011 30 Landsat 8 OLI

17th February 2021

30

2011, and 2021. This study intended to Landsat satellite data to quantify the change in Land Use Land Cover over the years (Table 8.1). This map was produced using supervised classification, which categorises the entire image using the information on the types of land cover. Only when the raw data has been suitably pre-processed and ready to remove radiometric, atmospheric, and earth geometry problems can it be classified as an image. Finding and repairing broken lines, geometric correction, radiometric calibration, atmospheric adjustment, and topographic correction are all processes in the preparatory process. With this method, we only choose pixels that show recognisable patterns or patterns that can be recognised using data from Google Earth. The data, the classes we want to learn about, and the methodologies to be used must all be understood before we can begin choosing training samples. We could monitor and follow the classification of pixels by tracking them. You must first assume that the results are accurate in order to effectively compare an unclassified image or change detection map. Comparisons between classified and reference maps have been used to assess accuracy while classifying maps. The Overall Accuracy report and the three other reports—individual accuracy, user accuracy, and producer accuracy—used the Kappa coefficient to investigate and ensure the highest level of assessment accuracy. The GEE engine previews all of the datasets in this study, cutting down on labour time and providing the ideal image for the classification process with the aid of codes. For accurate automatic categorization when employing remote sensing techniques, representative ground truth data is necessary. This classification and regression procedure was carried out using the Random Forest classification method after pre-processing the Landsat data. The classification output from Random Forest, a straightforward algorithm structure, depicts the statistical mode of numerous decision trees, resulting in a more reliable than other models available which works on a single classification tree algorithm generated by the model. Accuracy increases with the number of trees added. Detecting changes is a key component of land-use classification. Comparisons of the classification of land cover were used in this study. The areas covered by each category of land cover for each of the periods were compared. For each of the many land cover types, these changes—both favourable and unfavourable—have since taken place. GEE’s subset operation is demonstrated in Figs. 8.2 and 8.3.

8 Monitoring Land Use and Land Cover Change Over Bhiwani District …

167

Fig. 8.2 Subset operation in Google Earth Engine

Fig. 8.3 Code for RF classification

8.4 Results and Discussion The land use and land cover map for 1991 is shown in Fig. 8.4. The majority of the land categories in the district are, according to the image classified using random forest classification, farmland (56%), fallow land (14%), built-up (2.89%), waterbody (0.58%), and scrubland (26.42%). Table 8.2 shows the accuracy for random forest categorization in 1991 by producer and user. The table reveals that agriculture (74.44%) has the lowest user accuracy and waterbody (100%) has the greatest. While

168

S. K. Singh et al.

Srubland had the lowest accuracy (72.22%) and Waterbody had the highest accuracy (97.56%). Figure 8.4 represents the land use and land cover map for the year 1991. The classified image using random forest classification exhibits that the majority of the LULC are cropland (74.34%), fallow land (10.1%), Built-up (9.1%), Waterbody (0.63%), and Scrubland (5.81%) area of the district, respectively. Table 8.3 presents producer and user accuracy for random forest classification in the year 2001. The table shows that waterbody (100%) has the highest user accuracy, and cropland (81.01%) has the lowest. The producers’ accuracy for the Built-up category was found to be the highest at 94.23%, while the Scrubland category exhibited the lowest accuracy at 82.356%. Built-up areas have been accurately classified as such in 94.23% of cases, while 92.45% of the areas designated as such on maps actually are. This suggests that the producer’s accuracy in this category is higher than that of the other classes of

Fig. 8.4 LULC map of the study area in 1991

Table 8.2 Accuracy assessment of LULC map for the year 1991

Classes

Producer accuracy (%)

User accuracy (%)

Built-up area

94.94

92.72

Crop land

90.56

90.56

Fallow land

85.71

90

Scrub land

72.22

97.5

Water body

97.56

100

8 Monitoring Land Use and Land Cover Change Over Bhiwani District … Table 8.3 Accuracy assessment of LULC map for the year 2001

169

Classes

Producer accuracy (%)

User accuracy (%)

Built-up area

94.23

92.45

Crop land

94.11

94.11

Fallow land

90.69

92.85

Scrub land

82.35

95.12

Water body

91.48

100

importance (94.23%). The 261 sample points show a total accuracy of 90.80%, which means that 90.80 times out of 100, a point on the 2001 LULC map (RFC) correlates accurately with what was actually happening in 2001 (Fig. 8.5). The map data adequately portrays what actually happened in the Bhiwani district in 2001, as indicated by the Kappa statistic value of 0.86, which indicates an 86% greater agreement than could have been achieved by chance alone (Table 8.3). Land use and cover are shown for the year 2011 in Fig. 8.6. The ground truth points (GCP) were collected for accuracy assessment as an added measure of validation. The image that was categorised using RF technology reveals that the district’s land is primarily utilised for agriculture (74.9%), followed by fallow land (11.4%), builtup land (11.79%), waterbodies (0.71%), and fallow land (4.17%), in that order. Errors of commission in a specific class are shown by user accuracy. Additionally, a special class of omission errors is shown by the producer accuracy. In other words, the user accuracy demonstrates the percentage of points on the map that actually

Fig. 8.5 LULC map of the study area in 2001

170

S. K. Singh et al.

match the user’s description. As a result, Table 8.4 displays both the producer and user accuracy of MLC categorised images. According to the table, fallow land has the lowest user accuracy—85.1%—and waterbodies have the highest user accuracy (100%) of any category. While Water Body (97.22%) and Srubland (85.41%) had the lowest producer accuracy rates. The user’s accuracy for the Waterbody class (100%) was higher than that of all the other classes in this examination. The 261 sample points display an overall accuracy of 90.98%, meaning that 90.98 out of 100 times, a point on the 2011 LULC map (RFC) corresponds correctly with what was truly on the ground in 2011. The overall accuracy of the 261 sample points is 90.98%, or in other terms, the result exhibits that 90.98 times out of 100, a point on LULC map corresponds with the ground survey correctly what actually happened in 2011. While 94.44% of the built-up areas have been accurately classified as built-up, only 92.44% of the areas marked as built-up on the map actually are built-up, which means that the producer’s accuracy for water bodies (97.22%) is higher than that of

Fig. 8.6 LULC map of the study area in 2011

Table 8.4 Accuracy assessment of LULC map for the year 2011

Classes

Producer accuracy (%)

User accuracy (%)

Built-up area

94.44

92.72

Crop land

87.3

85.93

Fallow land

93.02

85.1

Scrub land

85.41

95.34

Water body

97.22

100

8 Monitoring Land Use and Land Cover Change Over Bhiwani District …

171

the other Land Use Land Cover classes. The assessment’s Kappa statistic value was 0.86, indicating an 86% greater agreement than could have been achieved by chance alone. As a result, the map data adequately depicts the actual situation in the Bhiwani district in 2011. Figure 8.7 presents land use and land cover for the year 2021. The classified image using RF classification shows that most of the land used for Crop Land (74.34%), (12.1%) fallow land, (10.1%) Built-up, (0.63%) Waterbody, and (2.81%) Scrub Land, respectively. Table 8.5 shows that waterbody (100%) has the highest user accuracy, and fallow land (31.66%) has the lowest.The waterbody category demonstrated the highest producers’ accuracy at 97.56%, whereas the Scrubland category exhibited the lowest accuracy at 74.5%. The process of “change detection” demonstrates changes in the amount of land covered between 1991 and 2021 (Fig. 8.8). Through the use of remote sensing imagery that tracks changes in land use and land cover, it has been widely accepted

Fig. 8.7 LULC map of the study area in 2021

Table 8.5 Statistics of the accuracy assessment of LULC map for the year 2021

Classes

Producer accuracy

User accuracy

Built-up area

94.23

89.09

Crop land

89.61

77.52

Fallow land

82.5

31.66

Scrub land

74.5

92.68

Water body

97.95

100

172

S. K. Singh et al.

in both academic research on environmental management and studies of natural resources management. This outcome is explained by a rise in both population growth and the expansion of human wants. Residential areas and agricultural land increased, whereas woodland, bushland, and grassland all decreased. This trend suggests that as populations rise, so does the need for agricultural products. We found a shift in the district’s land use, particularly in terms of settlement and agricultural land. Between 1991 and 2021, the amount of built-up and agricultural land increased, while the amount of pasture, bush, and forest lands dropped (Table 8.6). The district’s extensive reservoir construction may be to blame for the water bodies’ rather constant distribution. The hydrology of the catchment may be impacted by the change in land use and land cover. A large portion of the district’s forestland has been replaced by farms and homes as a result of the district’s growing population.

Fig. 8.8 Temporal variation in the area of various land cover classes

Table 8.6 The calculated area from the year 1991 to 2021 (Km2 ) Classes

Total area in 1991

Total area in 2001

Total area in 2011

Total area in 2021

95.23

146.28

223.89

299.85

1851.63

2156.91

2368.68

2448.89

Fallow land

457.52

368.68

408.59

398.58

Scrub land

870.41

592.57

269.28

125.7

Water body

19.18

29.55

23.56

Built-up area Crop land

20.98

8 Monitoring Land Use and Land Cover Change Over Bhiwani District …

173

8.5 Conclusion and Recommendation In this study, the land use classification of the Bhiwani District over the preceding 40 years was examined. This study found that significant change was observed over the course of its investigation. Both the residential/urban and agricultural sectors showed signs of growth, but scrubland and fallow land showed signs of decline. According to the results of this experiment, converting scrubland and fallow land into agricultural and residential areas may have an impact on a number of the hydrological system of the basin, including streamflow, soil erosion, and groundwater quality. The move from farmland to fallow land, and then from fallow land to urban built-up regions, is the major change in land cover. These changes may have an effect on the local community’s way of life and the management of sustainable resources. By employing better land management practises, better agricultural inputs, integrated watershed management, and a community involvement plan, LULC dynamics in the basin can be avoided. The rapid development of infrastructure, the tourism industry, and population increase in the study area are likely to contribute to a continuation of this trend and the extent of land use and land cover change. GIS and remote sensing can be used to verify seasonal patterns of land use dynamics that can be used by planners and decision-makers, which will aid in the development of sustainable land management planning. The planners and decision-makers for the city master plan, land use management, and natural resource management may also find this study to be helpful. Conflict of Interest: The authors declare no conflict of interest.

References Anees MM, Mann D, Sharma M, Banzhaf E, Joshi PK (2020) Assessment of urban dynamics to understand spatiotemporal differentiation at various scales using remote sensing and geospatial tools. Remote Sens 12(8):1306. https://doi.org/10.3390/rs12081306 Banzhaf E, Kabisch S, Knapp S, Rink D, Wolff M, Kindler A (2017) Integrated research on land-use changes in the face of urban transformations—An analytic framework for further studies. Land Use Policy 60:403–407 Bera A, Taloor AK, Meraj G, Kanga S, Singh SK, Ðurin B, Anand S (2021) Climate vulnerability and economic determinants: linkages and risk reduction in Sagar Island, India; A geospatial approach. Quat Sci Adv 4:100038. https://doi.org/10.1016/j.qsa.2021.100038 Chamling M, Bera B (2020) Spatio-temporal patterns of land use/land cover change in the BhutanBengal foothill region between 1987 and 2019: study towards geospatial applications and policy making. Earth Syst Environ 4:117–130. https://doi.org/10.1007/s41748-020-00150-0 Desta H, Fetene A (2020) Land-use and land-cover change in Lake Ziway watershed of the Ethiopian Central Rift Valley Region and its environmental impacts. Land Use Policy 96.https://doi.org/ 10.1016/j.landusepol.2020.10468 Joy J, Kanga S, Singh SK, Sudhanshu S (2021) Cadastral level Soil and Water conservation Priority Zonation using Geospatial technology. Int J Agric Syst 9(1):10–26

174

S. K. Singh et al.

Kanga S, Rather MA, Farooq M, Singh SK (2021) GIS based forest fire vulnerability assessment and its validation using field and MODIS data: a case study of Bhaderwah Forest Division, Jammu and Kashmir (India). Indian Forester 147(2):120–136 Kumar A, Kanga S, Taloor AK, Singh SK, Ðurin B (2021) Surface runoff estimation of Sind river basin using integrated SCS-CN and GIS techniques. HydroResearch. https://doi.org/10.1016/j. hydres.2021.08.001 Leta MK, Demissie TA, Tränckner J (2021) Modeling and prediction of land use land cover change dynamics based on land change modeler (LCM) in Nashe Watershed, Upper Blue Nile Basin, Ethiopia. Sustainability, 13:3740. https://doi.org/10.3390/su13073740 Lv Z, Liu T, Wan Y, Benediktsson JA, Zhang X (2018) Post-processing approach for refining raw land cover change detection of very high-resolution remote sensing images. Remote Sens 10(3):2–19. https://doi.org/10.3390/rs10030472 Meraj G, Singh SK, Kanga S, Nazrul Islam Md (2021) Modeling on comparison of ecosystem services concepts, tools, methods and their ecological-economic implications: a review. Model Earth Syst Environ.https://doi.org/10.1007/s40808-021-01131-6 Ngondo J, Mango J, Ruiqing L, Nobert J, Dubi A, Heqin C (2021) Land-Use and Land-Cover (LULC) change detection and the implications for coastal water resource management in the Wami-Ruvu Basin, Tanzania. Sustainability 13:4092. https://doi.org/10.3390/su13084092 Ramachandra TV, Bharath HA, Durgappa DS (2012) Insights to urban dynamics through landscape spatial pattern analysis. Int J Appl Earth Obs Geoinf 18:329–343. https://doi.org/10.1016/j.jag. 2012.03.005 Rasool R, Fayaz A, Shafiq M, Singh H, Ahmed P (2021) Land use land cover change in Kashmir Himalaya: Linking remote sensing with an indicator based DPSIR approach. Ecol Indic 125:107447. https://doi.org/10.1016/j.ecolind.2021.107447 Schürmann A, Kleemann J, Fürst C, Teucher M (2020) Assessing the relationship between land tenure issues and land cover changes around the Arabuko Sokoke Forest in Kenya. Land Use Policy, 95.https://doi.org/10.1016/j.landusepol.2020.104625 Shen H, Huang L, Zhang L, Wu P, Zeng C (2016) Long-term and fine-scale satellite monitoring of the urban heat island effect by the fusion of multi-temporal and multi-sensor remote sensed data: a 26-year case study of the city of Wuhan in China. Remote Sens Environ 172:109–125 Spruce J, Bolten J, Mohammed IN, Srinivasan R, Lakshmi V (2020) Mapping land use land cover change in the lower Mekong Basin From 1997 to 2010. J Front Environ Sci 8:21. https://doi. org/10.3389/fenvs.2020.00021 Terefe T, Feyera S, Moges K (2017) The impact of land use/land cover change on ecosystem services in the central highlands of Ethiopia. Ecosyst Serv 23:47–54. https://doi.org/10.1016/j. ecoser.2016.11.010 Tomar JS, Kranjˇci´c N, Ðurin B, Kanga S, Singh SK (2021) Forest fire hazards vulnerability and risk assessment in Sirmaur District Forest of Himachal Pradesh (India): A geospatial approach. ISPRS Int J Geo Inf 10(7):447. https://doi.org/10.3390/ijgi10070447 Tsai YH, Stow D, An L, Chen HL, Lewison R, Shi L (2019). Monitoring land-cover and land-use dynamics in Fanjingshan National Nature Reserve. Applied Geography, 111, 102077. https:// doi.org/10.1016/j.apgeog.2019.102077 Xiong J, Thenkabail PS, Gumma MK, Teluguntla P, Poehnelt J, Congalton RG, Yadav K, Thau D (2017) Automated cropland mapping of continental Africa using Google Earth Engine cloud computing. ISPRS J Photogram Remote Sens 126:225–244.https://doi.org/10.1016/j.isprsjprs. 2017.01.019 Zhou W, Pickett STA, Cadenasso ML (2017) Shifting concepts of urban spatial heterogeneity and their implications for sustainability. Landscape Ecol 32:15–30

Chapter 9

Image and Perception of Royal Heritage and Eco-space of the Medium Towns in India: Reflection from Burdwan Royal Heritage Site Koyel Sarkar and Sanat Kumar Guchhait

Abstract With the rampant and rapid growth of urbanization of the developing countries like India, especially from the last decade of the twentieth century, urban green spaces are rapidly vanishing. With this looming scenario of the defacement of urban green space, the urge for preservation, restoration, and recreation of urban green space has an integral part of urban planning. However, the older royal cities of Delhi, Mysore, Udaipur, Jaipur, Jodhpur, Burdwan, Jhargram, etc. for their heritage appeal are still offering scenic landscapes, deep green coverage interspersed with Royal architecture. Those heritage sites are practically the oasis in the jungle of concrete. Burdwan, a medium-sized town in West Bengal located in the central portion of Bengal is currently experiencing stupendous urban growth. However, the royal heritage site developed by the Maharajas of Burdwan with its appealing deep green eco-space still persists due to the authoritarian outlook of present stakeholders. This inquiry attempts to unfold this reality through qualitative analysis through narratives and word clouds and quantitative measures through factor analysis. The whole inquiry ultimately crops up that heritage site conservation is necessary to arrest the rampant urbanization and also it will help to develop a sensible attitude towards nature which is drastically sandwiched by relentless urbanization. Keywords Royal Heritage · Green space · Designed landscape · Eco-space · Nostalgia · Emotion · Recreation

K. Sarkar (B) · S. K. Guchhait Department of Geography, The University of Burdwan, Purba Bardhaman, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_9

175

176

K. Sarkar and S. K. Guchhait

9.1 Introduction The idea of urban heritage site conservation as the emerging concept and identifying its historical perspective, exploring built-up architecture and eco-space coupled with environmental sustainability has emerged as a new paradigm with its appealing research investigation in the new millennium (Harvey 2001; Stubbs 2004). Historical sites play an important role in maintaining sustainable eco-space as well as the architecture of the past. The treatment of historic gardens by Baker and Chitty (2013), Sharma (2007) has given special attention to Delhi about conservation; keeping it as a historic site even after its transformation. Cloke et al. (2008) showed places having trees can afford emotional responses and can carry significant memories. All the pieces of literature prompt that wild or built-up natural ecology are found to occur in the deserted and or preserved heritage sites of the urban world which are of critical importance not only for urban environmental sustainability, but also for the development of cognitive image and perception coupled with emotional space that is incremental to socio-cultural sustainability (Cabezas et al. 2004; Norton et al. 1997). In this context, cultural awareness appreciates both humans and plants and also focuses on interaction in a truly ecological sense (Gagliano et al. 2015). This idea is exemplified in the foregoing analysis of the royal eco-space of Burdwan town in India developed by the kings of Burdwan and the sustainability of the eco-space maintained by the university authority for their authoritarianism.

9.2 Background of the Study The urban environmental setup, its change, and related issues in Burdwan were discussed by Tah (2016). Mitra (2007) gives an idea about aesthetics as the philosophical study to explain or connect the facts that relate beauty to nature with special reference to its ecological setup, aesthetic beauty, and diversity of species. The historical sites along with the ecological and aesthetic sense holding an important part of the culture of urbanites as well as heritage could be referred to as the built-up ecospace along with royal architecture made by the kings within their kingdom that hold an imprint on the royal heritage even today. Chronologically, Burdwan has been ruled by different rulers from time to time; since the eighteenth century, Burdwan Raj along with other local kingdoms was more a result of Nabawi patronage rather than military adventurism (McLane 2002). The trace of their descent could be dated back to 1657 when Sangam Rai came from Punjab and settled in a village about 8–10 km away from Burdwan. Burdwan Raj Dynasty flourished for about 350 years from 1657 to 1954 both under the Mughals as well as under British rule. Chitra Sen Rai was bestowed with the title of ‘Raja’ in the year 1741 when he was the holder of the Raj Estate (Feilden 2013). The philanthropic built-up nature along with a cluster of several eighteenth century temples enriched the district through their historical, cultural, religious, and ecological importance. Thus,

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

177

the journey of the Burdwan Raj dynasty which started as traders from the ‘Kapoor family’ of Punjab to ‘Zamindars’ or feudal lords and lastly rulers by gaining the title of the ‘King’ made a significant impact on the establishment of Burdwan as a historic town (Sarkar 2014). Among the most important historical places built by the Burdwan Raj Dynasty, the Rajbati was the residential place of Burdwan Maharajas (Konar 2000). In the year 1840, Mahatab Manzil was established by Mahatabchand which was the royal palace for the royal family that has now been converted into an administrative and official building of the University of Burdwan; Golapbag campus which is renowned as an educational hub since 1960; which was earlier the recreational garden as well as zoo established by the King Tejchand from 1809 onwards (Sarkar 2014); Krishna Sayar was built-up by King Krishnaram Rai in 1691 to overcome the hardships at times during famine and also for the benefit of its subjects subsequently changed into a recreational park with time. Dalton Hooker visited Golapbag in the year 1884 and declared it a botanical as well as zoological park enlisting 128 species of plant (Ganguly and Mukherjee 2016). Ramna Bagan was a natural forest patch during the time of the Raj dynasty rule and was the property belonging to them during is now under the control of the divisional forest officer of Burdwan. At present, it has established itself as a zoological park. Therefore, the built-up nature of royal sites indicates their place of signature and is the source of image, emotion, and recreation at present. People of different castes, creeds, religions, and ages have an image and emotions for such an eco-space, and therefore, vehemently support the preservation and maintenance of those sites satisfactorily done by the university authority and the forest department. The major objective of this study is to find out the dominant factors controlling the functioning of royal sites at present based on their importance and performance; to identify the important attributes associated with the importance and performance of these sites based on their functioning through perception analysis and lastly to find out the diversity of the species at present. Thus, it tries to explore the legacy of an ecological image based on its present functioning Ganguly and Mukherjee (2016) that highlighted the aesthetic rejuvenation of the tree species once planted by the royal dynasty within the Golapbag campus to know their practical value in the design of the landscape. The contribution of ruling families to the development of Burdwan was put forward by Dhillon (2014). The Royal Palace and the gardens in Golapbag and Tarabagh, Ramna Bagan, age-old temples and mosques, palatial buildings, trading centres, and the peripheral rural settings have proved to be an optimum matrix presenting the ‘unity in cultural diversity’ hand in hand with plant diversity of esteem’ as stated by Dey et al. (2017).

178

K. Sarkar and S. K. Guchhait

9.3 Study Area and its Rationale Burdwan is well-known for its historical importance. Numerous historical sites, mosques, tombs, and temples predominate in this region. The Rajbati, Hawa Mahal, Dilkhusa Palace (Golapbag Campus), Curzon Gate including Bijoy Bahar, and other historical sites of the Burdwan Raj dynasty are among them. The Royal sites have had an impact on the town’s culture and history since ancient times, but they have also served as justification for the creation of contemporary educational institutions and ecological sites with various types of social value. Burying this royal palace and its related architecture, the royal heritage sites are beautifully landscaped by numerous large water bodies (Sayar or large pond) surrounded by deep green cover mostly of exotic species, a small but rich forest area (Ramna Bagan), and the recreational garden of king marked with island garden (Hawa Mahal), moat, zoo and tree line. The whole area with wilderness and tranquillity not only possesses scenic beauty, but also offers recreation and healthy life (walkable paths are used by the urbanites in the morning and evening). Therefore, major the administrative area and the surrounding region are taken for the study (Fig. 9.1). On the basis of Geo-historical and ecological importance of the Royal sites at present, the selected sites such as Golapbag Campus, Rajbati, Ramna Bagan, and Krishna Sayar Park have been considered for study (Fig. 9.2). The Golapbag area at present is often designated as the “Lungs of Burdwan” due to its ecological richness that can be witnessed in this area even today. The following ward numbers 2, 20, 21, 26, 27, 28 and 30 under Burdwan Municipality are selected for the study.

9.4 Materials and Methods 9.4.1 Questionnaire Design and Sampling To find out the ecological essence of Burdwan Raj Royal Sites along with its historic importance, questionnaires related to different ecological parameters and historical importance were given to the respondents, and were asked to rank based on their level of perception. The data sources were various target groups and data was collected through Questionnaire surveys, in-depth interviews, observation, and participant observation. Questionnaires were based on a Likert scale (5-point grade), open-ended questions, interviews, and recording of an individual’s perceptions, etc. Among the respondents, 50% were attached to the Royal sites as a working place and another 50% were the respondents who were residents living nearby to the Royal sites which included housewives, businessmen, others such as shopkeepers, music teachers, fishermen, etc. not associated with the Royal sites. Questionnaire surveys of elderly persons above 60 years of age were also conducted. The other sources of databases include

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

179

Fig. 9.1 Location of the study area

different journals, books, newspapers, etc. Forest officers of Burdwan zoological park, officers attached to Krishnasayar Park, and Rajbati officers were surveyed.

180

K. Sarkar and S. K. Guchhait

Fig. 9.2 Location of the major study points

9.4.2 Socioeconomic Profile of the Respondents The target group is of critical importance. Here the minimum age is not less than 15 years. Respondents are of different socio-economic backgrounds with familiarity with this area in the context of historical perspective and like to visit the area whenever possible. It also considers persons attached to the royal sites including—University

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

181

teachers, students, scholars, officers attached to the royal sites, etc. Respondents who were not attached to the royal sites included residents of the surrounding area residing at least fifteen years in this town through which they acquired a level of cognition about the royal architecture and eco-space of this area.

9.4.3 Methodology Adopted The study incorporates qualitative and quantitative analyses to make a common convergence of image and perception. Qualitative analysis is enriched by narrative analysis and word cloud thereafter that ascertain the core area of emotion and perception subjectively. This subjective analysis is unable to justify the objective reality. Therefore, to confirm the subjective outcomes, Exploratory Factor Analysis (EFA) is employed. This is done by identifying the important variables which can justify the image and perception concerning royal heritage and eco-space of this area. Narratives are an important source for identifying the variables for EFA. Here, the responses of the study area are coded with a 5-point Likert scale ranging from 1 to 5. To glean out the image and perception, altogether 30 variables are finally taken for which respondents gave priority. These are: Safeguarding the environment (i), Aesthetic Beauty Decreasing (ii), Evidence of conservation (iii), Rich Biodiversity (iv), Animal Diversity Decreasing (v), Conservation of species (vi), Contribution of Burdwan Raj Dynasty (vii), Ecological Balance Stable (viii), Plant Diversity (ix), Ecosystem Harmony (x), Historical Importance (xi), Ecological Balance (xii) Ecological Stability (xiii), Animal Diversity (xiv), Noise Free Zone Status (xv), Aquatic life in the environment (xvi), Animal Diversity Increasing (xvii), Wetland (xviii), Plastic free zone status (xix), Food Items Restrictions (xx), Recreation Facilities Increasing (xxi), Migratory birds Increasing (xxii), Awareness Development (xxiii), Aesthetic Beauty Increasing (xxiv), Plastic free zone would be (xxv), Recreation Facilities Decreasing (xxvi), Cleanliness (xxvii), Pollution free environment (xxviii), Environmental Programmes (xxix), and Migratory bird decreasing (xxx). All the variables are rated on a five points Likert scale taking the responses of urbanites within the study area. The dominant dimensions are sorted out to get the image and perception of royal architecture and eco-space. The respondents were selected through purposive sampling considering the occupational structure, educational level, and attachment to royal sites as a criterion for selection. Finally, 212 respondents were extracted from the study area from a large sample size through direct open-ended interviews, semi-structured interviews, questionnaire surveys, and participant observation until theoretical saturation is gained. Thus, although both Quantitative and Qualitative methods are used in the process of data collection, the Triangulation approach is used by analysing three quantitative statistical techniques to get an appropriate result: (i) Firstly, Factor analysis has been done to identify the dominant factors influencing the royal heritage site and ecospace based on performance first and importance thereafter. A sample size of 212

182

K. Sarkar and S. K. Guchhait

was considered for the study where the minimum requirement for such analysis is 150 (5 times the variables, i.e., 30 * 5 = 150). The respondents were the local urbanities selected based on both purposive as well as convenience sampling. Figures and tables are done by the author. Therefore, their sources are from primary survey and fieldwork.

9.5 Results and Discussions 9.5.1 Narrative Analysis Narrative analysis is adopted here to spot the core area of the image and perception of the urbanites through their instant spontaneous reaction about this place, its history, or their feeling whenever he or she comes here and about the preservation and sustainability of this site. Out of 212 spontaneous responses, 36 are presented here in which image and perception are deeply embedded (Table 9.1). Those statements are analysed through NVivo-12 software, which pictorially and geometrically (more frequent; large in size) depict the hierarchical level of perception. The word cloud profoundly expresses the image and perception of the royal architectural heritage and eco-space (Fig. 9.3). Feel is the most vibrant cognition. Other important attributes related to image and perceptions are visited, whenever, place, sites, happy, relaxed, nostalgic, memory, historical, etc. Place and sites stand for eco-space and royal architecture. The historical perspective is expressed by nostalgia, royalty, and emotion, whereas recreation and relaxation relate to the designed landscape and eco-space. Word cloud gleaned through narratives, therefore, encircles three important dimensions—historical legacy, royal heritage, and designed eco-space.

9.5.2 Exploratory Factor Analysis Based on that perception of narrative analysis, several questions (41) are framed with a rating scale where the respondents will answer the questions and the appropriate value is assigned. In finding out the correlation among the 41 variables, some variables are rejected for their low response throughout the correlation matrix. Therefore, altogether 30 variables are taken for exploratory factor analysis.

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

183

Table 9.1 Narratives of respondents R1

I feel nostalgic when I visit those sites and whenever get free time, I go out with my family in those areas covered with green trees. It makes me feel happy and relaxed at the same time

R2

Having those places for recreation and living in Burdwan is a matter of pride for us, it is our identity

R3

Positive vibes come into my mind I never I visit them

R4

I like the place a lot for recreation. Actually, I feel very happy, energetic, and relaxed when I visit those site

R5

A situation like that makes me forget my problems of daily life and think of things that are outside of my frequently problematic life

R6

I feel sad and nostalgic

R7

No more the legacy is present, no more historical flavour, too much administrative work, not well maintained few parts, and not cleaned properly…

R8

I feel deep emotion with nostalgia when I take a look at this place

R9

My heart is filled with joy whenever I come here

R10 I miss those days as I am unable to go outside my house due to age and distance R11 The city retains its identity, along with its historical significance, to the point where we are really very proud whenever we visit it R12 An enormous number of nostalgic memories are associated with this area of the city R13 I feel happy and proud R14 It is a place for recreation and happiness R15 I feel an emotional attachment R16 I miss those days as I am unable to go outside my house R17 Very happy and at the same time it is a matter of pride for us as it is a royal place R18 Positive Vibes come into my mind. I visit them and I feel happy and relaxed R19 I feel connected to this place mentally for its wilderness and greenery R20 I forget my worries R21 The age-old trees have a lot more to tell R22 People of Burdwan still feel that the image of the area is developed by the king… R23 Nostalgic, happy, proud… R24 Happy and emotional… R25 Positive vibes come into my mind whenever I visit them R26 I feel nostalgic when I visit those sites and whenever get free time I go out with my family in those areas. It makes me feel happy and relaxed at the same time R27 Places make me emotional and nostalgic at the same moment. I feel very happy whenever I visit R28 Now due to age and distance, I am unable to go outside my house, but I miss those days of my life R29 Whenever I visit these sites, I feel charmed (continued)

184

K. Sarkar and S. K. Guchhait

Table 9.1 (continued) R30 I feel very sad whenever I come to this site as it reminds me of my childhood days, but at the same time, I feel happy that it is being maintained by the university authority R31 I feel very happy and relaxed whenever I come here R32 I have a deep attachment with this place and, at the same time, I feel very proud seeing all of those

Fig. 9.3 Word cloud

9.5.3 Factor Loadings and Reliability Statistics To identify the important factors related to the Raj royal heritage site and ecospace based on their performance and importance, exploratory factor analysis was conducted through Varimax rotation using Kaiser Normalization. Cronbach’s alpha was calculated to find out the reliability of the data and internal consistency which in both cases showed high reliability among the variables (Table 9.2). Three components are extracted through dimension reduction and the variance explained by each factor is 53.9%, 32.81%, and 11.28%, respectively, of the total variance explained, i.e., 100% indicating an error in the system. Cronbach’s Alpha is used to measure high internal consistency or reliability among the data; here the value of 0.90 indicates the high reliability of the test items. As

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

185

Table 9.2 Reliability statistics Importance of attributes

Performance of attributes Cronbach’s Alpha

0.90

Cronbach’s Alpha

0.89

Cronbach’s Alpha based on standardized items

0.90

Cronbach’s Alpha based on standardized items

0.92

N of items

30

N of items

30

Table 9.3 Eigen values and variance explained

Component

Initial eigen values Total

Percentage of variance

Cumulative percentage

1

18.594

61.979

61.979

2

7.554

25.181

87.160

3

2.784

9.279

96.439

4

1.086

3.561

100.000

illustrated in Table 9.3 out of the variables that contribute to the performance or satisfaction from Burdwan Raj Royal sites the very first component is aligned with attributes such as conservation of species, ecological balance, maintenance, species diversity, safeguarding nature, etc. In this case, both positive as well as reverse score items are considered. The Factor Analysis of Attributes based on their level of Importance can be divided into four major components that explain 61.979%, 25.181%, 9.279%, and 3.561%, respectively, out of the total variance. The total number of variables is 30 as illustrated in Table 9.3. Out of thirty variables four are most important that contribute higher loadings through the first component: those are cleanliness, environmental programmes, and increase in aesthetic beauty and pollution-free environment having high loadings above 0.9 that is due to the management strategy of the present stake holders (University Authority and Forest Department). The deep green environment and cleanliness of the sites along with norms such as a plastic-free environment, and the prohibition on smoking and food items within the Ramna Bagan area (presently Burdwan Zoological Park) maintained by the forest department is enduring for ecological balance and species biodiversity within this area. The priority is given to the plants and animals here and their maintenance. No heavy or light vehicles are allowed here and the permissible sound level through horns from vehicles reduces noise pollution. Picnics are not allowed within the area and even screaming, shouting, or disturbing the animals is restricted here. The highest factor loading value of cleanliness that is 0.97 is not only due to the maintenance of this area, but also because the respective authorities are taking continuous measures and initiatives to increase the aesthetic beauty of these sites. The other attributes in this component are the prospect of a plastic-free zone, awareness development,

186

K. Sarkar and S. K. Guchhait

safeguarding environment, conservation, and stability of ecological balance with factor loading values more than or equal to 0.80. Increase in species diversity and migratory birds have factor loadings above 0.7. Similarly, the loading value of 0.68 is for the present status of the plastic-free zone which is controlled at present to some extent and the urbanities have an urge for a more plastic-free zone in the future. Thus, the very first component can be named Sustainability and Safeguarding of the Environment. This has also been the reason why the only green space within the town could be seen here. The plant diversity within these royal sites especially Ramna Bagan and Golapbag Campus has formed a mini hotspot over here. The second component named Green space quality and biota and Historical legacy has both positive and negative loadings. The question was asked whether green space quality, animal diversity, and recreational facilities are decreasing. The respondents go against it; therefore, the loading is highly negative (recreation facilities-0.926, decrease in animal diversity-0.921, decrease in migratory birds-0.894, decrease in aesthetic beauty-0.877, and ecological instability-0.857). It is really a matter of concern, though responses are the normal trend regarding contemporary environmental issues. The plant diversity has high loading positive loadings (0.883) along with the richness of biodiversity (0.837). Other important categories in this component include Historical Importance with a loading value of 0.807 and Contribution of the Burdwan Raj Dynasty with a loading value of 0.719 (Table 9.4). This shows that the legacy of the Burdwan Raj Dynasty is still present in the minds of the residents through their philanthropic works as well as their contribution to the development of Burdwan which is properly maintained by the University Authority and Forest Department. Thus, all these attributes highlight the importance of Burdwan Raj even today. The third component states Aesthetic Beauty and Recreation. The Ramna Bagan at present taken care of by the Forest Authority—the headquarter of the district forest department—has changed the reserved forest into a deer park and recently into the zoological park. The whole area of Ramna Bagan is modified to form Burdwan zoological park with new cages and animals along with new amenities without harming the ecology of this area. It was declared a Wildlife Sanctuary in 1981 and at present, the wildlife within the Ramna Bagan with an area of 14.31 hectares includes Emus, Crocodile, Bears, Vulture, Porcupine, etc. apart from spotted deer. A large variety of local, indigenous, and exotic trees species such as Sal (Shorea Robusta), teak (Tectona Grandis), Red Cotton tree (Bombox Ceiba), Ankola (Alangium Lamarckii), Bhringraj (Eclipta Alba), Rain tree (Samanea Saman), Malabar Palm (Syzygium Cumini), etc. are found within the forest area, Golapbag campus and Krishna Sayer that act as a source of recreation from nature (Bhattacharya 2009). The wetland also has a high loading value of 0.787 along with the increase in natural recreation facilities with a loading of 0.982 (Table 9.4). Here, recreation from nature is increasing day by day rather than artificial means of recreation. The zoo acts mostly open space for wild animals than restricting animal rights. Here the animals and their habitat are given more priority. The created wetlands by the kings of Burdwan such as Krishna Sayar, Rani Sayar, Kamal Sayar, and Shyam Sayar act as large water reservoirs for Burdwan. The aesthetic beauty of

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

187

Table 9.4 Factor analysis of attributes based on their level of importance Rotated component matrix

Component 1

2

3

4

Cleanliness

0.97

0.157

0.182

0.032

Environmental programmes

0.927

0.358

0.11

0.003

Aesthetic beauty increasing

0.909

0.338

0.105

0.221

Pollution free environment

0.903

0.258

0.339

0.05

Plastic free zone would be

0.88

0.39

Awareness development

0.858

Safeguarding the environment Evidence of conservation

−0.271

0.033

0.347

0.243

0.291

0.85

0.489

0.068

0.184

0.806

0.522

0.127

0.25

Ecological balance stable

0.803

0.546

0.053

0.234

Migratory birds increasing

0.774

−0.015

0.623

0.117

Conservation of species

0.714

0.627

0.006

0.312

Plastic free zone status

0.68

0.082

0.603

0.41

0.076

−0.926

0.14

Animal diversity decreasing

−0.388

−0.921

−0.029

−0.005

Migratory birds decreasing

−0.27

−0.894

−0.316

−0.166

Recreation facilities decreasing

0.342

0.249

0.883

−0.396

0.037

Aesthetic beauty decreasing

−0.249

−0.877

−0.392

−0.126

Ecological balance unstable

−0.205

−0.857

−0.467

−0.069

Rich biodiversity

0.445

0.837

−0.301

−0.104

Historical importance

0.469

0.807

−0.353

−0.065

Initiatives and maintenance

0.65

0.755

−0.032

−0.08

Contribution of Burdwan Raj Dynasty

0.682

0.719

−0.094

0.094

Ecosystem harmony

0.66

0.711

−0.119

0.211

Plant diversity

Recreation facilities increasing

−0.007

−0.058

0.982

0.177

Animal diversity

−0.138

−0.065

0.973

0.172

Wetland

0.588

0.023

0.787

0.185

Animal diversity increasing

0.617

0.07

0.754

0.217

Food items restrictions

0.265

−0.09

0.5

0.819

Noise free zone status

0.124

0.019

0.68

0.722

Aquatic life in the environment

0.494

0.027

0.519

0.697

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization a. Rotation converged in 11 iterations

188

K. Sarkar and S. K. Guchhait

these wetlands along with a large number of species diversity both terrestrial and aquatic provide a suitable habitat for migratory birds (Siberian birds), especially in Krishna Sayar which makes the area more appealing during winter. The last component is based on Eco-Authoritarianism where the ecological balance of this area is maintained under the prudence of the present authority and no it is allowed for degradation in the sequel to popular development. The prohibition of food items within the zoological park and the attempt of making it a plastic-free and no-smoking zone could be seen. Several awareness programmes are raised every year to increase human awareness and conserve these royal sites with an ecological signature. The urbanities are seen to support these measures taken by the present stakeholders. The noise pollution is also controlled within this area for restricted entry of vehicle and therefore pollution level is low showing high loading values of 0.819 and 0.722 (Table 9.4). The perception regarding the increase in aquatic life has no significant loading because apart from Krishna Sayar which is under Acchi Parishad, Burdwan University other wetlands are used for multiple purposes related to domestic chores and they are facing degradation as the authoritarian control is not present there. Therefore, authoritarian legacy is the most important consideration for the conservation and maintenance of royal architecture and also the preservation, conservation, and maintenance of eco-space. The investigation depicts that this area is not unique compared to other parts of the city, but its historicity and ecological setup have a huge impact on social and psychological domains. Narrative analysis confirms that respondents have deep emotion and an urge to access the area, even regularly for the morning and evening walkers. Even under momentary urbanization at present, cities that have such a setup should come under authoritarian principles through which conservation, preservation, and maintenance should be possible. EFA ascertains that four dominant factors influencing the Burdwan Raj Sites based on their level of importance in the present scenario are Safeguarding the Environment, Historical legacy and Importance, Aesthetic Beauty and Recreation, and Eco-Authoritarianism. Based on performance, three important factors are Ecological Stability and Maintenance of Burdwan Raj Sites, Environmental Ethics and Recreation, and lastly Awareness among respondents. All these components play a dominant role in the functioning of Burdwan Raj Sites and maintain their popularity even today.

9.6 Conclusion and Recommendations Heritage Geography has opened up new horizons with a huge boundary. Urban heritage sites offer a delicate inquiry into their historical event(s) and environmental perspective. Urban heritage site conservation encapsulates the history, culture, and ecological transformation and conservation which connotes topophilia relating to cultural dimension and ecological past. In the present context, architectural heritage

9 Image and Perception of Royal Heritage and Eco-space of the Medium …

189

surrounded by built-up green space represents its vibrant historical image and ecological persistence and also imprints. Urbanites of this area give almost equal importance to the ecological image and royal heritage and therefore, environmental importance are more profound which has become possible due to the authoritarian principle of present stake holder (University authority and Forest Department) for their endless effort of preservation and conservation of heritage sites. The main stakeholder of this area i.e., the Burdwan University Authority is taking proper attention, management, and strategies for conservation as well as the development of the wetland ecosystem as well as green space. Another reason is that this eco-space face human interaction during peak periods (10 am to 5 pm) with a regulated behaviour, and in the off-peak period (5 pm to 10 am), it almost becomes a shadow zone. Therefore, wild creatures have room for a peaceful existence. Similarly, the cultural landscape is guided by educational functions where rules and regulations are so strict that apart from an educational perspective no other performing activities like trade, manufacturing, or market operations are not allowed. Burdwan Raj royal sites despite some threats and weaknesses have evolved into a new physical and social space where the positive externalities are getting more priority than the negative ones forming a distinct position in the minds of the urbanities even today. This inquiry clearly prompts that for future green space development in the city space, urban green corridor development—a popular notion of eco-urbanism—can persist not only by the maintenance and preservation, but also by the people’s urge to appreciate the preventive and protective measures through authoritarian principles.

References Baker D, Chitty G (2013) Managing historic sites and buildings: reconciling presentation and preservation. Routledge Bhattacharya A (2009) Ecofloristic studies on forests in Bardhaman District. The University of Burdwan, West Bengal Cabezas et al (2004) Sustainability: ecological, social, economic, technological, and systems perspectives. In: Technological choices for sustainability. Springer, pp 37–64 Cloke et al (2008) Memorial trees and treescape memories. Environ Plan: Soc Space 26(1):107–122 Dey et al (2017) A contribution to the study of mural flora of Burdwan in West Bengal state of India. Indian J Sci Res 13(1):151–155 Dhillon A (2014) Historical account of Burdwan Raj—a tribute to Mahtab family of Burdwan. Indo-Canadian Friendship Society Feilden C (2013) The Indian Empire Royal Book: containing a true account of the commercial relations of Great Britain and the Indian Empire. Michigan State University, BPBC Gagliano et al (2015) Breaking the silence—language and the making of meaning in plants. Ecopsychology 7(3):145–152 Ganguly S, Mukherjee A (2016) A census of the tree species in the Golapbag campus of Burdwan University, West Bengal (India) with their IUCN Red List status and carbon sequestration potential of some selected species. Indian J Sci Res 7(1):67 Harvey D (2001) Heritage pasts and heritage presents: temporality, meaning and the scope of heritage studies. Int J Herit Stud 7(4):319–338 Konar G (2000) Bardhaman Samagraha, vol 1. 37A, College street, Kolkata-73, Mitram

190

K. Sarkar and S. K. Guchhait

McLane (ed) (2002) Cambridge University Press Mitra A (2007) A dissertation on aesthetics: Anandavardhana and Kant. The University of Burdwan, Burdwan Norton et al (1997) Sustainability: ecological and economic perspectives. Land Econ 553–568 Sarkar (2014) Bardhaman Raj Itibritta, vol 1. Bardhaman Sharma JP (2007) The British treatment of historic gardens in the Indian subcontinent: the transformation of Delhi’s Nawab Safdarjung’s Tomb complex from a Funerary garden into a public park. Garden Hist 210–228 Stubbs (2004) Heritage-sustainability: developing a methodology for the sustainable appraisal of the historic environment. Plan Pract Res 19(3):285–305 Tah (2016) Urbanization and environmental issues: a comparative study of Barddhaman and Durgapur Municipal areas, West Bengal. The University of Burdwan, West Bengal. http://hdl. handle.net/10603/204521

Chapter 10

Governance and Floodplain Extent Changes of Yamuna River Floodplain in Megacity Delhi Shobhika Bhadu and Milap Punia

Abstract Rivers across the globe are the most prosperous regions and loci of development of civilisations, e.g., New York on the Hudson, Paris on the Rhine, and Delhi on the bank of Yamuna River are some of the important million-plus cities located along the bank of rivers. Urban transformations of the 21st-century along the river bank create ecological problems in the natural riparian system. Delhi, a megacity, presents a complex case being the national capital of a developing nation with an ever-increasing population. In contrast, the planning of megacities is mainly rooted in the theories of the functioning of cities in the developed world. Delhi’s urban plan has divided Delhi into various planning zones, and the ecologically sensitive Yamuna River floodplain has been designated as Zone ‘O’. A land-hungry city and the artificial structures in the river’s floodplain obstruct, encroach and divert flow and reduce its natural evolution and extent. Such human interventions through casualness in planning for urban floodplains have necessitated the renewal of our understanding regarding the impact of recent urban transformations in the riparian environment and vice-versa and the role of governance. Remote sensing has been used to analyse these changes in the riverbed lying in Delhi and for understanding the planning and governing aspects of zone ‘O’. Keywords Yamuna river · Floodplain extent · Megacity Delhi · Governance · Floodplain zone ‘O’

10.1 Introduction In this age, characterised by geologists as the Anthropocene, patterns of human settlements are the most significant influencing forces on the environment and climate. The major portion of human settlements and infrastructure mainly occupies the urban sites S. Bhadu (B) · M. Punia CSRD, School of Social Sciences, Jawaharlal Nehru University, New Delhi 110067, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_10

191

192

S. Bhadu and M. Punia

across the globe. Urban centres of the developing world are overwhelmed as urban planning policies are not at the pace with the rate of migration. The urban population in megacities has far exceeded the carrying capacity of the designed infrastructure. It creates the problem of environmental sustainability and issues which are challenging urban governance. Therefore, the need of hour is to preserve ecosystem services by integrating the Ecological Urbanism approach in our governance structures, which emphasises nature during urban planning and provides solutions to the multiple social, economic and environmental challenges facing the 21st Century cities. In Delhi, the Yamuna River is reduced to a poorly managed resource, absent from the urban landscape and urban imagination due to the nonexistence of the ecological urbanism approach in policy formulations regarding urban governance (Manna 2019). Now, let’s look at the history of the urban growth of Delhi. We can see from the census results that since 1951, Delhi’s population has grown from 1.74 million to 16.75 million in 2011 as Delhi is the centre of art, culture, markets, Industries, commercial and administrative head (GNCT of Delhi 2013). Gurgaon and Noida are Information Technology and business centres, respectively. With the growth of population, the urbanisation of Delhi has extended into the adjacent states of Haryana and Uttar Pradesh. Because of such high population growth, the infrastructure of Delhi is working beyond its limits, and it is facing problems such as air pollution, water pollution, soil pollution, space acuteness, and social issues arising because of population pressure. One of the major problems is the occupation of floodplain and the construction of permanent structures, which may lead to significant flood damage in the future, and ecological problems like loss of aquatic biodiversity and decline of groundwater level because of surface cover changes over groundwater recharging zone. The biggest problem of third-world countries is their large population size (Pearce et al. 2013). Lack of employment opportunities and low productivity in agriculture in rural India, and attraction for higher wages and good services in urban centres led to significant streams of rural–urban migrations. But a city has a threshold of carrying capacity for a fixed number of people beyond which it cannot accommodate more of them. So, in Delhi, the initial residents of the floodplain were the people who could not afford decent houses. Yamuna Pushta is an example that was demolished in 2005 in the name of good governance (Baviskar 2006). After the 1990s economic liberalisation, the governance system has become favourable towards the free market concept (Kohli 2006). But in recent years, the urban planning agency Delhi Development Authority (DDA) has been facing the space crunch for providing housing facilities for a continuously growing population. For any system to develop, grow and thrive, the concept of governance is essential. Governance in any system means establishing policies and then regular monitoring of the proper implementation of those policies by the member of the governing body of an organisation of that system. “Governance has become a concept that includes more and more phenomena related to the steering of societal developments”. According to United Nations Human Settlements Programme (UN Habitat 2002), Urban governance is the sum of the ways in which individuals and institutions, public

10 Governance and Floodplain Extent Changes of Yamuna River …

193

and private, plan and manage everyday affairs and resources of the city (DFID 2002). It is a continuing process that can accommodate conflicting or diverse interests, and corporative action can be taken. It includes formal institutions and informal arrangements and the social capital of citizens. It is vital to understand urban environmental changes by implementing urban policies from the perspective of geography especially when these implementations change the map’s content of the respective urban area. Being a developing country, the high growth rate of urbanisation in India poses enormous challenges for urban governance. The government of India has announced the program of making 100 smart cities in India (Praharaj et al. 2018). The main goal of making the city a smart city is to rejuvenate the urban system, improve urban infrastructure quality of life and achieve sustainable and inclusive development besides making them the magnetic centre to attract foreign direct investment. Urban planning in India has been very complex and faces the lack of planned development most of the time. The rate of urbanisation in India is 2.47%. As per Mckinsey Global (2010), by 2025, Mumbai and Delhi will be number two and three in the world ranking, with an expected population of 26.4 and 22.5 million, respectively (Bholey 2016). Ahluwalia et al. have observed the consequences of unplanned urban development, which are highly unsatisfactory concerning building and maintaining urban infrastructure and delivery of public service (Ahluwalia et al. 2017).

10.2 Statement of Problem Human interventions in the fluvial system are creating problems of ecological sustainability, especially for urban areas located next to the river course. Urban planning in India is a complex process. Cities face poorly planned development most of the time. Delhi, located next to Yamuna Riverbank, is the national capital and has unique challenges most acute of them is land scarcity (Paul et al. 2021). Land hungry megacity needs land to absorb the migrant population and organise international level events, such as Commonwealth Games. Therefore, this study focuses on measuring the extent of human interventions in the Yamuna River floodplain from 1893 to 2015 and developing an understanding of urban policymaking and its effects on the river environment in Delhi.

10.3 Study Area River Yamuna is a Himalayan River that originates in the Yamunotri glacier near Bander Poonch peaks (38°59, N, 78°27, E) in the Mussoorie range of the lower Himalayas at an elevation of about 6387 m above mean sea level in district Uttarkashi, Uttarakhand.

194

S. Bhadu and M. Punia

The Yamuna River stretch is around 23 km from the Wazirabad barrage upstream to the Okhla barrage downstream in the national capital region (NCR). The riverbed is 2–3 km wide, whereas the waterway in this stretch confines up to 450–800 m during the non-monsoon period as shown in the Fig. 10.1. The nature of rainfall is primarily uncertain in India, and also it is unevenly distributed. This erratic pattern causes heavy rain leading to maximum damage in flash floods and landslides. Extensive knowledge of rainfall patterns is essential in the planning for water resource development. Moreover, another preventive way to stay protected against such unpredictable rainfall frequency and intensity and flash flood afterwards are to keep the floodplain free from permanent constructions. Delhi lies in a semi-arid climatic regime facing a high-temperature and rainfall range between winter and summer. Yamuna River in Delhi passes through the eastern part of the city, surrounded by high-density residential and commercial areas. The most suitable and least harmful land use activity in the riverbed is agriculture, carried out on the Yamuna river floodplain. The other activities like construction of the permanent structures and power stations, fly ash dumping, cremation ground, bathing ghats, sanitary landfills, waterworks, and sewage treatment plants cause deterioration and environmental damages to the riverbed.

Fig. 10.1 Study area location; Source Survey of India and DDA

10 Governance and Floodplain Extent Changes of Yamuna River …

195

10.4 Relevance of Floodplain in Megacity of Delhi The floodplains have ecological importance in urban settings, especially in the capital city of India. It is crucial for a megacity like Delhi in terms of hydrology, ecology, and recreational point of view as floodplains are environmentally sensitive and ecologically productive areas and hydrologically vital wetland regions. “The floodplain enhances the resilience of the river basin against climate and anthropogenic change and improves water quality, increases flood safety and ecosystem services for society” (Kiedrzy´nska et al. 2015). Delhi suffers from water scarcity, lousy water quality and waterlogging during the rainy season. Extreme weather events due to shifting climate are increasing in frequency and magnitude all over the globe (Field et al. 2012). Additionally, floodplains are important sites for groundwater recharge in general and particularly in urban areas because they offer a porous surface. In contrast, most of the urban areas are concretised and hence impervious. Moreover, the whole riverine system is an essential habitat for many flora and fauna. Naturally flowing floodplain rivers are among the most dynamic ecosystems on earth (Kingsford 2000). The flow regime of a river system and its connections to floodplain wetlands govern biotic responses, channel formation, and sediment transfer. Also, floodplains provide recreational sites and scenic value for the community. Suppose the Yamuna floodplain in Delhi keeps on shrinking; in that case, water logging problems will aggravate, water quality will worsen, and water scarcity will increase. Because extra floodwater would not get space to spread, waterlogging will increase in the sewage system of areas, especially around Yamuna Floodplain. If the river gets less room to spread its excess water, the groundwater quantity will reduce, and the quality will worsen (Kingsford 2000).

10.5 Literature Review Although numerous definitions exist, a floodplain can be broadly defined as a landscape feature periodically inundated by water from an adjacent river (Opperman et al. 2010b). Floodplain ecosystems are now proven to support biodiversity and primary productivity more than purely terrestrial or aquatic ecosystems (Ward et al. 2002). The periodic inundation due to floodwaters is mainly responsible for high floodplain productivity (Tockner and Stanford 2002). During periods of inundation, floodplains provide very different habitat conditions than those found in the adjacent river channel. Water velocity generally slows considerably as the flow moves from the river onto the floodplain, allowing the sediment to drop out of suspension (Ward and Stanford 1995). As a result, floodplain water is often less turbid than river water, thus supporting higher photosynthesis rates (Lewis et al. 2000). This primary productivity supports the increased productivity of

196

S. Bhadu and M. Punia

zooplankton and aquatic invertebrates. Floodplains are also known to provide high levels of ecosystem services (Brown et al. 2018). With all urbanisation process-led changes, there is a shift in the type of economic activities over the floodplain. Floodplains are the most suitable sites for cultivation. The economic lives of people living on the floodplains revolve around the rise and fall of local rivers (Sparks 1999). The annual flooding of low-lying areas has important implications for the overall soil fertility due to the deposition of silt and mineral-rich sediments washed down from upstream. With development, the cropping pattern is changing from traditional crops, like rice and corn, etc., to commercial crops like vegetables and nursery gardening. At some places, sand mining has also started in the floodplain. Despite floodplains’ immense ecological and economic values, they have been disconnected from river flows and converted to other land uses in much of the world (Griffith et al. 2008). Also, levees prevent river flows from entering floodplains, whereas dams can significantly alter floods’ magnitude, frequency, and duration. The floodplains are vulnerable to changing land-use patterns—like the expansion of cities and agriculture and flow regulation by dams and barrages. “Soon, floodplains will remain among the most threatened ecosystems, and will disappear faster than any other wetland type” (Tockner and Stanford 2002). River dynamics are affected by the construction of stream break protection structures (Surian et al. 2009) and human encroachment of the flood plain. The bunds, protection belts and artificial barriers alter the river’s natural flow regimes, changing river flow dynamics. Also, a river’s decreased flow and sediment supply bring about remarkable channel changes. In numerous catchments worldwide, human intervention has strongly altered natural river dynamics. A study at the National Remote Sensing Centre (Bhatt et al. 2016) has used remote sensing and GIS to identify changes in the bank line, erosion and deposition. Quantification of changes has been done for the stretches along with the Brahmaputra, Ganga, and Kosi rivers using satellite images from different sources under the above study. Geometric rectification and digitisation and integration of bank lines have been done to get the results. The mapping of three stretches of the Brahmaputra River was done in Assam between 2002 and 2010. The changes in the Ganga River from 1990 to 2004 were mapped in Bihar, and for the years 1987, 1992, and 1997 these changes were mapped in West Bengal. Changes in the Kosi River have been mapped for 1997–2011 in Bihar. The author has suggested that planners use remote sensing for planning along the river bed, monitoring and constructing flood control structures. This technology can produce models, river course changes, and mitigation measures. Different kind of remote sensing data and images has been used in this paper for mapping (Sahana and Patel 2019). Nation Institute of Hydrology and the Indian Institute of Technology Roorkee participated in a research project to quantify the actual bank erosion and deposition along the Brahmaputra riverbank in India from Dibrugarh to Dhubri, a 620 km long stretch for a period of 18 years from 1990 to 2008 (Archana et al. 2012). The whole stretch has been divided into sections for study and then mosaicked together.

10 Governance and Floodplain Extent Changes of Yamuna River …

197

Opperman in his study about ecological functionality of floodplains proposes a conceptual model for analysing hydrological connectivity between the river and the floodplain (Opperman et al. 2010a). He concluded that although floodplains support high levels of biodiversity and are one of the most productive ecosystems on earth, they are also among the most converted and threatened ecosystems. For this reason, the focus has recently been shifted to conservation and restoration plans of floodplain globally. The above-explained works show the utility and application of RS GIS to study the dynamics of the riparian system. This information, combined with climate data and flood data, can be used to generate a model to control flood hazards and erosion. The unchecked development leads to widespread degradation and destruction of riparian habitats and impacts hydrological regimes, alluvial aquifer storage, water quality, geomorphologic patterns. Till 2012 most ecological studies on urban rivers have focused on the smaller systems within urbanised catchments rather than heavily urbanised and engineered rivers (Francis 2012). New ideas about urbanisation are necessary because the city is a human-dominated ecosystem and poses significant social and environmental challenges. Bhan studied the paradigm shift in the history of urban settlements and concluded that urbanisation as a process has transformed over time (Bhan 2009). He has described the recent process of evictions of slum areas which are legally notified areas, jhuggijhopri clusters, which are not notified under the Slum Area Act. The non-poor unauthorised settlements are legalised in situ, and the new name given to them is “regularized—unauthorised colonies”. Further he has used Roy’s concept of informality which says that it is not something outside the planned/formal system; in fact, the state itself produces it and suggested that new inclusive politics must begin where poor and their advocates find appropriate legal, political and cultural representation. These much needed studies may be able to address the exclusion of the urban poor. At this stage of urban growth, it is vital to understand the accompanying environmental and social threats it poses towards sustainable and harmonious urbanisation. Cities are losing their harmony with nature and social systems, which manifests through the damage to the ecosystem, giant ecological footprints, disparities in environmental and economic infrastructure, social and economic insecurities, poverty, and crime (Heinrichs et al. 2009). A study on the Red River of Vietnam presents an excellent learning example for other cities in seasonal rainfall regions located next to the river that non-structural measures should be considered (McElwee et al. 2017). This river used to cause flooding in the rainy season, so a dike of 14.5 m was constructed to prevent flooding and get water for irrigation. With time, the quantity of alluvium made the water level higher progressively. This and other construction developments in the floodplain have become one of the primary reasons to increase the risk of inundation in this zone. French geographer Pierre Gourou observes that, on the one hand, the Red River has been ignored while planning the urban development of Hanoi (Phong 2015). On the other hand, this river is not considered an important natural component for sustainable development.

198

S. Bhadu and M. Punia

Follmann elaborated that, after creating the informalities in the urban environment, the state starts to plan or remap the urban landscape for desired development projects through un-mapping (Follmann 2015). Informality has become the mode of urbanisation (Roy 2005). The state across the globe seems to be working for the neoliberal agents by intentionally amending existing environmental laws and regulations to facilitate urban expansion on flood plains. “To transform their economic landscape and attract investments, cities all around the globe are redeveloping their riverfronts, and many times urban mega-projects are playing an important role in riverfront revitalization” (Follmann 2016). In line with a similar idea, permission has been granted to construct shopping malls and grand hotels in the reserved ridge forest. In his book, “Naturally Tread Softly on the Planet”, Soni advocates stating that conserving the ridge forest as a water sanctuary can beat the real-estate market in cost-benefit analysis. This benefit is based on one condition non-invasive, ‘preserve and use’ method. The second ecologically important zone is the Yamuna floodplain covering an area of 100 km2 and having an average depth of 40–50 m. Floodplain must be a protected zone for water recharge. River floods its bank in the rainy season, depositing sand on the floodplains. This sandy layer is over 5 km wide and 40 m deep for the Yamuna River. This floodplain can hold 1.5 billion cubic hectares. If kept undisturbed, the natural environment of floodplains provides more significant economic benefits than real estate. Preserving the natural aquifer is better than recycling the water using more technological and capital inputs (Soni 2015). The minimum flow in Monsoon Rivers is required during monsoon and nonmonsoon times so that it can perform all its natural functions. These functions are like avoiding the algal choking, sediment transportation, biodiversity of the aquatic environment, recharge flow, dilution flow, and minimum flow. And for the Yamuna River, the level of water withdrawal has crossed the mark of overexploitation. The author has suggested that in the lean season, that is, non-monsoon season, the water requirement must be fulfilled by water harvesting. More advanced instruments must be used for irrigation in agriculture applications. There is a tussle between different departments about the re-delineation of the boundary of the ‘O’ zone, which is a floodplain area because DDA is planning to take out some chunks of land from this zone. Moreover, the present infrastructure and facilities available for sewage and solid waste management are not enough for the size of the population residing in the city. So, this waste is dumped into the Yamuna river without even treating it, making the Yamuna a “dead river” (Iyer 2015).

10 Governance and Floodplain Extent Changes of Yamuna River …

199

10.6 Data Sources 10.6.1 Floodplain Extent Changes The following data sources have been used for the analysis to find out the results concerning changes in the floodplain boundaries. 1. The researcher obtained old maps of the Yamuna floodplain in Delhi for 1893 and 1912 from a series of Printed and Published maps from National Archives of India, Delhi • PP00294, DELHI—Delhi and Vicinity, Scale 4,, = 1 mile, 1912. • PP00291, DELHI—Cantonment, Civil station, City and Environs of Delhi, Scale 6,, = 1 mile, 1867–68, Corrected up to 1893. 2. Toposheet Maps having a scale of 1:50,000, from Survey of India • For 1975, 53 H/2 and 53 /H6. • For 2007, 53 HX/1, 53 HX/2, 53 HX/6, and 53 HX/7. Maps acquired from National Archive and topographical sheets were scanned and then georeferenced in Arc GIS. In the next step, vectorisation work was completed to get the floodplain, riverbank line, river centreline, and agriculture land using maps. For the vectorisation of the floodplain, different criteria were used for different years according to the information provided in the maps. For a detailed analysis of floodplain extent change, it is challenging to study using the whole floodplain area together. In order to understand how the extent of the floodplain is changing in Delhi, it has been divided into six sections. The boundary of the section is taken as Barrages, flyovers, and railway bridges for convenience. The time periods for the current study is taken as 1893, 1912, 1970, and 2007. As Delhi’s areal extent was significantly less in 1893, so out of six, only three sections covered the Yamuna floodplain area. And similarly, out of six, five sections covered the floodplain in 1912. For analysis area for each section was calculated in Arc GIS for all years using the maps mentioned in this section. Floodplain area is divided into these following sections: 1. 2. 3. 4. 5. 6.

Study area to Wazirabad Barrage Wazirabad Barrage to Yudhishthir Setu (GT Road) Yudhishthir Setu (GT Road) to Vikas Marg Vikas Marg to Delhi Noida Direct (DND) Flyover DND flyover to Okhla Barrage Okhla Barrage to Haryana Border in South.

Criteria used for floodplain delineation For vectorisation of floodplain, the best possible criteria were used for different years according to the information provided in the maps. These criteria have been explained below for every period taken.

200

S. Bhadu and M. Punia

Fig. 10.2 Stepwise methodology used for the study of floodplain extent changes

For 1893—For the year 1893, the northern, southern, and eastern boundary of the floodplain is the map’s boundary itself. The western boundary is along with structures and high ground information given on the map. For 1912—For 1912, the eastern boundary is the eastern Yamuna canal, and the western boundary is the contour of 675 feet (205.74 m). Northern and southern boundaries are boundaries of the map itself. For 1970 and 2007—For 1970 and 2007, the eastern and western boundaries are embankment and outer ring road, respectively. The northern boundary is the topographic map’s boundary, and the southern boundary is the boundary of Delhi. For analysis, the area for each section was calculated in Arc GIS for the mentioned time periods. For better visualisation of the scenario, bar graphs, line graphs, and pie charts have been computed in excel. Methodology to find out changes in the floodplain extent is explained in flowchart form in Fig. 10.2.

10.6.2 Governance/Planning Documents, information, and maps area provided by the Delhi Development Authority. For this study we have used reports and literature by different agencies like Master Plan for Delhi-2021 by DDA etc. Methodology for understanding governance is explained in flowchart form in Fig. 10.3. Governance of floodplain zone in Delhi has been studied by understanding the structure of the government, urban governance, and environmental governance on the national, state, and local levels. The map for flood return has been prepared using a map provided by the Delhi Development Authority. Analysis has been done using the literature survey and reports of different governing agencies like Zonal

10 Governance and Floodplain Extent Changes of Yamuna River …

201

Fig. 10.3 Showing methodology used for understanding governance

development plan and Master plan for Delhi-2021 by DDA and reports submitted to NGT regarding the management and planning of the Yamuna river floodplain.

10.7 Analysis/Discussion 10.7.1 Floodplain Extent Changes Floodplains are created through the overflow of excess water of the river over its bank. This situation where water flows beyond the river’s capacity is known as a flood. The floodwater flows according to the slope of land around the riverbank, and the area covered under the floodwater and sediments carried and deposited is known as a floodplain. The recent trend, especially in urban areas, is to control the flood inundated area through embanking and increasing the river channel depth, where the focus is to maximise the water flow through the main channel. But these landscape changes are not healthy for an ecological system because they lead to surface and sub-surface flow changes. In recent times due to haphazard growth and increasing frequency of extreme climatic events it is almost impossible to eliminate flooding. “But with the better management of river floodplains may significantly contribute in minimising the consequences” (Kiedrzy´nska et al. 2015). Thus the false notion of security created through embankments is not a sustainable solution to prevent flooding. Moreover, it reduces the width of the floodplain, which ultimately is harmful to a river system and its surroundings. Section-wise analysis of changes in floodplain extent First Section As we can see from Fig. 10.4, the first section covers the floodplain for 1912, 1970 and 2007. The first section’s north boundary is the study area boundary, and in the south, it extends up to Wazirabad Barrage. The area for the first section is decreasing over time, as we can see from the bar graph Fig. 10.5. There is a significant decrease in the area towards south-east and east from 1912 to 1970, and there is a consistent decrease in the area from north to south on the eastern side of the floodplain from

202

S. Bhadu and M. Punia

1970 to 2007. Similarly, there are changes in the western side in both periods, 1912 to 1970 and 1970 to 2007. To understand changes, if we look at the map in Fig. 10.4, there was a small settlement named Saadatpur on the east side of the floodplain in 1912 and on the west side also a tiny settlement named Jagatapura. But now if we see the situation in 1970, towards south-east direction more settlements have started emerging and embankment has also been constructed towards the east and west side. While in 2007, the settlement area started extending towards the riverbed. An embankment has been constructed just next to the river course. In this way, the floodplain is shrinking more towards the river course in the first section. There is a 1.5747 km2 decrease in the area of the first section from 1912 to 1970 and 9.3925 km2 decrease from 1970 to 2007. In the later period, the rate of decrease in the area is alarmingly high. One of the reason for the decrease in the area is the high rate of urbanisation and population growth.

Fig. 10.4 Floodplain boundary changes in the first section for years; 1912, 1970, and 2007

10 Governance and Floodplain Extent Changes of Yamuna River …

203

25

Area (in sq kms.)

20

15

10

5

0 1912

1970 Years

2007

Fig. 10.5 Change in the area in the first section (km2 )

Second Section The second section lies between the Wazirabad barrage and Yudhisthir Setu as we can see in Fig. 10.6. For the year 1893, the northern and eastern boundary of the floodplain is the boundary of the map itself. For 1912 the eastern boundary is the eastern Yamuna canal, and the western boundary is the contour of 675 feet (205.74 m). For 1970 and 2007, the eastern and western boundaries are embankment and outer ring road, respectively. From 1893 to 1912, the western boundary of the floodplain has shifted 0.1 km towards the east, i.e., towards the river. There is a westward shift of the eastern boundary of the floodplain from 1912 to 1970, with 2.74 km in the north-eastern part and a distance of 1.27 km in the southeastern part. And this area is now covered with settlements. The western boundary of this section for the year 1912 has shifted towards the east direction with a distance of 0.39 km in the north-western part but not any significant shift in the southwestern part as the contour line of 205.74 m was taken as a criterion to draw the floodplain. We can see the changes from Fig. 10.7 bar graph. For 1893, the area was minimum, which contradicts the fact that when there was no settlement around the floodplain, the extent of the floodplain must be significant. The reason for such a result is that this map is taken from National Archive and is an ancient map, so it is not covering the eastern part of the river, which is a floodplain. The eastern boundary of the map

204

S. Bhadu and M. Punia

Fig. 10.6 Floodplain boundary changes in the second section for years; 1912, 1970 and 2007

is the river course itself. In this case, the floodplain boundary has been drawn at 40– 50 m away from the river course. Another reason is that by that time, the northern and southern extent of the city was significantly less as the process of urbanisation started after the city was declared as the country’s national capital by British rule in 1911. Third Section The third section lies between Yudhisthir Setu and Vikas Marg and it is shown in Fig. 10.8. The Eastern and western boundary of the floodplain for 1912, 1970 and 2007 is the embankment, and the western boundary for 1970 and 2007 is the Mahatma Gandhi road. And for 1912, this boundary is based on the 205.74 m contour. The western boundary in the north part is almost fixed for all periods because it is the boundary of the Red Fort and Salim Garh Fort.

10 Governance and Floodplain Extent Changes of Yamuna River …

205

Change in Area Second Section 18 16

Area (in sq kms.)

14 12 10 8 6 4

2 0 1893

1912

1970

2007

Years Fig. 10.7 Change in the area in the second section (in km2 )

There is a difference of 0.02 km between 1912 and 1970. This is because of the combination of errors. This error is the sum of the mapping error, distortion during storage of maps, scanning error, and georeferencing error. The western boundary of the floodplain in 1912 has shifted eastwards in the southwestern part of this section, with a distance of 0.59 km, where the floodplain boundaries for 1970 and 2007 lie. The floodplain boundary on west side in 1912 is moving east and west following the contour line. The western boundary of the floodplain has shifted 0.07 km from 1893 to 1970 and 2007 towards the east. Fourth Section This section extends between Vikas Marg in the north to Delhi Noida Direct (DND) flyover in the south direction as we can see from Fig. 10.10. The Eastern boundary of the floodplain is almost coinciding for the year of 1912, 1970, and 2007 under the embankment. While the floodplain for 1893 extends up to one-fourth of this section in the north, i.e. this was the maximum southern extent of Delhi in 1893. While in the western direction, the Floodplain boundary position has changed continuously from the northern part to southern part with a distance of 1.14 km in north, 0.92 km in middle and 1.02 km in the south; from 1912 to 1970 and 2007.

206

S. Bhadu and M. Punia

Fig. 10.8 Floodplain boundary changes in the third section for years; 1912, 1970 and 2007

This area is now under the network of railway lines and the commercial area of Delhi. The western boundary of the floodplain for 1970 and 2007 is coinciding because it is Mahatma Gandhi Road for these years. Fifth Section The fifth section’s northern boundary is the DND flyover, and the southern boundary is Okhla Barrage as shown in Fig. 10.12. In this section, the eastern boundary of the floodplain in 1912 is the map’s boundary. The boundary of the floodplain in 1970 was shifted towards the west in 2007 with a maximum distance of 0.88 km and a minimum distance of 0.09 km. In the northern part, the western boundary of the floodplain has shifted with a distance of 1.01 km from 1912 to 1970; this area is under settlements like Friends Colony and Bhatnagar. The Floodplain boundary has moved 0.09 km from 1970 to 2007and this shift is mainly because of the shift in the course of the river as for 1970 and 2007, where the western boundary of the floodplain passes through a very narrow corridor and passes just next to river course.

10 Governance and Floodplain Extent Changes of Yamuna River …

207

Change in Area Third Section 14 12

Area (in sq kms.)

10 8 6 4

2 0 1893

1912

1970

2007

Years Fig. 10.9 Change in the area in the third section (in km2 )

In the middle part of this section, the western boundary of the floodplain for all three time periods has shifted towards the east, where the Okhla canal emerges from the right bank of the Yamuna River and flows towards the south. The floodplain map for 1912 ends before Okhla Barrage in the southern part. Floodplain boundary for 1970 goes along the Okhla canal, and boundary for 2007 has moved towards river course i.e., moving eastwards with a distance of 0.61 km. Sixth Section The sixth section extends from Okhla Barrage to the Haryana border in the south. The sixth section has the floodplain boundaries only for 1970 and 2007 as shown in Fig. 10.14. The southern extent of Delhi in 1893 and 1912 was not up to Okhla Barrage. Agra canal starts from the right bank of the river from Okhla Barrage, and while flowing southwards, it joins the Okhla canal. The western boundary of the floodplain in 1970 has shifted towards east by a distance of 1.88 km. This area was covered by Madanpur Khadar and Jhuggi Jhopari cluster in 2007. The Eastern boundary of the floodplain in 2007 is along the embankment, which was not there in 1970. In 1970 the eastern boundary in this section was taken along the roads connecting the small settlements. The floodplain boundary has shifted by 0.36 km towards the west along

208

S. Bhadu and M. Punia

Fig. 10.10 Floodplain boundary changes in the fourth section for years; 1912, 1970 and 2007

the embankment. The changes in the area in this section can be seen by the bar graph in Fig. 10.15. The bar graph in Fig. 10.16 shows the changes in the area of each section from 1912 to 1970. And second bar graph Fig. 10.17 shows the difference in the area from 1970 to 2007. We can see from the bar graph that from 1912 to 1970 there is a reduction of 1.57 km2 for the first section, 6.46 km2 for the second section, 0.79 km2 for the third section and 5.40 km2 for the fourth and 0.96 km2 for the fifth section. The floodplain area for 1912 was up to only the fifth section. The following graph in Fig. 10.17 shows the difference in the area from 1970 to 2007 for each section. In these years, floodplain is covering all six sections completely.

10 Governance and Floodplain Extent Changes of Yamuna River …

209

Change in Area Fourth Section 25

Area (in sq kms.)

20

15

10

5

0 1912

1970 Years

2007

Fig. 10.11 Change in the area in the fourth section (in km2 )

The changes are most significant for the first, fifth and sixth sections, 9.39 km2 , 3.44 km2 and 8.27 km2 , respectively. For the second, third and fourth sections it is very small, (i.e. 0.011 km2 , 0.004 km2 , 0.007 km2 , respectively). The reason behind the very small change in area is that these three sections are confined by embankment in the east direction and ring road, Red Fort and Mahatma Gandhi Road in the west. Figure 10.18 is showing the boundaries of floodplains for all four time periods, 1893, 1912, 1970 and 2007. We can see from map that the extent of Delhi was significantly smaller in 1893. From 1893 to 1912 the extent increased considerably. From 1912 to 1970 the area of the floodplain is reducing mainly in the north-eastern direction. Similarly, from 1970 to 2007 the area of floodplain is constricting in the north-eastern and southern section.

210

S. Bhadu and M. Punia

Fig. 10.12 Floodplain boundary changes in the fifth section for years; 1912, 1970 and 2007

10.7.2 Governance Delhi has two type of status; first, it is the National Capital Territory, and second, constitutionally its legal status is of Union Territory. Because of its dual status, the process of the governance system of Delhi is very complex. If we see India’s Urban Environmental Governance system, the central government’s power is limited because states own control over land. In this way, the state government controls all urban and regional planning decisions. But NCT and union territory, megacity Delhi is jointly governed by the Central Government, Government of National Capital Territory and five local bodies, namely the North Delhi Municipal Corporation, South Delhi Municipal Corporation, East Delhi Municipal Corporation, New Delhi Municipal Corporation and the Delhi Cantonment board. In Delhi, the power over land and urban planning work is under the Central Government’s Ministry of Urban Development. They execute all of its work by

10 Governance and Floodplain Extent Changes of Yamuna River …

211

Change in Area Fifth Section 12

Area (in sq kms.)

10

8

6

4

2

0 1912

1970

2007

Years Fig. 10.13 Change in the area in the fifth section (in km2 )

instructing the Delhi Development Authority. While none of the five local bodies has any role in the planning process of the Yamuna floodplain zone, DDA and its subordinate agencies have an important part in decision-making for the planning, implementation, and maintenance of the floodplain. To understand the governance process, it is essential to understand the structural composition of DDA. The Ministry of Urban Development is the key agency under which several agencies, four of which, play an important role in the river zone planning. First is DDA, second is Delhi Urban Arts Commission, third is Land and Development Office, and fourth is Central Public Works Department. River governance structural framework To understand the Yamuna River governance in Delhi, it is important to understand the Administration system, Urban Governance and Environmental Governance because at some point, all three are involved. There are three flow charts, the first shows the national level structure (Fig. 10.19), the second shows the interstate and state-level structure (Fig. 10.20), and the third shows the local level structure for administration

212

S. Bhadu and M. Punia

Fig. 10.14 Floodplain boundary changes in the sixth section for years; 1912, 1970 and 2007

and government, Urban Governance and Environment Governance (Fig. 10.21). The agencies that play an active role in river planning have been discussed here. Delhi is a National Capital Territory and a union territory simultaneously. The President of India appoints the Lieutenant Governor of Delhi. Delhi, being NCT, has its government. Lieutenant Governor is the administrative head of the state government of Delhi. Lieutenant Governor is head of Delhi Development Authority, which plays a significant role in river governance. DDA is a subordinate agency under the Ministry of Urban Development. Sub-ordinate agencies under the Ministry of Urban Development a. Central Public Works Department Central Public Works Department works to construct and maintain the public built up, environment and infrastructure all over India, excluding Railways, Defence, Communication, Atomic Energy and Airports departments.

10 Governance and Floodplain Extent Changes of Yamuna River …

213

16 14

Area (in sq kms)

12 10 8 6 4 2 0 1970

Years

2007

Fig. 10.15 Change in the area in the sixth section (in km2 )

b. Delhi Urban Arts Commission The primary function of the Delhi Urban Arts Commission is to advise the central government on preserving, developing, and maintaining the aesthetic and quality of urban environmental design in Delhi. The development projects carried out by DDA or any other local body in Delhi are supposed to be reviewed and cleared by the Delhi Urban Arts Commission. c. Land and Development Office Land and Development Office is responsible for the administration of land resources of the government of India in Delhi, maintenance of lease records, the allotment of land to government agencies, selling vacant land and developing properties, and removal of encroachment under its purview. d. Delhi Development Authority The Chairman of DDA is the Lieutenant Governor of Delhi. The Vice-chairman is a senior Indian Administrative Service officer who operates all the DDA’s operations. DDA has all the power to acquire land and can give this land further for development to any public agencies or private developers. There are different subordinate departments in DDA, for example, the planning department, horticulture department, landscape department etc. The landscape department of

214

S. Bhadu and M. Punia

2 1

Area (in sq kms.)

0 -1 -2 -3 -4 -5 -6 -7 1

2

3 4 Sections

5

6

Fig. 10.16 Changes in the area from 1912 to 1970 for all sections

Area (in sq kms.)

Fig. 10.17 Changes in the area from 1970 to 2007 for all sections

0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 1

2

3 4 Sections

5

6

10 Governance and Floodplain Extent Changes of Yamuna River …

215

Fig. 10.18 Temporal changes in floodplain boundary over 1893, 1912, 1970, and 2007

DDA plans the landscape design and green spaces for the area. The horticulture department maintains the plantation and greenery work. The planning department of DDA has prepared the Master Plan for Delhi and its subordinate Zonal Development Plans. The whole area of Delhi has been divided into 15 planning zones from ‘A’ to ‘P’, except ‘I’. In urban Delhi, ‘A’ to ‘H’, six in urban extension ‘J’ to ‘N’ and ‘P’ and one for River Yamuna as zone ‘O’. Any plan prepared by the planning department becomes a legal planning document after getting approval from the Ministry of Urban Development. But before putting it for approval from the Ministry of Urban

216

S. Bhadu and M. Punia

Fig. 10.19 National level structure (Follmann 2015)

Fig. 10.20 Interstate/State level structure (Follmann 2015)

Development, the draft of the plan has to be put public to get suggestions, objections and reviews from the public.

10 Governance and Floodplain Extent Changes of Yamuna River …

217

Fig. 10.21 Local level structure (Follmann 2015)

Zonal Development Plan for ‘O’ Zone For Yamuna River Floodplain Zone ‘O’, DDA has prepared the Zonal Development Plan. This zone has unique characteristics and ecological significance, for which various studies have been conducted to give suggestions to this plan in the form of reports. The ZDP for Zone ‘O’ has been divided into three distinct parts. 1. River bed, i.e., the area under River water 2. River flood plain, i.e., the area between River watercourse and embankments 3. Riverfront, i.e., the area outside the embankments These three areas are segregated by embankments constructed over the years by the Irrigation and Flood Department, GNCTD, based on the activities and floods in Delhi. The present study has focused on the first and second parts. In the following section, we will briefly discuss the essential parts of the Zonal Development plan for the ‘O’ zone. First, we will discuss the proposals and recommendations for the floodplain development in the draft of the zonal development plan for river zone ‘O’, which is subset of the Master Plan for Delhi-2021. In the Master plan for Delhi-1962, it has been pointed out that the entire area towards the north and south direction of the Wazirabad barrage is floodable. So the western bank of Yamuna River south of Wazirabad Barrage can be used for developing district parks, playgrounds and open spaces. Later in the Master Plan for Delhi-2001, it was recommended that large recreational areas should be developed on the floodplain (Zonal Development Plan (as per MPD-2021) | DDA n.d.). The river is an integral part of the city and recommends strict enforcement of the Water Pollution Act to keep the river clean. NCR—Regional Plan 2021 identifies Yamuna River Zone as a natural conservation zone and recommends that the water bodies be kept free from encroachment. Construction activities must not be permitted over the floodplain. At the same time, agriculture and horticulture, pisciculture, social

218

S. Bhadu and M. Punia

forestry, and regional recreational activity with 0.5% construction area have been the recommended activities in this zone (National Capital Region Planning Board, n.d.). According to the Yamuna Action Plan of Slum Wing and Municipal Corporation, low-cost toilets, Sewage treatment plants, electric crematoria, bathing Ghats and plantation work can be done in a floodplain area. In the Master plan of 2021, the emphasis has been put on the rejuvenation of the River Yamuna by ensuring minimum flow downstream by upper riparian states. Forest areas and water bodies must be conserved, no unplanned urban development should occur in these regions, and the riverfront must be fully integrated with the city (Master Plan for 2021 | DDA n.d.). Studies Undertaken for the ‘O’ Zone The first study was conducted by Central Water and Power Research Station, Pune, in 1993. In this study, the river’s channelisation has been recommended in three phases; first, Indraprastha Barrage to Okhla Barrage, the second phase is from Indraprastha barrage Wazirabad barrage and in the third phase from Wazirabad to Palla. Simultaneously the problems associated with the channelisation have been mentioned (Vivekanandan 2015). Inland Waterways Authority of India recommends that instream navigation be integrated with the scheme for development, pollution reduction and flood control. Study on planning and development of Yamuna River bed by the School of Planning and Architecture New Delhi has used three approaches. In the Eco-system based approach, it has been suggested that the water recharge potential of floodplains must be increased, pollution in the river must be reduced by installing the sewage treatment plants at the outfall points of large drains and natural areas must be conserved. The second approach is the integrated development scenario, which suggests that all kinds of development activities must involve the community of that area. The third approach is the post channelisation scenario, which incorporates the reduction in peak floodwater release and pumping regulation to prevent backflow during flood flow (Commissioner and Authority 2014). This study supports the partial channelisation, and some suggestions like water sports encouragement etc., which are not feasible until the pollution level of the river is not reduced. Another study was conducted by the National Environmental Engineering Research Institute (NEERI) for DDA on the Environmental Management Plan for Rejuvenation of River Yamuna. NEERI recommended sub-zone wise development. Categories of particular land use and particular activities have been suggested in the segments of subzones (Gill 2014). In the Environment Governance structure, the Ministry of Environment, Forest and Climate Change and Ministry of Water Resources, River Development and Ganga Rejuvenation are two national-level agencies, and their subordinate agencies are primarily responsible for governing India’s major rivers. The Ministry of Environment, Forest and Climate Change plays an important role in making policies regarding rivers at the national level and granting environmental clearance for urban megaprojects through Environmental Impact Assessment reports. Environmental Impact Assessment (EIA) is a tool of evaluating the positive and negative environmental, economic and social impacts of a project. It ensures the optimal use

10 Governance and Floodplain Extent Changes of Yamuna River …

219

of natural resources for sustainable development and minimises the negative effects on environment. The Ministry of Environment, Forest and Climate Change forwarded the project documents to the Environment Appraisal Committee, consisting of an independent group of experts. This committee assesses the project’s impact on the environment based on the information given in the project documents by the developers and, if needed, can conduct a site visit. Clearance is given based on some conditions. The Central Pollution Control Board is the principal agency for monitoring rivers’ water quality. National River Conservation Directorate is the key agency for implementing the National River Conservation Plan, known as River Action plans, Ganga Action Plan, and Yamuna Action Plan. Yamuna Standing Committee and Upper Yamuna River Board are two interstate agencies. The Yamuna Standing committee deals with the flood protection measures along the Yamuna River bank in Uttar Pradesh, Haryana and Delhi. At the same time, the Upper Yamuna River Board is responsible for maintaining minimum flow in the river and water sharing between the Yamuna riparian states. Delhi Jal Board and Municipal Corporation of Delhi are local-level agencies responsible for implementing the Yamuna Action Plan. The irrigation and flood control Department works for flood protection in the city. Delhi Forest Department is in charge of the forest area maintenance in the river zone. Development of Akshardham Temple Complex and Commonwealth Games Village in Yamuna Floodplain Despite the several rules and regulations for water body conservation river corridor conservation, the two urban megaprojects were carried out successfully in the Yamuna Floodplain in Delhi. The first one is the Akshardham temple complex built from 2000 to 2005. Besides the main temple area, this temple complex has a large parking area, gardens, a musical fountain, exhibition halls, a movie theatre and a food court. Follmann has explained the whole story of the making of the Akshardham temple complex in his PhD work (Follmann 2016). The Temple Trust approached DDA from 1968 to 2000 to get land at the location where it is located today. There was a public protest against the temple’s construction by farmers firstly because they had to lose their lands for this temple complex. Secondly, it was challenged in the Supreme Court by the Uttar Pradesh Irrigation Department through a petition which argued that construction of such a large complex would affect the groundwater recharge. The question was raised against DDA that the land allotted by DDA for the temple belongs to the Uttar Pradesh Irrigation Department. The second urban megaproject was the Commonwealth Games Village. The construction of the Akshardham temple complex paved the way for DDA to make significant profits through the development of Commonwealth Games Village. It was an opening door for Delhi to become a World Class City by hosting these Games. DDA allocated the land next to the Akshardham temple complex to construct the Commonwealth Games Village. DDA changed the land use of a few areas to a residential area to construct apartments for players and some areas to a commercial area to construct five-star hotels. Objections were created against the ad hoc way of

220

S. Bhadu and M. Punia

converting land use and DDA’s way of taking the land from the floodplain by making embankments. When DDA approached MoEF (Ministry of Environment and Forest) to get clearance for Commonwealth Games Village construction, it was pointed out that the site chosen by DDA was below the high flood level of the river. In reply, DDA proposed to fill the site using ash from nearby power plants by one meter to raise the level of the site. The objection was created by an Expert Appraisal Committee (EAC), saying that proper Environmental Impacts have not been studied before choosing the site and suggested that DDA should look for another site because this location was an important area of groundwater recharge zone. But DDA argued that it could not think about the other site because of time constraints. Lastly, MoEF gave clearance for the Commonwealth Games Village under strict conditions. Overall the clearance was given in a hurry because we were running out of time for the organisation work for the event. From these two stories of urban megaprojects, it is clear that the key planning agency did not take the proper course of environmental impacts of these projects over riparian zone very seriously; instead, it focused on getting profit out of these projects. Re-Delineation and Re-Zoning of Floodplain and Riverbed On 7 February 2007 the Central Government authorised MPD-2021 was notified. MPD-2021 stipulates phased evaluation and monitoring on a recurring basis. The Master Plan for Delhi 2021 includes review and monitoring phases. Since September 2011, the first five-year review has been performed. Several suggestions were made to relocate residential areas, villages, etc., in Zone ‘O’ to neighbouring zones E, F, PII, etc., as residents in Zone ‘O’ cannot access essential physical and social infrastructure. Under the review of MPD-2021, the Advisory Group has discussed the subject of redefining the boundary of Zone ‘O’. According to decisions taken in the meeting conducted on 15th June 2012 regarding the re-delineation and re-zoning of floodplain zone ‘O’. Table 10.1 shows the details of localities to be deleted from Planning Zone ‘O’ and added to the adjacent Planning Zones. “Hon’ble L.G. Delhi stated that wherever old embankment/bund has been constructed, the protected area should be excluded from Zone ‘O’. The example of Jaitpur Ext., a village Abadi, has been wrongly shown in Zone ‘O’ in Zonal Development Plan was quoted and hence there is a need to re-fix the Zone ‘O’ boundary. Likewise, Sonia Vihar Residential Colony behind the protection embankment would merit such inclusion” (Master Plan for 2021 | DDA n.d.). It has been mentioned that people living in Meethapur and Jaitpur are not getting permission for reconstruction and repair of the buildings because this area lies in the ‘O’ zone where any type of construction is not permitted. In these unauthorised colonies, MCD is not allowed to build schools, dispensaries and community halls, which is allowed in other unauthorised colonies. One more reason was that some of these colonies claimed to exist away from the river. These are the main reasons for

10 Governance and Floodplain Extent Changes of Yamuna River …

221

Table 10.1 Details of localities deleted from planning zone ‘O’ and added to the adjacent planning zones. Source Draft agenda for authority, Delhi development authority File No.: F.20(12)/2013-MP Sl. No

Land Parcels to be Excluded from Current Zone ‘O’

Area to be excluded from current zone ‘O’ (Ha.)

Area to be excluded from current zone ‘O’ (km2 )

Excluded areas to be included in adjacent Zones

1

Rajghat

213

2.13

A (Walled City)

2

I P Power Stations

112

1.12

D

3

Millennium Bus Depot

33

0.33

D

4

Sonia Vihar Area

718

7.18

E

5

Shahtri Park DMRC Land

104

1.04

E

6

Akshardham Temple, CWG Village

109

1.09

E

7

Yamuna Bank DMRC Land

40

0.4

E

8

Batla House Area

0.73

F

9

Jaitpur, Meethapur, Okhla

10

Jagatpur

11

Area Under Circulation

73 1310

13.1

F

385

3.85

P-II

12

0.12

D and F

taking out all of the following localities shown in the table and the map from Zone ‘O’. Now, if we look at Fig. 10.22 the location of areas has been shown in the ‘O’ Zone map; some areas like Shastri Park DMRC land, Yamuna Bank DMRC land, IP Power station, Batla House area, Sonia Vihar area are very close to the Yamuna river course. After subtracting these areas from the floodplain, the extent of the floodplain will be reduced. This way and logic of taking out the land from the river zone have paved a way further for the same practices of reducing the area of this zone in the next period of land crisis. The largest area taken out is Jaitpur, Meethapur, and Okhla combined, 13.10 km2 , largely residential. This area has been included in the ‘F’ Zone. The second-largest area is Sonia Vihar, 7.18 km2 , which is also a residential area and has been added to the ‘E’ Zone. Totally, 31.09 km2 area has been taken out from the ‘O’ zone and added to ‘D’, ‘E’, ‘F’ and ‘P-II’ zones, located around the ‘O’ zone. The above decision indicates that embankment is the measure to define the floodplain boundary. If this is the case that constructing an embankment can change the extent of the floodplain, then it keeps the chances open for future construction of the embankment towards the river course and further constricting it. Moreover, the focus of DDA through MPD-2021 for the river zone ‘O’ is protection against the flood that was also through constructing embankment. No other measures like removing unauthorised colonies and permanent structures built up so far and restoring the original river ecosystem have been considered important. Another critical aspect

222

S. Bhadu and M. Punia

Fig. 10.22 The areas to be deleted from River Zone ‘O’. Source Produced by the researcher using information given by DDA

that DDA has overlooked is the effort to maintain the floodplain extent. This is a critical percolation zone to replenish the groundwater reservoir of a megacity like Delhi. Another map was created by overlaying the flood return map given by DDA and areas to be taken out of the floodplain zone, ‘O’ from Wazirabad to Okhla which is shown in Fig. 10.23. The location map showing the places to be taken out from the river zone was overlapped with the flood return map to understand the situation more closely. As we can see in Fig. 10.23, a few places like Yamuna Bank DMRC land, Commonwealth Games Village area, Shastri Park DMRC land Batla House area, and Millennium Bus Depot are located in the 25-year flood return area.

10 Governance and Floodplain Extent Changes of Yamuna River …

223

Fig. 10.23 Flood return map showing proposed areas to be deleted from the floodplain. Source Adapted by the researcher using information given by DDA

These all locations are vulnerable to flooding once in every 25 years. And after taking these areas out of the river zone, they can be developed without any regulations, which can cause loss of property and structure at the time of the flood. “DTC, DMRC has constructed embankments/raised the land level for locating depots so that they are not affected by floods though located within 300 m from River Yamuna. Given this, similar areas and locations, based on ground realities, could be examined for excluding from ‘no construction zone’” (Master Plan for 2021 | DDA n.d.). Even the Batla house area is a residential area with permanent concrete construction. At the time of heavy rainfall in Delhi and the surrounding catchment of the

224

S. Bhadu and M. Punia

Yamuna River, these areas will face heavy life and material losses. All these practices point towards the critical issue of narrowing the floodplain as there is no clarity about the policy regarding the floodplain boundary. Under such conditions, more land will be diverted for construction, further increasing the problems. On the one hand, DDA provides the regulations to disallow any permanent construction, and on the other side, DDA has not decided the final boundary of the river zone. DDA has proposed to take out some areas within the limit of 300 m away from the river channel for construction purposes despite knowing that the Central Ground Water Authority has declared the Yamuna floodplain a vital part of the groundwater recharge zone. In a report published on 28th March 2016, it has been mentioned that the National Water Mission under the National Action plan on Climate Change recommended the zoning of the floodplain to avoid climate change-induced disasters. In the same report, Shekhar, head of the National Green Tribunal’s Principal Committee on Yamuna and secretary of the ministry of water resources, says that developed countries like Germany and England have started to gradually remove embankments and dykes to be ready to prepare against floods. Shekhar also explained that the panel was aware of floodplain constriction, but there was no law to protect the floodplain against encroachment. The land belongs to three agencies; DDA, Delhi Irrigation Department, and Uttar Pradesh Irrigation Department. Similar issues regarding ownership are persisting in 2017 also. On 16th June 2017, another report published in TOI talks about the ambiguity of ownership. This report is about Jaitpur and Mithapur area. It has been mentioned in the report that Jaitpur is now among the concrete extensions in the ecologically important Yamuna River zone in addition to the Commonwealth Games Village, Akshardham Temple Complex, and Millennium Bus Depot. DDA, a land-owning agency in Delhi, says that these locations are not development areas but are part of a private land, so South Delhi Municipal Corporation is responsible for preventing unauthorised construction. In reply to DDA’s response, South Delhi Municipal Corporation says DDA has the primary responsibility to protect the Yamuna floodplain. National Green Tribunal (NGT) was established in 2009 to handle and expeditiously dispose of cases about environmental issues. It draws its inspiration from Article 21 of the Indian Constitution, which states, “No person shall be deprived of his life or personal liberty”. The Hon’ble Supreme Court has passed various orders giving a liberal interpretation to Article 21, including the Right to a Clean and Healthy Environment in Right to Life and Liberty. The Supreme Court has ordered that ‘Maily Yamuna’ be transformed into a healthy and pure Yamuna in Writ Petition No. 725 of 1994 and other related matters. At least between Hathnikund in Haryana and the Taj Mahal Monitoring Station in Agra, the water should be the least polluted. The Yamuna has been transformed into a drain conveying sewage, domestic garbage, and industrial and commercial effluents because authorities lacked the essential will to execute instructions, plans, and programmes in a sincere and effective manner. The more concerning and disturbing situation is that the state has failed to fulfill its constitutional and statutory obligations, while citizens have also failed to fulfil their fundamental obligation to protect the environment, particularly in regards to the Yamuna River.

10 Governance and Floodplain Extent Changes of Yamuna River …

225

The Supreme Court in the case of Subhash Kumar v. State of Bihar and Ors. (1991) held that: the “right to live is a fundamental right under Article 21 of the Constitution. It includes the right to enjoy pollution-free water and air for full enjoyment of life” (Manoj Misra vs Union Of India on 31 May, 2021, n.d.). And thereby drawing from this judgment, NGT has held that maintaining an environmental flow regime in river Yamuna even during non-monsoon months is a non-negotiable obligation on the part of the state to provide its citizens with a clean and healthy environment. Also, at the same time, the significance of storm water drains has been highlighted as they are the natural channels that maintain the water flow in the River Yamuna in this region. It is therefore essential that they must be kept obstruction and pollutionfree. Hence, dumping solid waste in close proximity to the river is strictly prohibited, and a heavy penalty is imposed on the violators. It cannot be disputed that the present status of Yamuna is highly polluted river due to the lack of freshwater flow, discharge of untreated or partly treated domestic and industrial waste, and due to dumping of debris on its banks and in it. Its flood plains are highly truncated and degraded, resulting in depletion in most of its natural bio-diversity. Hence, NGT has directed all the concerned stakeholders that all the natural and artificial drains in Delhi should be kept clean, free from obstruction, dumping any material and municipal waste should be strictly prohibited.

10.8 Conclusion It is clear from the analysis that the area under the first section is reducing continuously from 1912 to 1970 and then from 1970 to 2007. The area for 1893 in the first section is not comparable because of the limited extent of the map for 1893. The first section is the northernmost part of the study area and which was not part of the city in 1893. Over decades the city is expanding on the way to urbanisation and this part of the floodplain became part of Delhi and is shrinking. The second section’s area has also reduced with time. The floodplain map for 1893 is starting in the upper middle part of the second section. So we can measure changes from 1912 onwards and the area is reducing from 1912 to 1970 with a big difference and from 1970 to 2007 with a minor difference. For the third section, the trend is the same from 1912 to 1970 and 1970 to 2007, reduction in the extent with a smaller difference in the earlier time period. For the fourth section, the areal reduction from 1912 to 1970 is more than the third section but from 1970 to 2007 it is similar to the third section. The area of the fifth section is showing an increment from 1912 to 1970 because the map extent for 1912 is not covering the fifth section completely. While from 1970 to 2007 the area of the fifth section is reducing. From 1970 to 2007 the area under the sixth section has reduced by more than eight square kilometres. From 1912 to 1970, the reduction in area from one section to the next section is abrupt while from 1970 to 2007 the reduction was the highest in the first section, it is small and almost equal for the second, third and

226

S. Bhadu and M. Punia

fourth section, which shows that the area under these three sections is almost constant in 2007 as it was in 1970. After understanding the complex governance system of the river zone and from the above analysis concerning the floodplain extent changes and documents related to planning and management of the Yamuna River floodplain by different governing agencies, we can conclude that some studies in their reports are critical about the proposals given for the development of floodplain in the Zonal Development Plan for the ‘O’ zone. At the same time, few of them are supporting the proposal. Some institutions have prepared the reports after conducting a visit to the river zone, and different institutions also give suggestions for the zonal development plan for the river zone ‘O’. Above all, there is a lack of clarity about land ownership among the Delhi Development Authority and Municipal corporations, as in the case of the TOI (Times of India) report on Jaitpur and Mithapur. Despite knowing about the associated risk, DDA is vaguely taking actions, e.g., DDA is taking out areas from the river zone located in 25 years of flood return period. Floodplain is shrinking because the landowning authorities keep themselves from taking responsibility for encroachments and concretisation in the Yamuna river zone. Governing authorities are, instead of working in coordination with each other, are working in a very complex way by transferring the assignments/responsibilities to each other. At the same time, Akshardham is the result of political pressure over DDA and other involved agencies and the Commonwealth Games Village is the result of lack of planning in time. We can sum up the consequences of all such practices in the following points: 1. The city is losing essential groundwater recharge and ecologically rich zone due to illegal and planned encroachment in the present climate change scenario. 2. The river is losing its way for hydrological evaluation because of a lack of space. 3. At the time of natural disasters like flooding the city is losing sponges which are wetlands regions that could have protected the city by giving passage to overflow. Acknowledgements This chapter and the research behind it would not have been possible without the exceptional support of my supervisor, Professor Milap Punia. His enthusiasm, knowledge and exacting attention to detail have been an inspiration and kept my work on track. This research was supported by University Grant Commission in form of Junior Research Fellowship and B. R. Ambedkar Central Library, Jawaharlal Nehru University in the form of resources.

References Ahluwalia IJ, Kanbur R, Mohanty PK (2017) Urbanisation in India: challenges, opportunities and the way forward, vol 91. Sage Publications Sarkar A, Garg RD, Sharma N (2012) RS-GIS based assessment of river dynamics of Brahmaputra River in India. J Water Resour Prot 2012(02):63–72.https://doi.org/10.4236/JWARP.2012.42008 Baviskar A (2006) Demolishing Delhi: world class city in the making by Amita Baviskar. Mute 2(3)

10 Governance and Floodplain Extent Changes of Yamuna River …

227

Bhan G (2009) This is no longer the city I once knew. Evictions, the urban poor and the right to the city in millennial Delhi. Environ Urbanization 21(1). https://doi.org/10.1177/095624780910 3009 Bhatt C, Rao G, Manjusree P, Bhanumurthy V (2016) Remote sensing and GIS technology for flood monitoring and management. Academia.Edu. https://www.academia.edu/download/73358979/ book_2017_BITS_Ch_12.pdf Bholey M (2016) India’s urban challenges and smart cities: a contemporary study. Scholedge Int J Bus Policy Governance 3(3). https://doi.org/10.19085/journal.sijbpg030301. ISSN 2394-3351 Brown AG, Lespez L, Sear DA, Macaire JJ, Houben P, Klimek K, Brazier RE, Van Oost K, Pears B (2018) Natural vs anthropogenic streams in Europe: history, ecology and implications for restoration, river-rewilding and riverine ecosystem services. Earth Sci Rev 180:185–205. https:/ /doi.org/10.1016/J.EARSCIREV.2018.02.001 Commissioner A, Authority DD (2014) Planning law & environment laws DFID (2002) Sustainable urbanisation: achieving agenda 21. In: UN-HABITAT 1792(August) Field CB, Barros V, Stocker TF, Dahe Q, Jon Dokken D, Ebi KL, Mastrandrea MD, Mach KJ, Plattner GK, Allen SK, Tignor M, Midgley PM (2012) Managing the risks of extreme events and disasters to advance climate change adaptation: special report of the intergovernmental panel on climate change. In: Managing the risks of extreme events and disasters to advance climate change adaptation: special report of the Intergovernmental Panel on Climate Change, 9781107025066, 1–582.https://doi.org/10.1017/CBO9781139177245 Follmann A (2015) Urban mega-projects for a ‘world-class’ riverfront – The interplay of informality, flexibility and exceptionality along the Yamuna in Delhi, India. Habitat Int 45(P3):213–222. https://doi.org/10.1016/J.HABITATINT.2014.02.007 Follmann A (2016) Governing riverscapes. Urban environmental change along the River Yamuna in Delhi, India [Franz Steiner Verlag]. http://www.uni-koeln.de/ Francis RA (2012) Positioning urban rivers within urban ecology. Urban Ecosyst 15(2):285–291. https://doi.org/10.1007/S11252-012-0227-6/FIGURES/2 Gill GN (2014) Environmental protection and developmental interests: a case study of the River Yamuna and the Commonwealth Games, Delhi, 2010. Int J Law Built Environ 6(1). https://doi. org/10.1108/IJLBE-03-2013-0010 GNCT of Delhi (2013) Economic Survey of Delhi 2012–13. In: Directorate of economics & statistics Tockner K, Bunn SE, Gordon C, Naiman RJ, Quinn GP, Stanford JA (2008) Á flood plains: critically threatened ecosystems. Griffith Research Online. https://research-repository.griffith.edu.au/bit stream/handle/10072/23618/53241_1.pdf;sequence=1 Heinrichs D, Aggarwal R, Barton J, Bharucha E, Butsch C, Fragkias M, Johnston P, Kraas F, Krellenberg K, Lampis A, Ling OG (2009) Adapting cities to climate change: opportunities and constraints (findings from eight cities). In: 5th urban research symposium, cities and climte change Iyer RR (2015) Living rivers, dying rivers. Oxford University Press. https://global.oup.com/aca demic/product/living-rivers-dying-rivers-9780199456222 Kiedrzy´nska E, Kiedrzy´nski M, Zalewski M (2015) Sustainable floodplain management for flood prevention and water quality improvement. Nat Hazards 76(2):955–977. https://doi.org/10.1007/ S11069-014-1529-1/FIGURES/4 Kingsford RT (2000) Ecological impacts of dams, water diversions and river management on floodplain wetlands in Australia. Aust Ecol 25(2):109–127. https://doi.org/10.1046/J.1442-9993. 2000.01036.X Kohli A (2006) Politics of economic growth in India, 1980–2005. Part I: The 1980s. Econ Polit Wkly 41(13):1251–1259. http://www.jstor.org/stable/4418028 Lewis WM, Hamilton SK, Lasi MA, Rodríguez M, Saunders JF (2000) Ecological detemination on the Orinoco floodplain: a 15-year study of the Orinoco floodplain shows that this productive and biotically diverse ecosystem is functionally less complex than it …. Academic.Oup.Com. https://academic.oup.com/bioscience/article-abstract/50/8/681/243272

228

S. Bhadu and M. Punia

Manna A (2019) Yamuna River Project: New Delhi urban ecology. J Landscape Arch 14(2). https:/ /doi.org/10.1080/18626033.2019.1673597 Manoj Misra vs Union of India on 31 May, 2021. (n.d.). Retrieved September 18, 2022, from https:/ /indiankanoon.org/doc/117961828/ Master Plan for 2021 | DDA. (n.d.). Retrieved September 18, 2022, from https://dda.gov.in/masterplan-2021 McElwee P, Nghiem T, Le H, Vu H (2017) Flood vulnerability among rural households in the Red River Delta of Vietnam: implications for future climate change risk and adaptation. Nat Hazards 86(1). https://doi.org/10.1007/s11069-016-2701-6 National Capital Region Planning Board. (n.d.). Retrieved September 26, 2022, from https://ncrpb. nic.in/regionalplan2021.html Opperman JJ, Luster R, McKenney BA, Roberts M, Meadows AW (2010a). Ecologically functional floodplains: Connectivity, flow regime, and scale. J Am Water Resour Assoc 46(2). https://doi. org/10.1111/j.1752-1688.2010a.00426.x Opperman JJ, Luster R, McKenney BA, Roberts M, Meadows AW (2010b) Ecologically functional floodplains: Connectivity, flow regime, and scale. J Am Water Resour Assoc 46(2):211–226. https://doi.org/10.1111/J.1752-1688.2010.00426.X Paul S, Saxena KG, Nagendra H, Lele N (2021) Tracing land use and land cover change in periurban Delhi, India, over 1973–2017 period. Environ Monit Assess 193(2). https://doi.org/10. 1007/s10661-020-08841-x Pearce D, Barbier E, Markandya A (2013) Sustainable development: economics and environment in the third world. In: Sustainable development: economics and environment in the third world. https://doi.org/10.4324/9781315070254 Phong LH (2015) The relationship between rivers and cities: influences of urbanization on the riverine zones – a case study of Red River zones in Hanoi, Vietnam. Sustain Dev Plan VII 1.https://doi.org/10.2495/sdp150031 Praharaj S, Han JH, Hawken S (2018) Urban innovation through policy integration: critical perspectives from 100 smart cities mission in India. City Cult Soc 12:35–43. https://doi.org/10.1016/J. CCS.2017.06.004 Roy A (2005) Urban informality. J Am Plan Assoc 71(2) Sahana M, Patel PP (2019) A comparison of frequency ratio and fuzzy logic models for flood susceptibility assessment of the lower Kosi River Basin in India. Environ Earth Sci 78(10). https://doi.org/10.1007/S12665-019-8285-1 Soni V (2015) Naturally:tread softly on the planet. HarperCollinsPublishersIndia. https://harpercol lins.co.in/product/naturally-tread-softly-on-the-planet/ Sparks RE (1999) Need for ecosystem management of large rivers and their floodplains. NCASI Tech Bull 2(781):507. https://doi.org/10.2307/1312556 Surian N, Ziliani L, Comiti F, Lenzi MA, Mao L (2009) Channel adjustments and alteration of sediment fluxes in gravel-bed rivers of North-Eastern Italy: potentials and limitations for channel recovery. Wiley Online Libr 25(5):551–567.https://doi.org/10.1002/rra.1231 Tockner K, Stanford JA (2002) Riverine flood plains: present state and future trends. Environ Conserv 29(3):308–330. https://doi.org/10.1017/S037689290200022X Vivekanandan N (2015) Comparison of EV1 and LP3 distributions using goodness-of-fit and diagnostic tests for estimation of design flood. Citeseer 116–126. https://citeseerx.ist.psu.edu/vie wdoc/download?doi=10.1.1.741.6860&rep=rep1&type=pdf Ward JV, Stanford JA (1995) Ecological connectivity in alluvial river ecosystems and its disruption by flow regulation. Regul Rivers Res Manage 11(1):105–119. https://doi.org/10.1002/RRR.345 0110109 Ward JV, Tockner K, Arscott DB, Claret C (2002) Riverine landscape diversity. Freshw Biol 47(4):517–539. https://doi.org/10.1046/J.1365-2427.2002.00893.X Zonal Development Plan (as per MPD-2021) | DDA (n.d.). Retrieved September 26, 2022, from https://dda.gov.in/zonal-development-plan-mpd-2021

Chapter 11

Assessing Urban Compactness Using Machine Learning and Earth Observation Datasets: A Case Study of Kolkata City Prosenjit Barman and Sk. Mustak

Abstract The development of a compact city is an alternate sort of city development plan for dealing with a quickly urbanizing metropolitan area. City compactness is essential in analysing the current situation so the city can develop sustainably. Urban sustainability has widely been introduced in several countries to achieve sustainable development goals (e.g., SDGs-11) in terms of efficient land use, sustainable transportation, socially interactive environment, economic viability, and is environmentally protected. The measures of city compactness are crucial for assessing urban form over time. Long debates have been about the necessity of studying city compactness concerning a place that has already achieved a dense urban. The main objective of this study is to assess the compactness of urban footprints using machine learning and earth observation datasets. This study used sentinel-2 and Landsat TM data during 1990–2021 which were used to map the urban footprint using machine learning algorithms (e.g., SVM, RF, etc.). Urban compactness has been measured from the building footprint using spatial density, spatial matrices, spatial intensity, functional compactness index (FCI), diagnosing sprawl, mixed land use development, multicriteria decision making, and byes theorem. In this study, the compactness has been measured using spatial matrices. The result shows that the compactness of the builtup area in KMA is non-linear during different decades, as it was sometimes sprawling and sometimes compact. This type of growth pattern shows peri-urban sprawling and city center approaching towards compact development. This study would help policymakers, city planners, and local governments to improve their understanding of the urban form in terms of urban compactness to implement development plans in Kolkata urban agglomeration area. Keywords Urban compactness · Machine learning · Earth observation · Object-based image analysis (OBIA) · Urban sustainability P. Barman · Sk. Mustak (B) Department of Geography, School of Environment and Earth Sciences, Central University of Punjab, Bhatinda, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_11

229

230

P. Barman and Sk. Mustak

11.1 Introduction The development of a compact city is an alternate sort of city development plan for dealing with a quickly urbanizing metropolitan area. Compact development is a significant solution for sustainable development in a growing city (Lan et al. 2021). The rising urbanization of the population and the proliferation of cities has generated a vital role in long-term urban development (Min et al. n.d.). After the creation of global industrial cities, a new notion of sustainable city development has emerged, requiring the development of a new theory or approach to developing city development worldwide (Yang et al. 2017). For this reason, cities should concentrate on long-term development based on environmental effects and management. The sustainable urban form can be defined as “the city to function within its natural and manmade carrying capacities—is user-friendly for its occupants and promotes social equity” (Willams 2001). There were many models developed for sustainable city development. Based on this, the seven concepts of compactness, such as sustainable transportation, density, mixed land use, diversity, passive solar design, and greening, have been explored to develop a sustainable city (Kotharkar et al. 2014). Compact city development is an ideal urban form defined by density, connectivity, and accessibility, which can be achieved through transit-oriented development (Atmadja and Bogunovich n.d.). Compact cities are the most environmentally friendly form of urban development, and this city is distinguished by its high density, land use diversity, accessibility, and public transportation efficiency. There is no standard measure for measuring compactness, although the impact of mixed land use development on compact city growth has been explored (Abdullahi et al. 2015a, b). Many kinds of literature exist in India and overseas that measure urban compactness at the city level. Stathakis and Tsilimigkas (2015) measured urban compactness in European countries and compared them to each other using new and existing metrics and fused data. The author used GIS and remote sensing technology to measure the urban form to discriminate between urban sprawl and compactness in a significant urban region. Size, form, continuity, density, scattering, and loss of green space were six-dimension indicators for measuring sprawl in a big urban area (Sim 2011). It also explored how land use data can be used to analyze urban sprawl, such as low density, low compactness, and so on (Steurer and Bayr 2020). For the urban extension analysis, a composite index was created based on the indices that affect urban compactness for an urban land analysis (Mubareka et al. 2011). Rahman et al. (2022) measured a compactness level using GIS in the urban form using six indicators: population density, evenness development, clustering development, land use diversity, floor use mix, and road network connectivity. The Multicriteria decision-making approach was used to develop the composite compactness index. The study area was classified into three categories: low compact, medium compact, and extremely high compact, with accuracy assessed using the study area’s travel behaviour. A comprehensive city compactness assessment employing GIS and radar technology was used to measure urban sustainability. Mixed land use development,

11 Assessing Urban Compactness Using Machine Learning and Earth …

231

urban density, and intensity are the main indicator for compactness analysis. Multicriteria decision-making and the Bayes theorem were applied to measure the overall compactness (Abdullahi et al. 2015a, b). The five crucial features of urban forms, such as type of clustered development, centralization, proximity, diversity, and the ratio of blocks to the urban space, were examined to distinguish between urban sprawl and compactness (Shamkhi Al-Khafaji and Abdul-Majeed Al-Salam 2018). Another compactness measuring index, the functional compactness index (FCI), was discovered utilizing physical and socioeconomic data. This FCI was calculated at the grid level in the study area, and it fundamentally distinguishes city compactness based on functioning (Lan et al. 2021). The study’s main goals are to map the urban footprint in the Kolkata metropolitan area \using machine learning and earth observation datasets from different years and to measure compactness using multiple spatial metrics.

11.2 Methodology The methodological section has been divided into three sections. This section has been discussed below.

11.2.1 Study Area For the compact urban analysis, the Kolkata metropolitan area has been selected. The study area mainly consists of 4 municipal corporations, 38 municipalities, and many other villages (Fig. 11.1). This is the 4th largest metropolitan city in India. Kolkata is the largest metropolitan city in the north-eastern region of India. This metropolitan region has mainly been growing along the Hooghly River on both sides. KMA extends over six districts situated on either side of the river Hooghly namely Kolkata, Howrah, Hooghly, Nadia, North 24 Parganas, and South 24 Parganas (Kolkata Metropolitan Development Authority).

11.2.2 Datasets The following datasets have been used to measure the urban compactness in Kolkata metropolitan area. For the LULC analysis of the Kolkata metropolitan area, the Landsat surface reflectance data have been collected in different years using the Google Earth Engine. (i) Landsat TM data in 1990 (ii) Landsat TM data in 2001 (iii) Landsat TM data in 2011 (iv) Landsat OLI data in 2021.

232

P. Barman and Sk. Mustak

India in Globe

Study Area

Fig. 11.1 Study area

11.2.3 Methods The following methods were employed to analyse the compactness of the KMA (Fig. 11.2).

11 Assessing Urban Compactness Using Machine Learning and Earth …

Landsat datasets

Review paper analysis

Splitting the Bands of Landsat data

Calculate NDVI and NDBI

Selection of class Normalize the Bands

Stack Bands

Compute image statistics

Training sample collection

Train SVM RBF ML model

Spatial metrices

Land use/cover map

Compactness analysis

Accuracy assessment

Conclusion and recommendation

Fig. 11.2 Methodological flowchart

233

234

11.2.3.1

P. Barman and Sk. Mustak

Land Use and Land Cover Mapping

The land use and land cover of KMA was done using SVM-RBF and satellite datasets in QGIS software. The land use/cover classes, such as water body, wetland, built-up, agricultural land, fallow land, vegetation, and open land, were selected based on the standard land use and land cover classification scheme. In this study, a minimum of 50 training samples per land use/cover class has been selected for the land use/cover classification. In comparison, 50 test samples per land use/cover classes have been used for accuracy assessment based on kappa and overall accuracy. Accurate land use/cover maps have been used for compactness analysis of the KMA.

11.2.3.2

Compactness Assessment

In this study, urban compactness assessment was done based on spatial metrices. Multidate LULC were used to calculate several spatial metrices to quantify urban compactness. These metrices were calculated using FRAGSTATS software at patch, class, and landscape level. Spatial metrics The metrics have been calculated using Fragstats open-source software. Many metrics have been calculated to analyze the urban compactness in the Kolkata metropolitan area. The matrices include area, shape, aggregation, and diversity metrics which were calculated using the Fragstats software. Area metrics Largest patch index (LPI) The largest patch index (LPI) defines the area of the largest patch of the corresponding patch type divided by the total landscape area multiplied by 100. LPI value range is 0–100, while 0 indicates the largest patch of the corresponding patch type that is increasingly small, and 100 deals with the entire landscape consisting of a single patch of the corresponding patch type. Mainly this metric measures the dominancy (Herkert 1999). a max(ai j ) j =1 LPI = (100) A where, aij = area of patch ij A = total landscape area. Class area (CA) The class area (CA) is the sum of the area of all patches of the corresponding patch type divided by 10,000. Its measures unit in hectare. The class area measures the

11 Assessing Urban Compactness Using Machine Learning and Earth …

235

landscape composition, especially how much of the landscape comprises a particular patch type. (Lehmkuhl and Raphael 1993). CA =

a 

 ai j

j=1

1 10000



where, aij = area of patch ij. Shape metrics Shape index (SI): Shape index (SI) equals to patch perimeter divided by the square root of the patch area adjusted by a constant to adjust for a square standard. Its range exists from one and above. One describes the patch as square and increases without limit as the patch shape becomes more regular. (Forman and Godron 1986). 0.25Pi j SHAPE = √ ai j where, Pij = perimeter of patch ij; aij = area of patch ij. Contiguity index: Contig equals the average contiguity value for the cells in a patch, i.e., the sum of the cell values divided by the total number of pixels in the patch minus 1, divided by the sum of the template value minus 1. Its value exists between 0 and 1. 0 reveals a one-pixel patch, and its increase to a limit of 1 a patch continuity and connectedness. This Index assesses the spatial connectedness or contiguity of cells within a grid cell patch to provide an index on patch boundary configuration and, thus, patch shape (LaGro 1991).  z CONTIG =

r =1 C i jr ai∗j

−1



V −1

Aggregation Metrics Euclidean Nearest-Neighbour Distance (ENN): ENN equals the distance to the nearest neighbouring patch of the same type based on a shortage edge-to-edge distance. Edge-to-edge distance means cell center to cell center. It measures unit meters. Its value ranges from 0 to the upper limit, while 0 reveals the nearest neighbour’s distance decreases. This ENN index is the simplest measure of patch context and is used to quantify patch isolation. ENN distance is defined using the simplest Euclidean geometry as the shortest straight-line distance between the focal patch and its nearest neighbour of the same class (McGarigal and McComb 1995). ENN = h i j

236

P. Barman and Sk. Mustak

hij = distance from patch ij to the nearest neighbouring patch of the same type based on the shortest edge to edge distance. Aggregation index (AI): Aggregation index (AI) equals the Number of like adjacencies involving the corresponding class divided by the maximum possible number of like adjacencies involving the corresponding class. This is achieved when the class is maximally clumped into a single compact patch multiplied by 100. Its measuring unit is percent. Its range is 0–100. 0 value demarcates that the focal patch type is maximally disaggregating. AI increases as the type focal patch type is increasingly aggregated and is equal to 100 when the patch type is maximally aggregated into a single compact patch. The aggregation index is calculated from an adjacency matrix which shows the frequency of different pairs of patch types. The aggregation index is based on adjacencies tallied using the single count method in which each cell side is counted only once (He et al. 2000). gii AI = (100) max − gii 

where, gii = no of like adjacencies between pixels of patch type based on the single count method; max − gii = maximum no of like adjacencies between pixels of patch type i based on the single count method. Landscape shape Index (LSI) Landscape shape Index (LSI) equals 25 times the sum of the entire landscape boundary and all edge segments within the landscape boundary involving the corresponding patch type, including some or all of those bordering backgrounds divided by the square root of the total landscape area. Its range is one and above. LSI is equal to 1 when the landscape consists of a single square patch of a corresponding type; the increase of LSI landscape shape becomes more irregular, and the length of edge within the landscape of the corresponding patch type increases (Patton 1975). LSI =

0.25

m ∗ e √ k=1 ik A

∗ where eik = total length of the edge in the landscape between patch type i and k; A = total landscape area

Patch cohesion index: The patch cohesion index measures the physical connectedness of the corresponding patch type. Patch cohesion increases as the patch type becomes more clumped or aggregated in its distribution, hence more physically connected. Cohesion equals one minus the sum of the patch perimeter divided by the sum of patch perimeter times the square root of patch area for patches the corresponding patch type divided by one minus one over the square root of the total no of cells in the landscape, multiplied by 100 to convert to a percentage (Schumaker 1996). Its range is 0–1.

11 Assessing Urban Compactness Using Machine Learning and Earth …



n

COHESION = ⎣1 −  n

237

⎤  Pi∗j 1 −1 ⎦ √ · 1− · (100) Z Pi∗j ai∗j

j=1

j=1

where pi∗j = perimeter of patch ij in terms of number of cell surface; ai∗j = area of patch ij in terms of number of cells; Z = total number of cells in the landscape Number of patches (NP): The number of patches (NP) of a particular patch type is a simple measure of the extent of subdivision or fragmentation of the patch type. Although the number of patches in a class may be fundamentally significant to several ecological processes, it often has limited interpretive value because it conveys no information about the patches’ area, distribution, or density. The Number of patches is probably the most valuable, however, as the basis for computing other, more interpretable metrics. The Number of patches equals the Number of patches of the corresponding patch type. Its value ranges from 1 and above (Jaeger 2000). NP = n i where ni = number of patches in the landscape of patch type i. Patch density (PD): Patch density (PD) is a limited but the fundamental aspect of landscape patterns. Patch density has the same essential utility as the Number of patches as an index, except that it expresses the Number of patches on a per unit area basis that facilitates comparisons among landscapes of varying sizes. Of course, if the total landscape area is held constant, then patch density and the Number of patches convey the same information. Its measuring unit is Number per 100 hectares. PD equals the corresponding patch type number of patches divided by the total landscape area multiplied by 10,000 and 100 to convert hectares (Jaeger 2000). PD =

ni (10000).(100) A

where ni = number of patches in the landscape of patch type i; A = total landscape area. Contagion index: Contagion is inversely related to edge density. When edge density is very low, for example, when a single class occupies a large percentage of the landscape, contagion is high and vice versa. Its range is 0–100. 0 reveals the patch type is highly disaggregated and interspersed, and 100 rely on maximally aggregated patches (O’Neill et al. 1988). ⎡ CONTAG = ⎣1 +

m m  i=1

k=1

    ⎤ Pi∗ mgik gik ∗ ln Pi∗ mgik gik k=1 k=1 ⎦(100) 21n(m)

238

P. Barman and Sk. Mustak

where Pi = proportion of the landscape occupied by patch type i; gik = number of adjacencies between the pixel of patch types I and k based on the double count method; m = number of patch types present in the landscape. Aggregation index: The aggregation index is calculated from an adjacency matrix at the class level. At the landscape level, the Index is computed simply as an areaweighted mean class aggregation index, where each class is weighted by its proportional area in the landscape. This Index is scaled to account for the maximum possible number of like adjacencies given any landscape composition. It measures unit in percentage. Value of this index ranges from 0 to 100. AI value is 0 when the patch types are maximally disaggregated. AI increases as the landscape is increasingly aggregated and equal to 100 when the landscape consists of a single patch (He et al. 2000).  AI =

m   i=1

gii max − gii

  pi (100)

where, gii = no of like adjacencies between pixels of patch type i based on the single count method; max − gii = maximum number of like adjacencies between pixels of patch type i based on the single count method; pi = proportion of landscape comprised of patch type i. Landscape shape index (LSI) The landscape shape index provides a standardized measure of total edge or edge density that adjusts for the size of the landscape. It has a direct implementation in contrast to the total edge, which is meaningful relative to the size of the landscape. It ranges from 1 and above. LSI = 1 when the landscape consists of a single square patch. LSI increases without limit as the landscape shape becomes more irregular and the length of the edge within the landscape increases (Patton 1975). LSI =

0.25E ∗ √ A

where, E* = total length of the edge in landscape includes the entire landscape boundary and some or all background edge segments; A = total landscape area Patch density Patch density is a limited but fundamental aspect of landscape patterns. Patch density has the same basic utility as the Number of patches as an index, except that it expresses the number of patches per unit area on the basis that facilitates comparison among landscapes of varying sizes. Its measuring unit is Number per 100 ha (Jaeger 2000). PD =

N (10000) · (100) A

where, N = total no of patches in the landscape; A = total landscape area.

11 Assessing Urban Compactness Using Machine Learning and Earth …

239

Diversity indices Shannon’s diversity index: Shannon’s diversity index: Shannon’s diversity index is a popular measure of diversity in community ecology to landscape analysis. Shannon’s Index is more sensitive to rare patch types than Simpson’s diversity index. SHDI value exists at 0 and above. SHDI = 0 when the landscape contains only a patch meaning no diversity. SHDI increases as the no of different patch types increases, and the proportional area distribution among patch types becomes more equitable (Shannon and Weaver 2009). SHDI = −

m 

( pi ∗ lnpi )

i=1

where pi = proportion of the landscape occupied by patch type i. Simpson’s Evenness index Simpson’s evenness index is expressed such that an even distribution of area among patch types results in maximum evenness. As such, evenness is the complement to dominance. Its value ranges are 0–1. SIEI = 0 when the landscape contains only one patch and approaches 0 as the area distribution among the different patch types becomes increasingly uneven. SIEI = 1 when the area distribution among patch types is perfectly even, i.e., proportional abundance is the same (Simpson 1949). m P2 1 − i=1 1 i SIEI = 1− m where Pi = proportion of landscape occupied by patch type I; m = number of patch types present in the landscape excluding the landscape border if present.

11.3 Results and Discussion 11.3.1 Land Use and Land Cover The Land use and land cover (LULC) map has been prepared for urban footprint analysis during 1990, 2001, 2011, and 2021 which provides land use and cover characteristics of the area (Fig. 11.3). Built-up developed in KMA over different decades is shown in Fig. 11.4. The built-up area has been increasing rapidly. In 1990, the built-up area was 39,219 hectares, and it reached to 47,843 ha in 2001, which establishes the high spreading of built-up growth. In 2011 the built-up area was 48,302 ha, and it reached 66,964 ha in 2021. The built-up area increased tremendously in the last decades (Fig. 11.4). The classification accuracy of the different land use and land cover classes is given in Table 11.1. The built-up area from the land use

240

P. Barman and Sk. Mustak

and cover map has been extracted using spatial metrics for the built-up compactness index.

11.3.2 Spatial Metrices to Assess Urban Compactness Several metrics have been calculated to measure the compactness in Kolkata metropolitan area, including area, shape, aggregation, and diversity in patch, class, and landscape level. Area Metrics In area metrics, two metrics have been calculated, one is a class area (CA), and another is the largest patch index (LPI) (Figs. 11.5, 11.6, Table11.2). The built-up area has been increasing in KMA day by day. This increasing rate has been high in the last decade, and the minimum increases from 2001 to 2011. In 1990 the total built-up area was 39218.47 ha, while in 2021, it reached 66978.63 ha. This increasing rate is so high basically from 2011 to 2021. This highly increasing building occurs in peri-urban areas and creates sprawl development (Figs. 11.5, 11.6, Table 11.2). Another important matrix is the largest patch index (LPI). In the built-up class, the LPI has been increasing. In 1990 LPI was 9.2131, which it reached 29.5218 in 2021. This value reveals that the built-up patches will be larger and become dominant. Small patches of built-up have been decreasing. That means the built-up area will be more compact than in previous years. Shape Metrix The shape and contiguity indexes are essential to measuring an area’s compactness. The shape Index (SI) has been calculated in patch-level analysis. Its value ranges from 0 to 1. SI value toward 1 indicates that the shapes are going to be regular. The continuity index 0 shows the one pixel or patch, and towards 1, it increases the connectedness of patches in a class. In 1990 the connectedness was high, and it decreased in 2001 and 2011, and now this connectedness is increasing. The connectedness of patches increases compactness. Aggregation Matrices Class level matrices The flowing metrics at class-level are explained below (Table 11.3). A very important metrix is to measure the compactness at KMA is the class-level Aggregation Index (AI) measure. This metrix relies on patch compactness or disaggregating. A 0 value indicates that the patch is segregating, and when its value goes toward 100, the patches will be compact. In KMA, the AI value was low in 1990 and increased till 2011; in 2021, this value decreased. It reveals that the patches reach compactness till 2011, and now the patches are starting to segregate. The landscape shape index is a standardized measure of the edge density; when the LSI value increases, the

11 Assessing Urban Compactness Using Machine Learning and Earth …

Fig. 11.3 Land use and land cover of KMA

241

242

P. Barman and Sk. Mustak

Fig. 11.4 Built-up development from 1990 to 2021 Table 11.1 LULC classification accuracy (1990–2021)

Year

Overall accuracy

Kappa efficiency

1990

81.02

0.70

2001

86.48

0.78

2011

92.69

0.85

2021

89.97

0.84

11 Assessing Urban Compactness Using Machine Learning and Earth …

243

class area

Fig. 11.5 CA metrix 80000 70000

Hectare

60000 50000 40000 30000 20000 10000 0 1980

1990

2000

2010

2020

2030

Year class area

Largest patch Index

Fig. 11.6 LPI metrix 35

Percentage

30 25 20 15 10 5 0 1980

1990

2000

2010

2020

2030

Year Largest patch Index

Table 11.2 Area metrix of built-up class

Year

Class area

1990

39218.47

Largest patch index 9.2131

2001

47842.74

12.8241

2011

48301.92

15.7312

2021

66978.63

29.5218

landscape shape becomes more irregular. In KMA the LSI value is decreasing day by day than the previous year, which indicates that the landscape shape of that area becomes regular and trends to compact. One class-level aggregation metric to measure compactness is called the patch cohesion index, which is very important. This Index measures the conceitedness of patches’ physical connectedness and is more compact in patch type. This Index measures the compactness of patches concerning the distinguished class. It’s a relative measure. In KMA, the patch cohesion index is very high and is increasing daily.

244

P. Barman and Sk. Mustak

Table 11.3 Aggregation metrics in class-level Year

Aggregation index (AI)

Landscape shape index (LSI)

Patch cohesion index

Number of patches (NP)

Patch density (PD)

Euclidean nearest neighbor (ENN)

1990

90.6288

62.7979

99.1246

4719

2.6279

140.7588

2001

93.1212

51.0569

99.4602

3646

2.0301

153.1847

2011

93.6342

47.5457

99.5846

3421

1.9048

157.8587

2021

93.4857

57.1107

99.7831

4289

2.3884

139.6647

In 1990 the cohesion was 99.1246, and in 2021 it is 99.7831. This value indicates that the built-up class patches are more compact and physically connected. The value increases, meaning this class’s patch type is aggregated and clumped; the distribution is also closed and physically connected. The number of patches and patch density are important metrics to measure the compactness of an area. The Number of patches is decreasing in KMA daily, which indicates that the respective class is going for compact development. However, the level of patches has been increasing in the last decade. Patch density is decreasing, which indicates the development is less compact and is going for sprawl development. ENN metrics are important to measure the compactness of a perspective class in an area. It’s a simple measure, and it’s the shortest distance between two patches in one class from edge to edge. In KMA, ENN has been increasing till 2011, and in the last decade, this has decreased. This metric demarcates that till 2011, the patch distance of the built-up area is more than the previous year. The largest distance of patches indicates that the development is isolated. In KMA, the built-up area was isolated, and now the patch distance is decreasing, which relies on the built-up development becoming compact. Landscape-level matrices To measure compactness, landscape-level matrices are significant (Figs. 11.7, 11.8, 11.9, 11.10; Table 11.4). Among these landscape level matrices contagion, Metrix is one of the best to measure compactness in an area. In KMA, this value exists in the medium; it reveals that the landscape of this area is normally distributed. The value increases, indicating that the landscape will aggregate develop from segregate development. But the increasing rate is low. So, for compact landscape development, this value has to increase highly. Aggregated Index in KMA reveals that in 2001 the patch compact was high compared to other decades. Generally, the value is going up, which defines the landscape to aggregate development to a one-patch development. LSI value will decrease so that the landscape patches will be irregular to regular development. Patch density has been going high in the last decade, which indicates that the development turns compact. But in 2011, the development was less compact than in other years comparatively.

11 Assessing Urban Compactness Using Machine Learning and Earth …

245

Aggregation metrix percentage

Fig. 11.7 Aggregation metrix

96 94 92 90 88 86 84 82 1990

2001

2011

2021

Year class level

landscape level

Landscape shape index (LSI)

Fig. 11.8 LSI metrix 100

percentage

80 60 40 20 0 1990

2001

2011

2021

Year landscape level

class level

Patch density 20

No per 100 hectarea

Fig. 11.9 Patch density metrix

15 10 5 0 1990

2001

2011

2021

Year class level

landscape level

Relationship Between Class Level And Landscape Level Metrices Both the class-level and landscape-level metrics show that built-up development approaches towards compact development at a slower rate (Figs. 11.7, 11.8, 11.9, 11.10).

246

P. Barman and Sk. Mustak

Fig. 11.10 Patch cohesion metrix

Patch cohesion index 100

percentage

99.5 99 98.5 98 97.5 97 1990

2001

2011

2021

Year class level

landscape level

Patch density

Patch cohesion index

Table 11.4 Aggregation index in landscape-level Year

Contagion index

Aggregation index

Landscape shape index (LSI)

1990

44.8073

86.9554

94.7045

15.7471

98.2147

2001

46.4725

88.1523

86.1998

15.3627

98.702

2011

47.0555

87.9708

87.4735

12.5414

98.7096

2021

47.9491

87.1776

93.0281

16.2465

98.9845

Diversity Indices Shannon’s diversity index is one of the important landscape matrices to measure diversity in landscape analysis. In KMA, the landscape diversity was comparatively higher in 1990 than in any other decade. The SHDI value is decreasing, indicating that the diversity is decreasing in landscape-level analysis, and the similarity will increase, which defines compact development (Figs. 11.11, 11.12; Table 11.5).

Shannon’s diversity index (SHDI)

Fig. 11.11 SHDI metrix

Information

1.64 1.62 1.6 1.58 1.56 1.54 1.52 1.5 1980

1990

2000

2010

Year

2020

2030

Shannon’s diversity index (SHDI)

11 Assessing Urban Compactness Using Machine Learning and Earth …

247

Simpson’s Evenness index (SIEI)

None

Fig. 11.12 SIEI metrix

0.92 0.91 0.9 0.89 0.88 0.87 0.86 1980

1990

2000

2010

2020

2030

Year Simpson’s Evenness index (SIEI)

Table 11.5 Diversity metrices

Year

Shannon’s diversity index (SHDI)

Simpson’s evenness index (SIEI)

1990

1.6226

0.9146

2001

1.5919

0.9043

2011

1.5816

0.8995

2021

1.5148

0.8702

Simpson’s evenness index is the most useful measure for the dominance analysis. In the KMA area, the landscape dominance was more in 1990, and this dominance has been decreasing daily.

11.4 Conclusions Spatial matrices are essential to landscape analysis in an urban area. To measure the compactness in an urban area, spatial metrics play a significant role. Kolkata metropolitan area is an urban agglomeration with a rapidly growing population and urban footprint. The city is increasing daily, mainly in the peri-urban area where the built-up area has grown tremendously. The compact urban analysis has been taken in the following manner. First, the land use and land cover map were prepared using the machine learning algorithm, and then only the built-up of various years overlapped to visualize the built-up extent in KMA. Later, different spatial matrices were measured to assess the compactness in an urban area. The result shows that the built-up area is growing, especially in the last two decades. The built-up development mainly occurs in peri-urban areas. In KMA, the central Kolkata city is highly congested, but most of the development is sprawling in the peri-urban area. This sprawling development creates many problems in cities. In 2011, the development was mainly sprawling, but the development during the last decades started with compact development. For sustainable urban development, compact built-up is very necessary. Sustainable urban development needs policies and planning models like Transit-oriented, inclusive,

248

P. Barman and Sk. Mustak

and smart city development. The compactness analysis would help policymakers and planners to develop the city sustainably to support sustainable development goal (e.g., SDG-11).

References Abdullahi S, Pradhan B, Jebur MN (2015a) GIS-based sustainable city compactness assessment using integration of MCDM, Bayes theorem and RADAR technology. GeOcarto Int 30(4):365– 387. https://doi.org/10.1080/10106049.2014.911967 Abdullahi S, Pradhan B, Mansor S, Shariff ARM (2015b) GIS-based modeling for the spatial measurement and evaluation of mixed land use development for a compact city. Gisci Remote Sens 52(1):18–39. https://doi.org/10.1080/15481603.2014.993854 Atmadja A, Bogunovich D (n.d.) Shaping compact cities for liveability, affordability and sustainability (L-A-S). A Comparative Assessment of TODs in Jakarta and Auckland. In: 55th ISOCARP World Planning Congress Jakarta-Bogor, Indonesia. International Society of City and Regional Planners Forman RTT, Godron M (1986) Landscape ecology. Wiley, New York, 619 pp—Google Search (n.d.) Retrieved Feb 26, 2023 He HS, Dezonia BE, Mladenoff DJ (2000) An aggregation index (AI) to quantify spatial patterns of landscapes. Landscape Ecol 15:591–601 Herkert JR (1999) Status and habitat area requirements of the veery in Illinois. NCASI Tech Bull 781 I, 233 Jaeger JAG (2000) Landscape division, splitting index, and effective mesh size: new measures of landscape fragmentation. Landscape Ecol 15:115–130 Kolkata Metropolitan Development Authority. (n.d.). Retrieved June 8, 2022, from http://www.kmd aonline.org/page/cms/map_of_kma_c1fed8 Kotharkar R, Bahadure P, Sarda N (2014) Measuring compact urban form: a case of Nagpur city, India. Sustainability (switzerland) 6(7):4246–4272. https://doi.org/10.3390/su6074246 LaGro J (1991) Assessing patch shape in landscape mosaics. Photogram Eng Remote Sens 57:285– 293—Google Search (n.d.). Retrieved Feb 26, 2023 Lan T, Shao G, Xu Z, Tang L, Sun L (2021) Measuring urban compactness based on functional characterization and human activity intensity by integrating multiple geospatial data sources. Ecol Ind 121.https://doi.org/10.1016/j.ecolind.2020.107177 Lehmkuhl JF, Raphael MG (1993) Habitat pattern around northern spotted owl locations on the Olympic Peninsula, Washington. J Wildl Manage 57:302–315. Google Search (n.d.). Retrieved Feb 26, 2023 McGarigal K, McComb WC (1995) Relationships between landscape structure and breeding birds in the Oregon coast range. Ecol Monogr 65(3):235–260. https://doi.org/10.2307/2937059 Min C, Suxia L, Liang Y (n.d.) Calculation and analysis of urban compactness using an integrated ARCGIS tool. LNEE 144 Mubareka S, Koomen E, Estreguil C, Lavalle C (2011) Development of a composite index of urban compactness for land use modelling applications. Landsc Urban Plan 103(3–4):303–317. https:/ /doi.org/10.1016/j.landurbplan.2011.08.012 O’Neill RV, Krummel JR, Gardner RH, Sugihara G, Jackson B, DeAngelis DL, Milne BT, Turner MG, Zygmunt B, Christensen SW, Dale VH, Graham RL (1988) Indices of landscape pattern. Landscape Ecol 1:153–162. Google Search (n.d.). Retrieved Feb 26, 2023 Patton DR (1975) A diversity index for quantifying habitat “edge”. Wildl Soc Bull 3:171–173. Google Search (n.d.). Retrieved Feb 26, 2023

11 Assessing Urban Compactness Using Machine Learning and Earth …

249

Rahman MH, Islam MH, Neema MN (2022) GIS-based compactness measurement of urban form at neighborhood scale: the case of Dhaka, Bangladesh. J Urban Manage 11(1):6–22. https://doi. org/10.1016/j.jum.2021.08.005 Schumaker NH (1996) Using landscape indices to predict habitat connectivity. Ecology 77(4):1210– 1225. https://doi.org/10.2307/2265590 Shamkhi Al-Khafaji, AJ, Abdul-Majeed Al-Salam N (2018) Measurement of urban sprawl and compactness characteristics Nasiriyah City-Iraq as case study. Int J Civil Eng Technol 9(9). http://iaeme.comhttp://iaeme.com/Home/issue/IJCIET?Volume=9&Issue=9 Shannon CE and Weaver W (2009) The mathematical theory of communication Sim S (2011) Measuring urban sprawl and compactness: case study Orlando, USA. https://www. researchgate.net/publication/307228230 Simpson EH (1949) Measurement of diversity [16]. Nature 163(4148):688. https://doi.org/10.1038/ 163688A0 Stathakis D, Tsilimigkas G (2015) Measuring the compactness of European medium-sized cities by spatial metrics based on fused data sets. Int J Image Data Fusion 6(1):42–64. https://doi.org/ 10.1080/19479832.2014.941018 Steurer M, Bayr C (2020) Measuring urban sprawl using land use data. Land Use Policy 97.https:/ /doi.org/10.1016/j.landusepol.2020.104799 Willams N (2001) Achieving sustainable urban form. Land Use Policy 18(2):202. https://doi.org/ 10.1016/S0264-8377(01)00010-2 Yang B, Xu T, Shi L (2017) Analysis on sustainable urban development levels and trends in China’s cities. J Clean Prod 141:868–880. https://doi.org/10.1016/J.JCLEPRO.2016.09.121

Chapter 12

Analysis of Ecological Vulnerability Behind the Land Conversion from Agriculture to Aquaculture of Purba Medinipur District in West Bengal, India Manishree Mondal , Ramu Guchhait, and Sk. Mustak

Abstract The illegal, unplanned, and forceful conversion of fertile cultivable lands into brackish water aquaculture is now becoming a serious ecological threat at the cost of instant economic profit for the last decade in Bhagawanpur-II Block-, Purba Medinipur district, West Bengal. This study mainly focused on the impact of the mushrooming of these unwanted fishing ponds and ecological vulnerability on land and concerned people. It tried to unveil how short-term economic gain destroyed the long-term indigenous means of sustainable livelihood of the local residents. Extensive relevant-literature review, change detection of LULC using Rs-GIS-GPS techniques, ground truth verification, intensive perception survey using F-G-D and P-A-R approach, and ecological cost–benefit analysis techniques etc. were used to reach the goal of the research. The survey showed that the aquaculture area increased by 1.31 km2 (8.65%) from 2011 to 3.56 km2 (23.52%) in 2021. Mass reduction in cultivable and pastoral land, extinction of native species, changes in chemical characteristics of soil as well as water quality, economic stratification, vandalism, and social conflicts were identified. What people thought of as their economic strength was their ecological vulnerability. This research tried to set a blueprint for future micro-level planning and development so that, this vulnerability will be converted into their strength. Keywords Conversion · LULC · Cost–benefit · Ecological · Vulnerability · Strength

M. Mondal (B) · R. Guchhait Department of Geography, Midnapore College (Autonomous), Midnapore, West Bengal, India e-mail: [email protected]; [email protected] Sk. Mustak Department of Geography, School of Environment and Earth Sciences, Central University of Punjab, Bathinda, Punjab, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_12

251

252

M. Mondal et al.

12.1 Introduction The fishery has a significant influence on the local as well as national economy because it creates job opportunities for the rural people and also helps in the socioeconomic development of a region. It has been recognized as a powerful source of income and employment generator (Long et al. 2005). It is also one of the cheapest sources of animal protein besides its capacity to earn foreign exchange. This sector is important for poverty alleviation and gender empowerment (Burgi et al. 2004). To protect the wetlands for the sake of the environment, inland fishing along with agriculture can be done and this practice can contribute something worthy to the socioeconomic development particularly for developing countries (Paul and Chakraborty 2016). Shrimp culture is now converted into a leading industrial activity in many shrimpgrowing countries of the world (Lebel et al. 2002). High profitability and generation of foreign exchange have provided the major driving forces in the global expansion of shrimp farming (Primavera 1993, 2006; Hein 2000) which leads to rapid conversion of different types of land into commercial shrimp ponds. Over the recent decades, aquaculture in India is now flourishing as a very high profit-making industry. Inevitably, it now replaces the life-sustaining traditional agriculture both in intensive and extensive types. The main causes are the increasing internal and foreign demands (Abraham and Sasmal 2009). The densely populated coastal areas particularly tropical Asia and Latin America have experienced this havoc conversion from agriculture to aquaculture (Gujja and Finger-Stitch 1996; Dewalt et al. 1996; Flaherty et al. 1999). The growth of fishery, directly and indirectly, affects the agricultural land classes in two ways such as, it has decreased the area of the agricultural land and, the productivity is affected due to continuous leakage, seepage, and overflow of water to the nearby agricultural fields (Ojha and Chakrabarty 2018). Furthermore, this conversion leads to the salinization of groundwater and agricultural land, biodiversity loss and social conflicts (Rahman et al. 2013; Ahmed and Glaser 2016; Jayanthi et al. 2018). Abandonment of large fish ponds is also a great issue for the local ecological system (Páez-Osuna 2001a, b; Ahmed and Diana 2015; Proisy et al. 2018). The highly profitable shrimp culture rapidly encroaches into the highly fertile paddy fields which creates a contradiction in the mind of the farmers, cultivators, fishermen, and environmentalists (Kagoo and Rajalakshmi 2002). In Purba Medinipur District, the central inland part of the district includes Egra I and II, Patashpur I and II, Bhagwanpur I and II Chandipur, Khejuri I, Contai III blocks which are very favourable for paddy cultivation in terms of natural environmental setup like fertile soil, monsoon climate, irrigation facilities and socio-economic environments like choice, demand and labour. This conversion of fertile rice fields into shrimp, prawn, and carp fish culture with or without legal permission is now very common. Therefore, the highly productive cropping land is now under threat (Das and Mondal 2020). The conversion is going on ignoring the land degradation issues and proper guidance although there are strict environmental rules and regulations. Such a type of

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

253

conversion has a great influence on the environment as well as society (Lambin and Meyfroidt 2010). Several authors have already expressed doubts about the adverse impacts of this type of land conversion on society, the economy and the environment (Hein 2002; Páez-Osuna et al. 1999; Mitra and Santra 2011; De et al. 2014). This type of land conversion might result in income disparities between different communities living in those villages (Akteruzzaman 2005; Akteruzzaman and Jaim 1999). This conversion of fertile rice fields into shrimp, prawn, vennamin and carp fish culture with or without legal permission is now very common. Therefore, the highly productive cropping land is now under threat. The conversion is going on ignoring the land degradation issues and proper guidance although there are strict environmental rules and regulations. Such a type of conversion has a great influence on the environment as well as society. Several authors have already expressed doubts about the adverse impacts of this type of land conversion on the environment (Hein 2002; Paez-Osuna et al. 2003; Chowdhury 2006; Azad et al. 2009; Mitra and Santra 2011). As most of the aqua cultural farms are intermittently situated with agricultural lands, seepage of salt water into agricultural lands, aqua cultural pond overflow, and leaching from sludge piles during rainfall are well documented (Mitra and Santra 2011). Salinization badly affects soil nutrient balance and productivity. Intensive aquaculture in many Asian countries such as Taiwan, China, the Philippines, Thailand, and Indonesia etc. has caused land resource degradation and water quality deterioration. Various environmental threats have been categorized in the following section. Firstly, the fertile topsoil was extracted permanently for ‘bheri’ (fish ponds) making which created massive ‘bio-physical’ losses and greater possibilities of soil erosion. Secondly, The farmers use salt water, an excessive amount of lime, potash, urea, and aqua-chemicals i.e. banned and restricted drugs, (antibiotics, pesticides etc.) to control the growth and development of the fish species only to earn more and more profit so, there is a large concern about its adverse effects of the wastewater on and in the surrounding cultivated land, water bodies, groundwater and human health. Thirdly, rice cultivation is not so-called less economical to the farmers because they receive so many direct benefits (like money by selling surplus food grain, food for the family, fodder for a pet animal (grass, straw, husk), fuel for cooking (straw, husk), seasonal fish as a source of protein, leafy vegetables) and indirect benefits (like milk, eggs and meats from animals and birds, money by selling of them). Finally, the rootless farmers who will lose their jobs have to migrate from the rural area to the outside. Besides it, various socio-economic impacts—like marginalization of rural poor, breakdown of traditional livelihood support system, food insecurity, and transfer of land and wealth to local and national elites are frequent. Hein (2000) pointed out that the high profitability of shrimp farming is the major driving factor of land conversion from mangroves into shrimp ponds. He tried to warn the world that the cyclone protection belt of coastal areas (mangroves) is reducing day by day due to the destruction of mangroves. To prove this issue, he showed that 80% area of the total Godavari delta had been converted into shrimp ponds. Highly

254

M. Mondal et al.

profitable shrimp farming can improve the Indian economy in a short time if its social and environmental costs are intolerable. Páez-Osuna (2001a, b), basically noticed on environmental impact of shrimp aquaculture, including causes and environmental mitigating action. Reduction of biodiversity, loss of habitats, coastal erosion, salinization of soil, water logging, biological pollution, disease, and inverters in the wild population are the major impacts, pointed out by Páez-Osuna. He concluded that continual expansion of shrimp culture usually results in catastrophic collapses due to viral and bacterial diseases, which obstruct the life span of shrimp (7–15 years) and that’s why the no. of abundant shrimp ponds is increasing day by day. According to him sustainable development of shrimp may be attainable, by taking environmentally non-degrading, technically appropriate, economically viable and socially acceptable practices.

12.2 Background Shrimp culture is expanding in low-lying coastal areas of the world and the research emphasizes its development and positive and negative impacts on society as well as the environment. Primavera (1997) focused on basically the impacts of commercial shrimp farming on the coastal mangrove of Indonesia. He assumed that the conversion of mangroves into shrimp ponds results in a vigorous change in the natural environment, loss of biodiversity and salinization of agricultural land due to saltwater intrusion making land uncultivable compelling the poor farmers to sell their lands. He pointed out that the life span of fish ponds doesn’t exceed 5–10 years in the case of intensive farming, due to self-pollution and diseases. The abrupt growth of shrimp culture has severe negative environmental effects, like loss of biodiversity, imbalance of the ecosystem, pollution of land and water etc. (Hossain et al. 2013). They found socio-economic impacts including exile of indigenous livelihood system, land insecurity, marginalization, rural unemployment, social instability, and conflict in the wake of shrimp culture (Hossain et al. 2004; Huang et al. 2017). Inundation of land by saline water for a long period leads to its percolation into the surrounding soil resulting in the alteration of soil chemistry (Mitra and Santra 2011). It has also been found that there was a direct relationship between soil salinity and the distance of sampling points from shrimp ponds and a buffer of 60 m around those ponds had been highly salinized. This type of aquaculture affects the food security particularly of poor people because of salt-encrustation (Clay 1996), seepage and leakage of chemicals and wastage from fish ponds badly affect the yield of nearby agricultural fields (Hein 2000; Shiva and Karir 1997). The rapid conversion of agricultural land into aquaculture has been already observed in the coastal district of West Bengal. The present study area is not an exception. This may put a severe threat to the available resources as well as to the environment. Earlier, it has been found that only low-lying, waterlogged areas which were not suitable for agriculture are mainly converted into aquaculture. Now due

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

255

to the greediness of a few land mafias, the fertile agricultural lands are converted into aquaculture overnight. Local people are also not against these practices at the first time because they have a handsome amount of money in their hands. Later on, they and other people who are against this practice realize the fact that although this practice is economically profitable ecologically it is devastating, particularly for future generations. The fishing ponds are now spreading like mushrooms capturing the agricultural lands by hook or by crook. Keeping all these in mind the present study focused on unveiling how shortterm economic profit destroys the long-term sustainable livelihood of the local rural inhabitants. What they perceive as their economic strength is the vulnerability of their indigenous sustainable livelihood. The specific objectives set for this research were: 1. To find out the main causes for the conversion of agricultural land to commercial fisheries. 2. To detect the rate of change in land use and land cover (LULC) by using RS-GIS techniques 3. To assess the economic benefit using cost–benefit analysis for measuring the strength of local people. 4. To examine the ecological threats due to this conversion by investigating soil, water and biodiversity to measure the vulnerability of local residents. 5. To set a blueprint for future micro-level planning and development so that their vulnerability will be converted into their strength

12.3 Study Area The astronomical location of Bhagwanpur-II Block of Purba Medinipur district in West Bengal, India is from 21.35°N to 21.53°N latitude and 87.39°E to 87.52°E longitude. It is bounded by Bhagwanpur-I and Chandipur Blocks in the north, NandigramII and Khejuri-I Blocks in the east, Contai-III and Egra-II Blocks in the south and Pataspur-I and Pataspur-II CD Blocks in the west. It is 35 km away from Tamluk, the district headquarter of Purba Medinipur District. Ten villages of Bhagwanpur-II Block, where the land conversion process is rapidly going on have been selected for intensive field survey for ground truth verification (Sau et al. 2018) (Fig. 12.1). This block is a part of the Indo-Gangetic plain and the relief is entirely flat. The climate is tropical monsoon type with moist hot summer and short cool winter. The soil is composed of younger fertile alluvium of medium to fine texture. The population statistics of the selected villages are given below (Table 12.1). The main occupation of the residents was cultivation. Paddy has stapled cultivated crops. The soil, climate and irrigation were well supported to cultivate paddy a minimum of two times a year. This life-supportive economic activity was gradually shrinking due to the recent mushrooming of shrimp cultivation (Plate 12.1).

256

M. Mondal et al.

Fig. 12.1 Study area

12.4 Materials and Methods This research was mainly based on a primary database obtained from an intensive perception survey, data explored from satellite images, topographical sheets, and laboratory experiments. Few supportive secondary-level data had been used from census handbooks and statistical reports.

12.4.1 Change Detection of Land Use Change detection of LULC had been assessed between 2011 and 2021 using union overlay analysis (Jayanthi et al. 2018). This detection matrix depicted the land use change transformation including trajectories.

1.05

0.62

251

Jiagodi

0.44

338

North Khasmulda

0.68

0.31

248

249

294

Ramchawk

Ranichawk

Raghunath chak 631

10

203

45

506

39

120

95

142

631

45,285

No. of households

2800

38

854

197

2387

183

472

394

613

2800

192,162

Total population

Population

1436

23

423

100

1999

87

238

201

304

1436

99,060

Male

Source District Statistical Handbook, Purba Medinipur District, 2011, and Census of India, 2011

4.23

2.9

0.45

247

339

Kantapukuria

Natagachia

0.21

252

337

Kharinet

South Khasmulda

4.23

294

Nayabasan/ Dhaipukuria

7251.9

Area (km2)



JL. No.

Population data

Bhagawanpur-II Block

Block/Villages

Table 12.1 Population statistics

1364

15

431

97

1188

96

234

193

309

1364

93,102

Female

662

123

1256

437

823

416

2248

635

584

662

1100

Population density (km2 )

140

02

43

10

119

09

24

20

31

140

538

No. of respondents

12 Analysis of Ecological Vulnerability Behind the Land Conversion … 257

258

M. Mondal et al.

Plate 12.1 Fishing ponds: a Dhaipukuria village, b Ramchawk village (Source Field survey)

12.4.2 Image Processing The satellite images had been collected from the Google Earth and Satellite Image (Sensor: Landsat −5, TM, No. of Bands: −7, Spatial Resolution: −30 m in (scale are 1:30000 and 1:40000) between 2011 and 2021 from December. These maps had been geo-referenced first and after then the land use classification was done (Richards and Jia 2006; Campbell et al. 2007). To identify land cover in the proper location, the images needed to be geo-referenced accurately. A total 200 of well-distributed points were marked and ground control points (GCPs) were put by using GPS. UTM projection system was used for georeferencing. These referenced images were classified into 5 land cover classes by the on-screen digitization method. Image classification and geo-referencing were done in Arc GIS-environment 10.7. For testing the accuracy level of classification, a field visit was done where-samples of each land cover class were randomly checked (Disperati and Virdis 2015; Rwanga and Ndambuki 2017). The total processing is depicted in the following flow chart (Fig. 12.2).

12.4.3 Soil Sample Collection For the collection of soil samples, those fish ponds were selected which were having aquaculture for at least the past five years and were adjoining rice fields. The total number of soil samples collected was 300, 30 for each village. The samples were collected from just inside the pond (0 m distance) to correspondingly at 50 m internal (50, 100, 150 m, and so on). These samples were collected from the month of late November to late February because at this time the aquaculture season was over. Samples were collected from 20 cm below from the soil top layer of 500 gm amount and kept in new polybags with the id number of sample sites. The experiments of

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

259

Satellite data

Fish pond identification

Image processing

Ground control points from topo sheet

Image classification

Ground truth verification

Ground truth verification Field verification

Google earth verification

Test of accuracy

Post classification

Land use and cover detection

Change detection Estimation of rate of change

Fig. 12.2 LULC change detection procedure

soil chemical properties were done as per the recommendation of Indian Agricultural Research Institute (IARI 2014) and electrical conductivity were tested in the department of Chemistry and Geography of Midnapore College (Autonomous).

12.4.4 Water Sample Collection To assess the water quality, a total of 100 samples of one litre each were collected. Fifty samples were collected from the inside of the fishing ponds and another fifty samples were taken from at least 200 m away from the fisheries and kept in pet jars.

260

M. Mondal et al.

12.4.5 Statistical Analysis Few qualitative statistical techniques such as Mean, Standard Deviation (SD), Coefficient of Variation (CV) and Regression Analysis etc. were used.

12.4.6 Perception Analysis The most important part of the methodology of this study is taking interviews to understand the perception level of local residents about ecological perspectives of these land conversion scenarios. Face-to-face personal interviews, focused group discussion, (FGD) and P-A-R (Problem-Action-Result) techniques were used for this purpose. PAR format is an effective method to present the responses during an interview, particularly in open-ended questions (Reason et al. 2001). Ten sample villages were selected based on the recent large growth of fishing ponds. The numbers of sample respondents were 538. The stratified random sampling method had been applied to choosing target respondents. The total procedure is presented in Fig. 12.3. This research had been done based on the exploratory primary database through intensive field visits in three phases 5–10 December 2020 (unlock phase 7.0), 15–20 March 2021 (unlock phase 10.0) and 20–25 June 2021 (unlock phase 13.0) with the help of semi-structured and open-ended questionnaire schedule of a pre-formulated set of questions. Three sets of F-G-D sessions had been arranged in three phases

Total population of the village (100%)

No. of respondents (5% of total 100%)

Respondents only engaged in fishery (ownership/rent basis) (33.33% of total5%)

Income strata (100% of total 33.33%)

Medium (33.33% of total 33.33%)

High (33.33% of total 33.33%)

Low (33.33% of total 33.33%)

Fig. 12.3 Sampling procedure

Respondents giving their agricultural land for fishery on rent basis (33.33% of total 5%)

Income strata (100% of total 33.33%)

Medium (33.33% of total 33.33%)

High (33.33% of total 33.33%)

Low (33.33% of total 33.33%)

Respondents engaged in cultivation of crops till now (33.33% of total 5%)

Income strata (100% of total 33.33%)

Medium (33.33% of total 5%)

High (33.33% of total 33.33%)

Low (33.33% of total 33.33%

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

(a)

261

(b)

Plate 12.2 Perception survey: a Madhakhali village, b Ranichawk village (Source Field survey)

of field visits to discuss the economical and ecological issues following the P-A-R approach. The perception of the respondents was analyzed by following the Likert” scale of perception measurement (Allen and Seaman 2007; Perla and Carifio 2007; Robbins and Heiberger 2011; Derrick and White 2017; Jovanovi´c and Lazi´c 2020) (Plate 12.2).

12.5 Result and Discussion The main aim of this paper was to investigate the actual scenario of economy and ecology behind the mushrooming of aquaculture in this study area. So, firstly, the economic scenario had been analyzed.

12.5.1 Ecological Impact Analysis For the assessment of the ecological impact of the mushrooming of aquaculture activities in the study area, the emphasis was given to land use and land cover change detections, laboratory experiments of the quality of soils particularly, cultivable soils; examination of the quality of water from various sources including potable water; the analysis of the status of biodiversity. Most importantly, the active perceptions of local residents on these aforesaid issues were taken to understand the actual reality.

12.5.2 Change Detection of LULC Land use land cover change: Agriculture is the dominant economic authority of the local residents. Five dominant LULC classes had been taken for change detection

262

M. Mondal et al.

which was shown in (Table 12.2) the significant change was found in brackishwater aquaculture and agriculture. The area under aquaculture increased nearly by about 15% in one decade which was quite alarming whereas the agricultural area decreased by 19% (Fig. 12.6). Mostly, aquaculture and settlement had invaded these agricultural fields. The waterbodies did not have any change. Dynamic land use change had been found mainly among aquaculture, and agriculture (Figs. 12.4 and 12.5). Aquaculture gradually replaced the fertile cropping lands. Even much freshwater aquaculture area was converted into brackishwater aquaculture. It was also found that few residents abandoned the freshwater fish ponds due to the low productivity and high cost of production and high price in the market. This assessment of this land use dynamics based on GIS analysis of bi-temporal satellite dates indicates the direction of change in the uneven, vigorous growth of aquaculture and conversion of different land use types. This unplanned growth invited various diseases which created tremendous problems not only in aquaculture but also in agricultural sectors also (Leung and Tran 2000). Small farmers could not able to fight against these diseases and got spoiled economically. On the other hand, this havoc growth of aquaculture increased the load of organic matter which led to self-pollution when it exceeded its carrying capacity. This will also cause disease outbreaks frequently and crop failure (Kautsky et al. 2000) (Fig. 12.6). Perception on Land Use Change: Group discussion sessions had been carried out to assess the perception of local inhabitants about the alarming land use change in their villages (Table 12.3). Their perception indicated that the encroachment of aquaculture in fertile cropping lands was instituted rapidly. They became afraid of the shrinkage of freshwater fish which was healthier than brackish water fishes and cropping diversity. The perception scenario had been reflected in the Table 12.3. Table 12.2 Detection of land use and land cover change from 2011 to 2021 Sl. No.

Land use land cover class

Area in km2 % of the Area in km2 area to (2021) (2011) total area

% of the area to the total area

1

Aquaculture

1.31

23.52

8.65

3.56

% of the area changed 14.87

2

Brick kiln

0.01

0.06

0.01

0.07

0.01

3

Cropland

9.39

62.10

6.53

43.16

−18.94

4

Rural settlement

1.18

7.79

1.79

11.85

4.06

5

Waterbody

0.25

1.64

0.25

1.64

0.00

Remarks

The total area of the study area is 15.12 km2

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

263

Fig. 12.4 LULC-2011 (Source Maps prepared from the superimposition of Administrative Maps and the Satellite Images (Sensor: Landsat −5, TM, No. of Bands: −7, Spatial Resolution: −30 m)

Fig. 12.5 LULC-2021 (Source Maps prepared from the superimposition of Administrative Maps and the Satellite Images (Sensor: Landsat −5, TM, No. of Bands: −7, Spatial Resolution: −30 m)

264

M. Mondal et al. Changes in LULC

70

% of area to total area in (2011)

60 50 40 30

% of area to total area (2021)

20 10 0

Aquaculture

Brick kiln

Cropland

Rural Settlement

Water body

-10

% of area changed (increase/decreased)

-20 -30

Fig. 12.6 Changes in LULC

Table 12.3 Perception on land use change Sl. No.

Perception on land use change

Strongly agreed (%)

Agreed (%)

Neither agreed nor disagreed (%)

Disagreed (%)

Strongly disagreed (%)

1

Decreasing agricultural land

17.5

39.75

14.25

25.5

3

2

Increasing aquaculture

27.5

72.5

0

0

0

3

Vanishing the 78.5 freshwater fishing

21.5

0

0

0

4

Decreasing the diversity of crop cultivation

17.5

1.5

1.0

0

80

Source Field survey

12.5.3 Soil Quality Assessment It was necessary to assess the soil quality because antibiotics and other therapeutic chemicals added to feed the fish had an immense effect on the soils as well as other organisms of surrounding areas of fish ponds. Detailed chemical properties such as pH , total nitrogen, potassium, phosphorous, organic carbon and electrical conductivity were measured thoroughly. Clear evidence of deterioration of soil quality had been found in this study area. Soil Salinity: Salinity is such an environmental parameter which can severely limit crop productivity. Salt concentration in the soil raises the salinity level of soil water

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

265

which can quickly sensitize the crops (Zeng and Shannon 2000). Moreover, the high salt concentration in soil water may cause the reverse water flow from crop roots to the soil. This effect can dehydrate the crops dangerously and causes a decline in yield or even crop death. Salinity can also reduce the nitrogen intake capacity of crops and nitrogen deficiency affects the growth and reproduction capability of cultivated crops. Saline soils are those which have a pH usually between 7.0 and 8.5 and ECe > 4 dS−1 (Richards 1954). Three hundred (300) soil samples were taken at the inner portion of fish ponds and away from fish ponds. Most importantly, it was found that the areas within five meters of fish ponds had high pH ranging from 8.25 to 9.25 (Table 12.4) and within twenty meters of the fish ponds, and saltwater draining channels, it had 8–8.5. The salinity of the soil decreased with the increase in distance from the saltwater fish ponds. The agricultural lands which were more than twenty meters had pH values of 7.0–7.5 respectively. This happened due to the continuous seepage of salt water to the nearby cultivable land. The rate of seepage was higher close to the fishery and draining channels. The presence of aquaculture in the vicinity was related to this enhancement of soil salinity in agricultural fields. Most of these fish ponds were situated in intermittent positions between agricultural fields. Significant correlations were found between Table 12.4 Soil salinity status Villages

Soil pH

Electrical conductivity (ECe -dSm¯1 )

At the fish ponds (0 m) N =5

Cultivated land (samples taken at 50 m intervals away from fish ponds to 250 m) N = 25

At the fish ponds (0 m) N =5

Cultivated land (samples taken at 50 m intervals away from fish ponds to 250 m) N = 25

Mean

Mean

Mean

Mean

Nayabasan/ Dhaipukuria

9.25

7.0

5.79

1.84

Jiagodi

9.75

7.0

7.23

1.59

Kharinet

8.50

6.5

7.74

2.87

South Khasmulda

8.75

7.5

5.65

2.54

North Khasmulda

8.75

7.5

5.75

1.92

Kantapukuria

8.75

7.0

5.62

1.73

Natagachia

8.75

6.5

5.78

2.29

Ramchawk

8.25

7.5

5.83

2.53

Ranichawk

8.50

7.0

7.19

1.98

Raghunath chak

8.75

7.5

4.62

3.11

Mean

8.80

7.1

6.12

2.24

Source Field survey

266

M. Mondal et al.

the soil salinity and the distance of the sampling point from the aquaculture farms. However, salinization trended up to a distance of 500 mt. from aquaculture was noticed. High salinity zones were situated mostly in the areas of higher-density of aqua farms. Due to seepage, leakage and poor farm management water reached the nearby agricultural land and affected the soil properties (Table 12.5 and Fig. 12.7). The increase in pH and salinity levels had an adverse effect on the productivity of paddy. An increase in soil pH and salinity had a negative impact on soil nutrients (N, P, K and organic carbon etc.) Another thing was observed in the fields. When the fish ponds were filled with brackish water and this existing water was drained out after fish culture, then the adjacent paddy fields were badly affected by the polluted water and sediments. Table 12.5 Relation between distance from fishponds and soil salinity parameters Distance from fishponds (m)

Average soil pH (No. of the sample of each category = 20)

Average Ec (ds/m) (No. of sample of each category = 20)

0–50

8.79

5.34

50–100

8.23

4.21

100–150

7.43

3.28

150–200

7.22

2.61

200–250

6.52

2.42

Source Field survey

Relation between soil pH and distance from fish ponds(m)

Relation between soil Ec (Ds/m) and distance from fish ponds(m)

PH Linear (PH)

0

100

200

300

Distance from fish ponds(m)

Soil Properties Ec Ph

y -0.0149x + 5.06 -0.0111x + 8.748

Average Ec (Ds/m)

Average soil pH

6 y = -0.011x + 8.748 R² = 0.979

10 9 8 7 6 5 4 3 2 1 0

y = -0.014x + 5.06 R² = 0.942

5 4 3

EC

2

Linear (EC)

1 0 0

100

200

300

Distance from fishponds (m)

R -0.97 -0.99

R² 0.9429 0.9793

Fig. 12.7 Relation between distance from fish ponds and soil salinity parameters (Source Field survey)

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

267

Soil Nutrient Status: The salinity of the soil may be the culprit for the unusual change of soil property. Soil pH plays an important role in plant nutrition. Continuous use of nitrogen particularly, ammonium nitrogen is one of the leading reasons for the lowering of the soil pH . When bacteria present in the soil starts to convert the ammonium present in the fertilizer into nitrate (nitrification) the release of hydrogen begins and increases the soil’s acidity. If the amount of ammonium nitrogen increases in the soil then the acidifying potentiality will also be increased. This process can severely reduce the pH level of the soil (Zafar et al. 2016) (Table 12.6 and Fig. 12.8). About 90% of soil samples had a low amount of nitrogen, while only 4% of samples had a high amount. Nitrogen in ammonium form was used for paddy farming. But here almost the whole region has a low amount of ammonium, which indicates that these agricultural lands were losing their fertility and becoming unsuitable for further cultivation. The amount of nitrogen is low (5.8–29.4 kg/acre) in the maximum portions of the region. The central portion of Dhaipukhuria, in a few portions of Jiagodi and the south-eastern portion of Katapukhuria, had medium to high amounts of nitrogen (29.4–40.0 kg/acre and 40.0–81.6 kg/acre) in the soil. The amount of nitrogen became mainly low due to the continuous increase of soil pH . Almost half of the soil samples had no phosphate content, while 27 and 14% of soil samples had medium and medium to high concentration phosphate respectively. Phosphorus, next to nitrogen, is the most fundamental nutrient for crop production. But the soil pH level highly influenced the nature of the phosphorus in the soil. Plants can’t do anything without phosphorus. But there are often withdrawal limits on how Table 12.6 Soil nutrient status At the fishery (0 m) N = 50 Cultivated land (far from the 250 m) N = 250 Soil parameters Mean Standard Co-efficient Mean Standard Co-efficient Remarks deviation of variation deviation of variation The (SD) (CV) (SD) (CV) recommended amount for high-yield of crops by I.A.R.I Organic carbon (% by wt.)

2.5

1.8

27.54

0.9

0.68

22.14

Above 0.75%

Total nitrogen (kg/acre)

79.10 11.56

14.55

13.15 2.30

17.88

20 lbs (un-irrigated soils and 40 lbs (irrigated soils)

Phosphate (kg/acre)

74.15

6.94

14.41

15.87 2.17

15.21

50–80 lbs/acre

158.76 43.13

23.17

79.38 9.54

13.47

250 lbs/acre

Potassium (kg/acre)

Source Samples collected from field areas

268

M. Mondal et al.

(a)

(b)

(c)

(d)

Fig. 12.8 a, b, c and d Change in soil nutrient status (2021) (Source Field survey)

much phosphorus they can get from the soil (Chakraborty and Prasad 2021). The soil pH level and the availability of phosphorus in the soil have an inverse relationship with each other. In acidic soil, the amount of soil phosphorus is very low. In acidic soil, phosphorus interacts with iron and aluminium, and as a result, the property of the soil started to change. On the other hand, if the soils are too alkaline in nature, phosphorus reacts with the calcium and becomes unfavourable for the tree as well as the crop productivity (Dean and Rubin 1947). Soils of most areas contained trace to a very low amount of phosphate, while in north-eastern portion and in some patches of the central portion, a medium amount of phosphate was found. Zones of high amounts of phosphate were rare and found in dispersed nature in some small patches. It can be concluded that the soils of most of the agricultural lands were deficient in phosphate and the phosphate content should be enriched by using fertilizers to gain well rice production. More than 84% of soil samples had a low amount of organic carbon and 15% had medium and only 1% of samples had a high amount of organic carbon. Soils of most of the area contain a low amount (0.5–0.58%) of organic carbon. Some of the

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

269

portions in the west and north have a medium concentration of organic carbon (0.58– 0.66%) and a little portion on the northern side has a high concentration of organic carbon (0.66–0.74%). It can be concluded that increasing salinization is responsible for this reduction of organic matter in agricultural soils. Soil organic matter (SOM) is the organic part of the soil which is constituted with the parts of organisms in various levels of decomposition. The main function of SOM is to neutralize the soil pH by releasing H+ from alkaline soil and fixing H+ within acidic soil. When organic matters start to decay, it causes a soil pH increase. Soil microbes change the soil pH (increases pH in mineralization and decreases pH in nitrification). Soil pH increases in alkaline soils with the decomposition of organic matter in greater amounts. In our study area, most of the samples (61%) have a very high amount of soil potassium. Too much potassium disrupts the uptake of other important nutrients, such as calcium, magnesium and nitrogen. This will stunt the growth of the plants. The medium and high amount of potassium is suitable for rice cultivation. Here only 30% of soils contain a medium and high amount of potassium. Potassium and phosphate availability in soil is mostly dependent on the soil pH . Calcium positively reacted with potassium and phosphate in alkaline soils (Sharpley 1989). If Potassium is available in greater amounts in soil then, calcium will move the potassium from the clay colloids and prepare it more water soluble for the crops. On the contrary, the amount of phosphate is gradually becoming insoluble because of the reaction of aluminium, ferrous and calcium ions and due to this, phosphate is unavailable for crops. Therefore, the most suitable use of potassium and phosphate is found in the pH range of 6–7 (Akinola et al. 1976) (Table 12.7). Table 12.7 Perception of respondents about soil salinization and soil nutrients Sl. No.

Perception on land use change

Strongly agreed (%)

Agreed (%)

1

Low yield of paddy

40

60

2

The increasing rate of disease in cultivable crops

30

65

Neither agreed nor disagreed (%)

Disagreed (%)

Strongly disagreed (%)

0

0

0

3

2

0

3

Increase in salinity

100

0

0

0

0

4

Forced to use more fertilizer

100

0

0

0

0

5

Draining out of existing water of post-fish culture into nearby agricultural fields

60

40

0

0

0

6

Harmful effects of the chemical used in fish ponds

10

20

70

0

0

Source Field survey

270

M. Mondal et al.

12.5.4 Assessment of Water Quality For the assessment of water quality; the following (Table 12.8) parameters had been tested in the laboratory of the Dept. of Geography and Chemistry of Midnapore College (Autonomous). Water logging problem: The water logging problem is a great threat to the environment. Before installing the fisheries, there was a natural slope which helps the smooth flow of water. But the water logging started when this natural slope is disturbed by installing artificial lowlands (ponds) on the agricultural lands. Now, the water can’t flow following the natural slope, rather it logged in comparatively low lands where it can. This scenario creates socio-environmental effects. At first, water stagnated and then overflow started when the spaces are filled. As this incident creates damaged biodiversity and broke the natural laws of the water system. As a result, local people face many problems. These are given belowFlood: During the rainy season overflow of water results in a flood, as in the absence of a natural slope the excessive water cannot be drained out fully from the region. This results in the rise in groundwater table and salinization: In rainy seasons excessive water logging results in rapid infiltration. As a result the level of the groundwater table starts to rise. And the soil always is in a saturated condition. Due to the excessive wetness of the soil, the salinity of the soil starts to increase. Salt contents of the soil start to rise from the lower layer to the subsurface layer capillary action and thus salinization takes place, which can badly harm crop production. Lack of aeration: Due to water logging, the soil is always saturated and as a result, the pore spaces between the particles of the soil are always filled up, which can obstruct the passage of air. Without air microbial activity can stop, plant roots degenerate and crops can die. Impact on soil: The availability of nutrient elements in the soil can reduce and leaching loss can be a major issue. Table 12.8 Measurement of water quality Water parameters At the fishery (0 m) N = 50 Mean

pH Dissolve organic carbon/matter (µM) DO (mg/liter)

Standard deviation (SD)

Co-efficient of variation (CV)

Far from the fishery (Drainage channel) N = 50 Mean

Standard deviation (SD)

Co-efficient of variation (CV)

8.6

0.6

7.46

8.2

0.5

11.25

860.4

398.4

13.24

377.13

231.17

13.69

19.47

11.25

2.36

22.32

9.3

1.97

Salinity (ppt)

14.4

6.4

15.14

16.5

9.8

23.40

Ammonia (µM)

48.63

28.47

17.83

37.52

21.34

23.45

Source Samples collected from field areas

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

271

Disease: Due to the water logging problem many diseases like skin problems, cholera, and malaria can be seen. Decrease in livestock: In the last ten years, 60% of the lands used for animal rearing decreased in the study area. Therefore, the number of livestock decreased in that area. According to a field perception study, the average number of livestock per household significantly decreased from 4.3 (2010) to 2.4 (2020).

12.5.5 Loss of Bio-diversity Due to the uncontrolled and unprecedented conversion of cultivable land to saltwater shrimp fisheries, the local rural bio-diversity had been threatened severely. Excessive saltwater intrusion in ponds, land and nearby water bodies like canals, bills, nayanjulies (drainage systems) etc. had seriously damaged the freshwater ecosystem. The freshwater fishes and turtles were killed or died. Many birds like saras, heran, kingfisher etc.… which were directly dependent upon the rural freshwater ecosystem will now also be under threat and they are either forced to migrate outside or stay there to die. The rural Purba Medinipur particularly, Bhagwanpur II block was historically famous for its rich freshwater local fish diversities but now it becomes a myth due to the destruction of freshwater bodies through the overwhelming spread of brackish water fish ponds. The changing status of a few prominently endangered animal species had been presented in the following figure (Fig. 12.9). Perception analysis on bio-diversity As expected most of the respondents presented their views on this alarming rate of biodiversity loss. This loss was not only influenced by the ecosystem but also by their livelihood. Maximum poor people particularly women supplied protein in their family diet by collecting small fish from freshwater bodies, especially from paddy fields (Table 12.9). Perception Analysis on Over Ecological Impact: The most devastating impact of this land conversion had been found on the overall deterioration of soil, water, biodiversity and agricultural productivity. Most of the respondents agreed upon these issues (Fig. 12.10). Few people could not perceive properly due to their ignorance and few were not ready to claim these types of issues for their self-interests. Various complaints were launched repeatedly to the administrations against this type of forceful land acquisition by land mafias but, all these efforts went in vein.

12.5.6 Ecological Cost Benefit (EECB) Analysis In this section, the authors tried to measure the balance between the economic and ecological benefits of this land conversion.

272

M. Mondal et al.

Name of the fauna/flora

Local name

Scientific name

2010

2020

Wild cat

Khatash/ Baghroll

Prionailurus viverrinus

Vulnerable

Extinct

Udbiral

Lutra lutra/ Lutra sumatrana

Rare

Extinct

Makroll

Paradoxurus hermaphroditus

Rare

Extinct

Desi sial/Khenk sial Canis aureus indicus

Vulnerable

Extinct

Vonda sial

Vulpes bengalensis

Vulnerable

Vulnerable

Gokhuro

Naja naja

Vulnerable

Vulnerable

Keute

Naja kaouthia

Vulnerable

Vulnerable

Dadash

Ptyas mucosa

Vulnerable

Vulnerable

matiali

Atretium schistosum

Vulnerable

Vulnerable

Dhora

Fowlea piscator

Abundance Vulnerable

Hele

Amphiesma stolatum

Abundance Vulnerable

Lizards

Gosap

Varanus bengalensis

Abundance Extinct

Mongoose

Newul

Herpestes javanicus palustris

Vulnerable

Frogs

Sona bang

Hoplobatrachus tigerinus

Abundance Vulnerable

Metho bang

Phrynoglossus Peters

Abundance Vulnerable

Lissemys punctata

Vulnerable

Fox/Jackal

Snake

Turtles /Indian flap shell turtle Kachchhap

Picture

Vulnerable

Extinct

Fig. 12.9 Bio-diversity status (Source Field survey)

Ecological Cost Index (ECI) Analysis: The following Table 12.10 showed the perception analysis of local residents regarding the economic and ecologic costs they had to pay due to this land conversion. Ecological Benefit Index (EBI) Analysis: The following Table 12.11 showed the perception analysis of local residents regarding the economic and ecologic benefits they had to enjoy due to this land conversion.

12 Analysis of Ecological Vulnerability Behind the Land Conversion … Birds

Freshwater fishes

Fig. 12.9 (continued)

273

Sparrow

Passer domesticus

Abundance Vulnerable

Kingfisher

Alcedo atthis

Abundance Extinct

Common Kingfisher

Alcedo atthis

Abundance Vulnerable

Shoul

Channa striata

Abundance Vulnerable

Chital

Chitala chitala

Abundance Vulnerable

Magur

Amblyceps mangois

Abundance Vulnerable

Khalsia

Colisa fasciata

Abundance Vulnerable

Bele

Awaous grammepomus

Abundance Vulnerable

Shinghi

Heteropneustes fossilis

Abundance Vulnerable

Nadose

Nandus nandus

Abundance Extinct

Foli

Notopterus notopterus

Abundance Vulnerable

Chela

Salmostoma acinaces

Abundance Extinct

Phasa

Setipinna Phasa

Abundance Extinct

Chanda

Chanda nama

Abundance Vulnerable

Cutchia/Barua

Monopterus cuchia

Abundance Vulnerable

Pabda

Ompok Pabo

Abundance Extinct

Chang

Channa orientalis

Abundance Extinct

Kachon punti

Pumtius ticto

Abundance Extinct

Shor punti

Puntius sarana

Abundance Extinct

Koi

Anabas testudineus

Abundance Vulnerable

Baim

Mastacembekas armatus

Abundance Vulnerable

Boal

Wallago attu

Abundance Extinct

Pakal

Macrognathus panchalus

Abundance Vulnerable

Mola

Amlypharyongo donmola

Abundance Extinct

274

M. Mondal et al.

Table 12.9 Perception on biodiversity Sl. No.

Perception on land use change

Strongly Agreed agreed (%) (%)

Neither agreed nor disagreed (%)

Disagreed (%)

Strongly disagreed (%)

1

Loss of native fish species

75

25

0

0

0

2

Loss of other aquatic species

75

25

0

0

0

3

Declining of terrestrial species

35

47

18

0

0

4

Change in the natural ecosystem

80

20

0

0

0

5

Declining the quality of habitat

20

80

0

0

0

6

Decreasing ecosystem productivity

30

70

0

0

0

Source Field survey

Strongly agreed (%)

Decreasing agricultural productivity Proliferation of diseases Occurrence of utrofication Desertification of local natural resources

Agreed (%)

Alteration of local hydrology Destruction of fluvio-coastal aquatic… Declining landscape sustainability Neither agreed nor disagreed (%)

Water and land pollution Change in soil nutrient status Increasing rate of soil salinity Biodiversity loss

Disagreed (%)

Disruption of natural drainage system Alarming change in land use change 0%

50%

100%

Strongly disagreed (%)

Fig. 12.10 Perception of overall ecological impact due to shrimp cultivation (Source Field survey)

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

275

Table 12.10 Ecological cost index (ECI) analysis Costs

Dimension/Variables

Weighted values (5)

Average weighted value

Specific cost index (SCI)

Ecological

1. Alarming change in land use

4.9

0.912

Ecological cost index (ECI) = 0.912

2. Threat to biodiversity

4.6

3. Increasing salinity of soil and water

4.8

4. Alteration of local hydrology

3.8

5. Deterioration of soil quality

4.7

Table 12.11 Ecological benefit index (EBI) analysis Benefits

Dimension/Variables

Weighted values (5)

Average weighted value

Specific benefit index (SBI)

Ecological

1. Use of uncultivated land

4.1

0.71

2. Utilization of low-productive cultivated land

3.0

Ecological benefit index (EBI) = 0.71

Ecological Cost Benefit Index Analysis (ECBI): The following Table 12.12 showed the perception analysis of local residents regarding the cost and benefits they had to pay or enjoy due to this aquaculture practice. One thing was very clear from the above index analyses that according to the perception of the local residents of the study areas, the ecological cost was very high rather than the economical benefits (Table 12.12). The cost was higher than the benefit of this type of land conversion. Many villagers did not agree to give up their lands for aquaculture but, they were helpless because of the pressure of land mafias and obviously of political pressure. So, there was an urgent necessity for administrative interventions to stop illegal aquaculture. The implementation of scientific management planning should be necessary by taking the views of all stakeholders, particularly for the benefit of low and marginal-income groups of the society. Table 12.12 Ecological cost benefit index analysis (ACBI)

Cost index (CI)

Benefits index (BI)

Ecological cost index (ECI) 0.912

Ecological benefit index (EBI) 0.71

Ecological cost benefit index (ECBI) EBI/ECI = 0.71/0.912 = 0.779

276

M. Mondal et al.

12.6 Recommendations of Road Map for the Development and Sustainable Coping Strategies The sudden growth of unplanned aquaculture brings huge problems to the environmental status of the surrounding areas. The economy has no doubt increased due to aquaculture but it disrupted the equilibrium of the natural environment. The rich people become richer and the poor became poorer. Socio-economic conflict among have and have not has altered and been assessed in society. In this context, the author has tried to suggest a model for management recommendation (Table 12.13) for the future betterment of the section of society. The main driving cause of land conversion of agricultural land to the fishery is the high profitability in terms of hard cash in the hands of the farmers. They used to prefer monetary income per year from the rent of the land given for fishery rather than risky and hazardous crop cultivation. A clear-cut economic differentiation among residents is created due to these land use changes. It is very much worthy in this context to implement such a concrete policy by which the conflicts between aquaculture and agriculture will be minimized. In case of land conversion and LULC change, concrete land use policy of govt. should be implemented. The role of local administrators, representatives, political parties and other chair holders is very vital regarding monitoring and taking strict steps against the illegal activities if any from the end of fishery farm owners for the benefit of particularly poor cultivators. Sustainable tools and techniques for this practice should be implemented from joint efforts of individuals, institutions, and organizations regarding aquaculture practices. More research, studies, and projects should be conducted for creating a well understanding of sustainable aquaculture and its far-sighted positive impacts and Table 12.13 Management recommendations Detection of the existing situation

Emphasize on the Assessment of perception of the impact local respondent

Rapid government initiation

Planning strategies

First of all the actual Situation of land conversion should be visualized and analyzed. All types of change including economy, society, ecology, and livelihood should be checked and detected

Emphasis should be given to the perception of local residents on a priority basis because only they can tell the actual truth

Government should take immediate action against illegal fish ponds without considering any political color and expand its hands to strengthen the economy of small ad marginal sections of society

To maintain the economic and ecological sustainability of the study area, all stakeholders should work together with full cooperation. Keeping in mind the overall benefit and development of sustainable livelihood of local respondents

Proper assessment should be done on economic cost and ecological cost so that a balance can be maintained between them

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

277

costs for maintaining economic sustainability. Awareness programmes have to be implemented among farmers about the long-run benefits of fertile agricultural lands as well as agricultural practices. The loss or uncertainty in agricultural sectors should be recovered systematically by conjugated cooperation of farmers and the government. Co-cultivation of fish and crops should be considered in the thinking, innovation and research for economic security through joint ventures of agriculture and aquaculture sectors.

12.7 Conclusion The rapid conversion of agricultural lands into aquaculture is the result of the selfinterest of a few businessmen, politicians, land mafias and even, a few persons in administration with the indirect support of the state. The poor and marginal sections of society are the main target for acquiring their land by hook or by crook. This conversion brings a hazardous impact on their livelihood. A few alarming social issues also came out in this research in the time of conversation with local residents. There has been no doubt that shrimp farming is more profitable than agriculture. People earn hard cash in hand without any investment. They just give their land on a lease basis and earn a comfortable amount of money per year. But this system encourages some negative impacts on society. It changes economic stratification which ultimately leads to social stratification or class stratification. Moreover, the agricultural lands of the study area are fertile. This type of land conversion leads to the complete destruction of fertile land, food insecurity, economic differentiation, and social differentiation. This short-term economic prosperity at the cost of devastating ecological exploitation cannot be considered the strength of the local residents because in this case, this life-supporting ecology is not capable to recognize social values. It welcomes the vulnerability of a life-sustaining system by means of economic stratification, class conflicts, and ecological destruction. Moreover, it will disrupt cultural traditions of significance for sustainable resource use (Kagoo and Rajalakshmi 2002) It is noticed that though shrimp farming is highly profitable than traditional rice cultivation, the unfavourable impacts of aquaculture on the environment create various implications. The soil pH of most of the agricultural lands (70%) is found to be 8.0 to 8.0, which indicates that most of the agricultural lands become salinized due to saltwater intrusion, seepage of salt water into agricultural lands from shrimp ponds, aqua cultural pond overflow, leaching from sludge pile during rainfall. Soil chemical properties are also changing their nature and becoming unsuitable for rice cultivation. Organic matter in soil tends to diminish, as soil organisms can’t live in highly salinized soils. Spatial differences in maps of different soil chemical properties can clarify that most of the agricultural lands are becoming unsuitable for prosperous rice production. Water logging and poor drainage condition due to obstruction of the natural slope of land by installing more and more shrimp ponds results in flooding, groundwater table rising, salinization due to capillary action and some diseases. And it is also noticed that the life span of shrimp ponds is not more than 5–10 years from installation, due

278

M. Mondal et al.

to self-pollution and catastrophic collapses due to viral and bacterial diseases. As a result, the number of abundant fisheries is increasing day by day, which results from the rapid conversion of agricultural lands into new shrimp ponds. Now it is a matter of thinking that if this trend continues the scenario of the whole area will be like undulated surface or degraded badlands, which cannot produce any kind of crop. Though the people are thinking that they will start again agriculture after reforming those lands, it is doubtful that either those lands will produce as they did in past, or they will fail to produce successfully. Besides such environmental issues, it results in the marginalization of the rural poor, social unrest and inequality (fishery owners become richer and paddy cultivators become poorer), unemployment, migration, and breakdown of the traditional livelihood support system. So it is high time to concern people about the adverse impacts of land conversion. Sustainable development of aquaculture may be attainable, if proper eco-friendly, scientific, economically feasible and socially admissible practices are taken into consideration. Acknowledgements The authors are sincerely grateful to the respondents of the studied villages for their kind cooperation, sparing their time and giving thoughtful opinions. The authors are thankful to the students of 3rd year Hons., Dept. of Geography for their selfless help in field visits and laboratory experiments. They are also thankful to the officers of different levels of administration for their kind help with secondary data. Credit also goes to Google for the supply of satellite images. Declaration There is no such fund that has been utilized for this research project. No personal or institutional conflicts of interest are associated with this study. This is completely an original work of the authors and has not been published anywhere.

References Abraham TJ, Sasmal D (2009) Influence of salinity and management practices on the shrimp (Penaeus monodon) production and bacterial counts of modified extensive brackish water ponds. Turk J Fish Aquatic Sci 9:91–98. https://www.trjfas.org/uploads/pdf_734.pdf Akinola A, Agboola, Oko B (1976) An attempt to evaluate plant available P in Western Nigerian soils under shifting cultivation. Agron J 68(5):798–801. https://doi.org/10.2134/agronj1976.000 21962006800050028x Ahmed N, Diana SJ (2015) Coastal to inland: expansion of prawn farming for adaptation to climate change in Bangladesh. Aqua Rep 2:67–76. https://doi.org/10.1016/j.aqrep.2015.08.001 Ahmed N, Glaser M (2016) Coastal aquaculture, mangrove deforestation, and blue carbon emissions: is REDD+ a solution? Mar Policy 66:58–66. https://doi.org/10.1016/j.marpol.2016. 01.011 Akteruzzaman M (2005) Shifting rice farming to fish culture in some selected areas of Mymensingh, Bangladesh: the process, conflicts and impacts. Bangladesh J Fish Res 9(1):97–99. http://hdl. handle.net/1834/33442 Akteruzzaman M, Jaim WMH (1999) The impact of privatization and deregulation on new technology diffusion and agricultural growth performance in Bangladesh-a macro view. Bangladesh J Agric Econ 22(454–2016–36639), 51–62. https://doi.org/10.22004/ag.econ.202090 Allen IE, Seaman CA (2007) Likert scales and data analyses. Qual Prog 40(7):64–65 Azad AK, Jensen KR, Lin CK (2009) Coastal aquaculture development in Bangladesh: unsustainable and sustainable experiences. Environ Manag 44:800–809

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

279

Burgi M, Hersperger AM, Schneeberger N (2004) Driving forces of landscape change-current and new directions. Landscape Ecol 19:857–868 Campbell DJ, Smucker TA, Olson JM, Wangui EE (2007) Contemporary challenges of participatory field research for land use change analyses: examples from Kenya. Field Methods 19(4):384–406 Chakraborty D, Prasad R (2021) Stratification of soil phosphorus forms from long-term repeated poultry litter applications and its environmental implication. Environ Challenges 5:100374.https://doi.org/10.1016/j.envc.2021.100374 Chowdhury MA (2006) The sustainability assessment of shrimp farming: a case study of the present farming systems in Bangladesh. Asian Institute of Technology (AIT), Thailand PhD Dissertation no. AQ-06-5 Clay J (1996) Markets potentials for redressing the environmental impact of wild captured and pond produced shrimp, manuscript, World Wildlife Fund, Washington, DC, 188 pp Das R, Mondal M (2020) Changing landscape ecology and dwindling environment due to the massive growth of shrimp farming: challenges to micro-level planning and regional development of fluvio-coastal Khejuri over South Bengal Basin. Progressive Publishers, Kolkata, pp 196–229 Dean LA, Rubin EJ (1947) Anion exchange in soil. Exchangeable phosphorus and anion exchange capacity. Soil Sci 63(5). https://doi.org/10.1097/00010694-194705000-00005 De D, Banerjee S, Ghosh S (2014) Assessment of land use and land cover changes in Panchrakhi Village, Hooghly District, West Bengal, India. IOSR J Hum Soc Sc (IOSR-JHSS) 19(7):120– 126, Ver. + IV. http://www.iosrjournals.org/iosr-jhss/papers/Vol19-issue7/Version-4/S01974 120126.pdf De Roy S (2013) Impact of fish farming on land relations: evidence from a village study in West Bengal. Indian J Agric Econ 68(2), April–June. https://www.researchgate.net/publication/287 633920 Derrick B, White P (2017) Comparing two samples from an individual Likert question. Int J Math Stat 18(3) Dewalt BR, Vergne P, Hardin M (1996) Shrimp aquaculture development and the environment: people, mangroves and fisheries on the Gulf of Fonseca, Honduras. World Dev 24(7):1193–1208 Disperati L, Virdis SGP (2015) Assessment of land use and land-cover changes from 1965 to 2014 in Tam Giang-Cau Hai Lagoon, Central Vietnam. Appl Geogr 58:48–64. https://doi.org/10.1016/ j.apgeog.2014.12.012 Flaherty M, Vandergeest P, Miller P (1999) Rice paddy or shrimp pond: tough decisions in rural Thailand. World Dev 27(12):2045–2060 Gujja B, Finger-Stich A (1996) What price prawn? Shrimp aquaculture’s impact in Asia. Environ Sci Policy Sustain Dev 38(7):12–39 Hein L (2000) Impact of shrimp farming on mangroves along India‘s East Coast. Unasylva J 51(203):48–55 ref. 20 (English ed.). www.cabdirect.org/cabdirect/abstract/20003028691 Hein L (2002) Toward improved environmental and social management of Indian shrimp farming. Environ Manage 29(3):349–359 Huang M, Tang Q, Ao H, Zou Y (2017) Yield potential and stability in super hybrid rice and its production strategies. J Integr Agric 16(5):1009–1017. www.sciencedirect.com Hossain MS, Otta SK, Chakraborty A, Kumar HS, Karunasagar I, Karunasagar I (2004) Detection of WSSV in cultured shrimps, captured brooders, shrimp postlarvae and water samples in Bangladesh by PCR using different primers. Aquaculture 237(1–4):59–71 Hossain MS, Uddin MJ, Fakhruddin ANM (2013) Impacts of shrimp farming on the coastal environment of Bangladesh and approach for management. Rev Environ Sci Bio/Technol 12(3):313–332 Indian Agricultural Research Institute (IARI) (2014) Handbook of soil testing kit. Kolkata Jayanthi M, Thirumurthy S, Muralidhar M, Ravichandran P (2018) Impact of shrimp aquaculture development on important ecosystems in India. Global Environ Change 52:10–21 Jovanovi´c V, Lazi´c M (2020) Is longer always better? A comparison of the validity of single-item versus multiple-item measures of life satisfaction. Appl Res Qual Life 15(3):675–692

280

M. Mondal et al.

Kagoo E, Rajalakshmi N (2002) Environmental and social conflict of aquaculture in Tamilnadu and Andhra Pradesh. J Soc Econ Dev 4(1):13–26. http://www.isec.ac.in/JSED/JSED_V4_I1_ 13-26.pdf Kautsky N, Ronnback P, Tedengren M, Troell M (2000) Ecosystem perspectives on management of disease in shrimp pond farming. Aquaculture 191(1–3):145–161 Lambin E, Meyfroidt P (2010) Land use transitions: Socioecological feedback versus socioeconomic change. Land Use Policy 27:108–118 (Elsevier). https://doi.org/10.1016/j.lan dusepol.2009.09.003. https://www.webambiente.gov.br/webambiente/wiki/lib/exe/fetch.php? media=webambiente:lambin_land.pdf Lebel L, Tri NH, Saengnoree A, Pasong S, Buatama U (2002) Industrial transformation and shrimp aquaculture in Thailand and Vietnam: pathways to ecological, social, and economic sustainability? AMBIO: J Hum Environ 31(4):311–323 Leung P, Tran LT (2000) Predicting shrimp disease occurrence: artificial neural networks vs. logistic regression. Aquaculture 187(1–2):35–49. https://doi.org/10.1016/S0044-8486(00)00300-8 Long H, Heilig GK, Li X, Zhang M (2005) Socioeconomic development and land-use change: analysis of rural housing land transition in the Transect of the Yangtse River, China. Land Use Policy 24:141–153, 222 (Elsevier). http://www.igsnrr.ac.cn/xwzx/jxlwtj/200702/W02009 0715580799873510.pdf Mitra R, Santra SC (2011) Influence of brackish water aquaculture on soil salinisation. Int J Res Chem Environ 1(2):166–168. https://www.academia.edu/4017799/Influence_of_Brackish_ water_aquaculture_on_Soil_Salinisation.10 Ojha A, Chakrabarty A (2018) Brackish water aquaculture development and its impacts on agriculture land: a case study on coastal blocks of Purba Medinipur district, West Bengal, India using multi-temporal Satellite data and GIS Techniques. Int J App Eng Res 13(11):10115–10123 Páez-Osuna F, Galván F, Guerrero SR, Ruiz ACF (1999) Discharge of nutrients from shrimp farming to coastal waters of the Gulf of California. Mar Pollut Bull 38(7):585–592. https://www.scienc ebase.gov/catalog/item/5053f0fde4b097cd4fcf7799 Páez-Osuna F (2001a) The environmental impact of shrimp aquaculture: a global perspective. Environ Pollut 112(2):229–231. https://doi.org/10.1016/S0269-7491(00)00111-1 Páez-Osuna F (2001b) The environmental impact of shrimp aquaculture: causes, effects and mitigating alternatives. Environ Manage 28(1):131–140. https://doi.org/10.1007/s00267001 0212 Páez-Osuna F, Gracia A, Flores-Verdugo F, Lyle-Fritch LP, Alonso-Rodrıguez R, Roque A, RuizFernández AC (2003) Shrimp aquaculture development and the environment in the Gulf of California ecoregion. Mar Pollut Bull 46(7):806–815 Paul P, Chakraborty DS (2016) Impact of inland fisheries on the socio-economic development: a focus on perspectives on development, Nadia District, West Bengal, India. Int J Fish Aquac Sci 6(1):59–76. ISSN 2248-9975 Perla RJ, Carifio J (2007) Psychological, philosophical, and educational criticisms of Harry Frankfurt’s concept of and views about “bullshit” in human discourse, discussions, and exchanges. Interchange 38(2):119–136 Primavera JH (1993) A critical review of shrimp pond culture in the Philippines. Rev Fish Sci 1(2):151–201. https://doi.org/10.1080/10641269309388539 Primavera JH (1997) Socio-economic impacts of shrimp culture. J Aquac Res 28:815–827. https:/ /citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.461.3150&rep=rep1&type=pdf Primavera JH (2006) Overcoming the impacts of aquaculture on the coastal zone. Ocean Coast Manage 49(9–10):531–545. https://doi.org/10.1016/j.ocecoaman.2006. 6.018 Proisy C, Viennois G, Sidik F, Andayani A, Enright JA, Guitet S, Gusmawati N, Lemonnier H, Muthusankar G, Olagoke A, Prosperi J (2018) Monitoring mangrove forests after aquaculture abandonment using time series of very high spatial resolution satellite images: a case study from the Perancak estuary, Bali, Indonesia. Mar Pollut Bull 131:61–71

12 Analysis of Ecological Vulnerability Behind the Land Conversion …

281

Rahman AF, Dragoni D, Didan K, Barreto-Munoz A, Hutabarat JA (2013) Detecting large scale conversion of mangroves to aquaculture with change point and mixed pixel analyses of highfidelity MODIS data. Remote Sens Environ 130:96–107. *******https://doi.org/10.1016/j.rse. 2012.11.014 Reason P, Bradbury H (eds) (2001) Handbook of action research: participative inquiry and practice. Sage Richards JA, Jia X (2006) Remote sensing digital image analysis: an introduction. Springer, Berlin Richards LA (ed) (1954) Diagnosis and improvement of saline and alkali soils. U.S.D.A. Handbook, 60 Robbins NB, Heiberger RM (2011) Plotting Likert and other rating scales. In: Proceedings of the 2011 joint statistical meeting, vol 1. American Statistical Association Rwanga SS, Ndambuki JM (2017) Accuracy assessment of land use/land cover classification using remote sensing and GIS. Int J Geosci 8(04):611 Sau S, Hazra N, Mondal M (2018) Social development: a block level study of Purba Medinipur District, West Bengal, India. Int J Recent Sci Res 9(3)(H):25188–25191. https://doi.org/10. 24327/ijrsr.2018.0903.1809. ISSN: 0976-3031 Sharpley AN (1989) Relationship between soil potassium and mineralogy. Soil Sci Soc Am J 53(4):1023. https://doi.org/10.2136/sssaj1989.03615995005300040006x Shiva V, Karir G (1997) Towards sustainable aquaculture: Chenmmeenkettu. Research Foundation for Science, Technology and Ecology, New Delhi, 133 pp Zafar MA, Haque MM, Aziz MS, Alam M (2016) Study on water and soil quality parameters of shrimp and prawn farming in the southwest region of Bangladesh. J Bangladesh Agric Univ 13:153. https://doi.org/10.3329/jbau.v13i1.28732 Zeng L, Shannon MC (2000) Salinity effects on seedling growth and yield components of rice. Crop Sci 40(4):996–1003. https://doi.org/10.2135/cropsci2000.404996x

Chapter 13

Environmental Change Analysis Using Remote Sensing and GIS: A Study of Upper Baitarani Basin, Odisha Tapas Ranjan Patra, Priyanka Chakraborty, Diptimayee Naik, and Ashis Chandra Pathy

Abstract Environmental change is directly associated with LU/LC changes and changes in vegetative cover. Therefore, Remote Sensing and Geographic Information Systems are commonly used for obtaining this kind of information. The present study focuses on the detection of environmental change in the upper Baitarani basin which is located in the Keonjhar district of Odisha state. In this study, LU/LC and NDVI (Normalized Difference Vegetation Index) are studied between 2000 and 2020 based on a Survey of India’s Topographic Map and temporal changes gathered from the Landsat 5 and 8 data and Sentinel 2 data. The study explores that the land under water bodies, agricultural activities, and irrigation reduces considerably due to mining activities and urbanization. On the other hand, the area under vegetative cover increased from 2010 to 2020 due to afforestation. Healthy vegetation accounts for minimum NDVI value. Therefore likewise the vegetative cover, and reclamation of water bodies is necessary to take necessary measures through local community participation. The experimental result indicates that over the specified period of study, the rate of increment and decrement in the urban built-up area, area of water bodies, vegetative cover, agricultural land, and other lands may have a significant impact on environmental sustainability. Keywords Land use and land cover · NDVI · Remote sensing · Geographic Information System · Landsat data

T. R. Patra · P. Chakraborty (B) · D. Naik Department of Geography, Rajendra University, Prajna Vihar, Balangir 767002, Odisha, India e-mail: [email protected] A. C. Pathy Department of Geography, Utkal University, Vani Vihar, Bhubaneswar 751004, Odisha, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_13

283

284

T. R. Patra et al.

13.1 Introduction The environment is the habitat of life on the earth and the existing growth, development, and functional activities of all living organisms including human beings are directed and determined by the environment. Thus, the environment is a complex of many components like biotic, and abiotic components which surround man as well as all living organisms. There are so many external factors like environmental, ecological, or simply living and non-living agents which affect the life of an organism including human beings that surround them. The four pillars of the environment are the atmosphere, hydrosphere, lithosphere, and biosphere. The present world is facing innumerable environmental problems due to urbanization, industrialization, population explosion, modernization, and technological advancement which impose enormous pressure on environmental resources. Thus, there is an urgent need for maintaining an equilibrium between the carrying capacity of the environment and the sustainable utilization of resources. “Environmental Impact Assessment (EIA) is the prior assessment of the future impact of the consequences of any decision on the quality of the total human environment on which man largely depends for his well-being” (Saxena 2017). In the studies of local, regional, and global environmental change land use land cover (LU/LC) change plays a vital role (Gupta and Munshi 1985; Mas 1999). Land cover refers to the natural cover of the Earth’s surface, endowed by forests, wetlands, impervious surfaces, agricultural land, grassland, and other types of land and water resources (Prakasam 2010). Land use refers to the bifurcation of land by the intervention of the functional role of the human being. Land use includes cultural landscapes like recreation areas, wildlife habitats, agricultural land, and built-up land (Reis 2008). Updated and accurate LU/LC maps have immense significance for proper planning, global change, environment monitoring, and the estimation of forest degradation. The consumption of natural resources is increasing with the pace of time due to the bouncing growth of the population in the world, and it seems that this situation puts increasing pressure on natural resources (FAO 1997). LU/LC changes are known to be the main focus of sustainable development (Lambin et al. 2000) and are very important concepts in natural resource management and monitoring (Sinha et al. 2015). The environment is the storehouse of resources over which all organisms including human beings depend for their basic needs like food, shelter, air, water, etc. Land and vegetation are important components of the environment. The land is one of the most precious assets in any region. The efficient and judicious utilization of land in general and that of urban areas, in particular, is critical. Land is required for various purposes in both urban and rural areas and all societies. It is a major factor of production and an important element for the socio-economic development of any country or society. The categories of land utilization are agricultural land, fallow land, forest areas, built-up areas including both rural and urban settlements, etc. (IGNOU 2019). We are currently living unsustainable lives because human beings are consuming resource energy and raw materials faster than the regenerative capacity of the natural

13 Environmental Change Analysis Using Remote Sensing and GIS …

285

system. For this reason, Land Use (LU) and Land Cover (LC) are changing continuously at a faster rate. Landuse changes are often non-linear and trigger feedback to the system, stressful living conditions and threaten the vulnerable society/community. Drivers such as population boom with additional needs for food, pressure on existing resources and urbanization, demand for more energy and fiber, and enhanced transportation network are the cause of deforestation, which fasters the Land use/Land cover (LU/LC) change. Remote sensing combined high-resolution images and image processing techniques to exhibit the LU/LC changes in a modernized way (Herold et al. 2003; Chaudhuri and Mishra 2016). Recently, aspects like vegetation cover, forest cover depletion, and urban expansion are analyzed using high-resolution satellite data (Mustafa et al. 2007). Conventional methods like field surveys are time-consuming and expensive and with the pace of time the produced maps become quickly outdated in a rapidly changing environment (Dash et al. 2015). Satellite data provides more valuable information about land use and land cover changes with less time and cost than the conventional method. However, the availability of satellite data sets is limited due to economic issues (Dwivedi et al. 2005; Gadrani et al. 2018). “Numerous methods have been developed by many researchers to review changes in the LU/LC (Al-doski et al. 2013; Singh 1989) including multi-temporal composite image change detection (Carmelo et al. 2012; Eastman and Fulk 1993) on-screen digitization of change (Sreedhar et al. 2016), vegetation index differencing (Shanmugam and Rajagopalan 2013), and postclassification change detection” (Belal and Moghanm 2011; Courage et al. 2013; Kafi et al. 2014). Mining and constructional activities may change land use and land cover pattern and may have negative impacts on the environment by fostering urbanization with dense transport and communication facilities. The Joda-Barbil area is located in the Joda block in the northern part of the Keonjhar district of Odisha. It hosts some of the richest iron ore deposits in India that have encouraged industrialization unprecedentedly. Although industrialization improves the economic life of people living in the Joda-Barbil area, it has also caused the living conditions to deteriorate steadily. This has degraded the land, soil, water, and atmosphere intensely and fastened the rate of weathering, and erosion resulting in the reduction of forest cover. It has affected the flora and fauna of the towns and also has the rivers and streams polluted by the effluents released from these processing plants. The atmosphere of Joda and Barbilis laden with dust due to mining activities. Therefore, as the two mining areas: Joda and Barbilare located in Upper Baitarani Basin, so, this basin area has been selected as a study area. Keeping in mind, the above problems of the study area, the following objectives are included: 1. to study the Spatio-temporal changes of land use land cover pattern in the upper Baitarani Basin. 2. to analyze the health of green vegetative cover from 2000 to 2020 in the upper Baitarani Basin. 3. to recommend suggestions to minimize the environmental degradation in the upper Baitarani Basin.

286

T. R. Patra et al.

13.2 Background Vivekananda et al. (2021) in their study “multi-temporal image analysis for LU/LC classification and change detection” have investigated the extent of land use change in Ananthapuram using multi-temporal satellite imagery between 1978 and 2018. From the study, they observed that the area under built-up land and other land is increased but the area under agriculture and waterbodies is reduced considerably. The major factor responsible for LU/LC changes in the study area is the decrease in agricultural activities and the expansion of built-up areas. Karaku¸s (2019) in his study “the impact of Land Use/Land Cover (LULC) changes on Land Surface Temperature in Sivas city center and its surroundings and assessment of Urban Heat Island” has revealed the relationship between LU/LC, NDVI, Land Surface Temperature, and Urban Heat Island in Sivas city and its periphery by using satellite images (1989–2010). It is observed from the study that the barren land has shown a decreasing pattern while land under agriculture and built-up area has shown an increasing pattern over the study period. Urban built-up areas and barren land record the highest LST in the study area due to the geographical location of the city, physio-morphological structure, and more importantly due to the construction material used. There is a negative correlation between the NDVI and LST in the study area accompanied by LU/LC changes. The LST changes show a decreasing tendency toward the rural areas from the urban areas. Lastly, it also demonstrates that there is a positive correlation between the urban built-up area and UHI in the study area. Attri et al. (2015) in their paper “Remote Sensing and GIS-based approaches for LU/LC change detection-A review” have suggested that the characteristics features of the study area; spatial resolution of the sensor, atmospheric impacts, and changes of the sun ray plays a pivotal role in the selection of the appropriate method to detect the change of any element or phenomena on the earth surface. So, they suggested in this paper pixel, features, and object level change detection techniques. They also recommended band image differencing and PCA techniques for detection. According to them, ANN/combination of change detection methods also can generate highquality change detection images. Therefore, it has been observed that most of the authors have done their works on LU/LC change and its impact on LST and UHI in the city area as well as recommended some modern LU/LC change detection techniques. Hence, the present paper is giving focuses on LU/LC and NDVI change over a period of time in a particular river basin area where mining activities are a predominant one that triggers the urbanization in Odisha-like backward states.

13 Environmental Change Analysis Using Remote Sensing and GIS …

287

13.3 Study Area The study area lies in the Keonjhar district of Odisha which is having a latitudinal extension of 21° 01' N to 22° 10' N and a longitudinal extension of 85° 11' E to 86° 22' E. It is sharing a border with administrative units of Jharkhand state in the north, Baleswar, and Mayurbhanj of Odisha in the East, Sundargarh in the west, and Dhenkanal district in the south of Odisha. The district is having 13 administrative blocks and three sub-divisional units (Keonjharsadar, Champua, and Anandapur) with the district headquarters at Keonjhar. The district has 8303 km2 of geographical area with well-connected roads to the major cities of Cuttack and Bhubaneswar. Undulating plains with isolated hillocks, high hills, and dome-shaped granite outcrops are different typical physiographic features found in the district. The district is mostly characterized by monsoonal rainfall from June to September months and has a hot and humid climate during summers and cold winters. About 30% of the district’s total geographical area is covered with dense forest mostly tropical moist deciduous type. The mountain ranges of the district are having rich mineral reserves of iron ore chromite and manganese for which mining activities in the area are quite dominant. The Baitarani is the major drainage basin draining the Keonjhar plains which are joined by some perennial and ephemeral tributaries namely Kanjhari, Sitanadi, Salandi, Khairibandhan, DeoNadi, etc. Due to high runoff and infiltration, the western part of the basin has a high drainage density. The river Baitarani originates from Guptaganga hills at Gonasika village in Keonjhar. Most of the portion of the Baitarani River basin (95%) lies in Odisha, while a small part of it (5%) lies in Jharkhand State. The upper Baitarani Basin, which includes the Panposh-Keonjhar-Pallahara plateau, is one of the parts of “The Central Plateau.” The Upper Baitarani Basin is situated between 85° 10' E to 86° 23' E longitude and 20° 53' N to 22° 15' N latitude and covers an area of 1812 km2 with an elevation varying from 375 to 1116 m above MSL showing high relief variation in the study area. The maximum and minimum temperature varies from 35 °C in the month of May to 10 °C in the month of January (Bardhan and Rao 2022). The Baitarani basin, with its abundant mineral and agricultural resources and cheap labor, provided a perfect setting for the construction and operation of numerous enterprises. The shifting cultivation is practiced in the upper catchment of the Baitarani River. The major urban centers in this region are Keonjhar, Joda, Champua, Swampatna, and Anandpur. The study area is a part of the Pre-Cambrian Banded Iron Ore Formation (BIF) belt comprising the Keonjhar Iron Ore province of the Eastern Indian shield of Odisha. The important rock types of the area are Banded Hematite Jasper (BHJ), shale, mineralized manganese phyllites, dolerites, limestone, massive iron ore, basic lavas, and Kolhan conglomerates. The various flow systems originating from the study area and its catchments constitute well-developed perennial and ephemeral rivers, tributaries, streams, springs, monsoon rivulets, and drainages of diverse patterns. The Baitarani River constitutes the main perennial river flowing from south to southeast over a conglomerate in the southern catchments of the study area. The Karo River

288

T. R. Patra et al.

originating from the altitude of 570 m and draining the entire northern sector of the study area, flows over shales and phyllites having trends from south-west to north direction. The area has been cut by several nalas, which originate from the river Karo running from North-West to South. The entire flow systems in the study area are controlled by minor local cracks, joints, rock fractures, and major tectonic lineaments (Fig. 13.1).

13.4 Materials and Methods 13.4.1 Data Collection The topographical maps were collected from Survey of India for the demarcation of the study area moreover satellite data (Landsat TM, ETM+, OLI and Sentinel) of the Upper Baitarani river basin region were downloaded from USGS Earthexplorer (https://earthexplorer.usgs.gov/) for the calculation of NDVI and LU/LC. This study ranges from 2000 to 2020 for the Spatio-temporal analysis of LU/LC Change and vegetative cover. Hence we have taken satellite images of the years 2000, 2010, and 2020, respectively. All the data are processed using ArcGIS 10.4 and Google Earth Engine (Table 13.1).

13.4.2 Method 13.4.2.1

Normalized Difference Vegetation Index (NDVI)

NDVI is widely used to understand the live green vegetation cover of the target of interest using the Red (Visible band) and Infrared (Near Infrared band) of the electromagnetic spectrum. As shown below, Normalized Difference Vegetation Index (NDVI) uses the NIR and Red bands of the Sentinel Satellite image in its formula. NDVI =

NIR − RED NIR + RED

(13.1)

For vegetation health monitoring data were accessed from USGS earth explorer and processed in ArcGIS 10.4 using the direct NDVI option from the toolbox.

13 Environmental Change Analysis Using Remote Sensing and GIS …

289

Fig. 13.1 Location map of study area

13.4.2.2

Land Use/Land Cover (LULC)

The land use/land cover map of the research region was created using the supervised classification approach (using ArcGIS), which included picking training sites that

290

T. R. Patra et al.

Table 13.1 Landsat and Sentinel images used in the study Satellite/ Factors

Sensor

Resolution (m)

Data source

Acquisition year

Format

Landsat 5 and Landsat 8

L5

30 × 30

USGS Earth Explorer

2000, 2010, 2020

Raster

Sentinel 2

S2

10 × 10

USGS Earth Explorer

2000, 2010, 2020

Raster

DEM

ASTER DEM

30 × 30

Global Mapper

2020

Raster

were generally representative of the different cover classes. The study area was classified into seven land use and land cover classes. The classifications were validated for the year 2020 using ground control points collected from the region. For LU/LC Landsat image has been used (Fig. 13.2).

Fig. 13.2 Flowchart demonstrating the methodology followed in present study

13 Environmental Change Analysis Using Remote Sensing and GIS …

291

13.5 Result and Discussion 13.5.1 Land Use/Landcover The study of Land use and land cover is critical in terms of understanding environmental changes. Increased population in recent times has greatly influenced the exploitation of natural resources which have resulted in increased cutting down of forests, loss of biodiversity, increase in the built-up area, and increased problems of natural disaster. Hence, it becomes vital to study Land use and Land cover for planning and decision-making of environmental management. In Fig. 13.4, there is a fluctuation in the areal extension (km2 ) of the water body over the two decades. From the year 2000 to 2010, the area (km2 ) of the water body increased from 145 km2 (9.0%) to 276 km2 (17.2%) out of the total area of the basin. On the other hand, in 2020 there is a sharp decline in the share of area (km2 ) of the water body. That is, it becomes 190 km2 approx (11.8%). Water bodies registered a growth in 2010 because the Landsat imagery was collected after monsoon month or else the pattern of changes in water bodies is modest throughout the study period. Vegetation cover has increased from the year 2000 (10.3%) to the year 2010 (18.4%) but again it declines to 11.2% out of the total area. Therefore, the area under vegetative cover is reduced due to the lumbering activities for commercial purposes. Forest areas of the study area have registered a negative growth from 190 km2 (11.9%) to 168 km2 (10.5%) for the years 2000–2010 due to mining activities which were quite high during that decade. Forest areas increased for the last decade (2010–2020) from 168 km2 (10.5%) to 338 km2 (21.1%) because of the reforestation and plantation activities carried out by the forest department of the government of Odisha. Vegetation of the region has some uneven patterns because of the clearing of green patches for mining activities. Agricultural fields were drastically diminished by almost 50% within two decades (2000–2020) from 310 km2 (19.3%) to 175 km2 (10.9%). There is the magnetic pull of mining activities to establish large-scale industries. Industrialization acts as a centripetal force to migrate people from different parts of the state/country for the sake of a job. It triggers the growth of transport and communication network as well as the urbanization process. Thus, the urbanization in mining areas fosters the magnitude of urban sprawling and the share of the urban built-up area increased from 210 km2 (13.11%) to 247 km2 (15.4%) within the two decades. There is a continuous downfall of the share of Barren lands and irrigated land in the last two decades in the upper Baitarani Basin area (Table 13.2). In the analysis of 2000, 2010, and 2020 Landsat satellite images, Agricultural and moderately Barren land accounted for the greatest share of land covering with a value of 19.3% and 19.7% respectively in the year 2000. But in 2010 and 2020, it gets reversed, i.e., in 2010 vegetative cover registered the highest position with 18.4% area, and forest cover deals with the lowest share of the total area (10.5%). In 2020 due to plantation activities and urban expansion, forest cover increased accounting for 21.1% of the shared area whereas agricultural land share is reduced to 10.9% of the total area which is the lowest one (Table 13.2).

292

T. R. Patra et al.

Table 13.2 Temporal variation of land use/land cover changes, indicator-wise, 2000–2020 Id

Land use/ land cover

Year 2000 Area (km2 )

1

Water

Year 2010 Area (%)

144.566

9.0

Area (km2 ) 275.927

Year 2020 Area (%) 17.2

Area (km2 ) 189.562

Year 2000–2020 Area (%)

Diff in Area (km2 )

11.8

44.997

2

Forest

190.054

11.9

168.583

10.5

338.408

21.1

148.354

3

Vegetation

165.331

10.3

295.227

18.4

178.929

11.2

13.598 −134.833

4

Agricultural

310.029

19.3

183.826

11.5

175.196

10.9

5

Irrigated

266.966

16.6

264.839

16.5

236.637

14.8

−30.329

6

Barren land

316.037

19.7

179.616

11.2

237.202

14.8

−78.834

7

Urban area Grand total

210.478

13.1

235.443

14.7

247.526

15.4

37.048

1603.461

100.0

1603.461

100.0

1603.461

100.0

0.000

Source Extracted from LU/LC map

In Fig. 13.3, it is observed that the cluster of urban settlement/built-up area was confined mainly to the northern part of the basin area. But with the pace of time, the built-up area is continuously spreading from the northern to the southern part along the eastern margin of the basin area. The location of the water bodies is mainly concentrated in the western part of the Baitarani basin region but unfortunately, the areal coverage of it is gradually decreasing from 2000 to 2020 (Fig. 13.4).

13.5.2 Normalized Different Vegetation Index (NDVI) NDVI is calculated from the Near Infrared (NIR) and Red band of the Sentinel 2 image. It is the ratio between the difference and sum of green and NIR bands. Here all the lower values are the vegetated zone. From the time series data, it is clear that vegetated zones are increasing with time. That increase may be due to some error in our calculation or artificial plantation by mining companies. Red shows the presence of mining areas in the figure whereas yellows show the other mixed features. A satellite-based measure of vegetation growth and cover, the normalized difference vegetation index (NDVI) is one of the most extensively used and recognized indicators of vegetation growth and cover. Near Infrared/Red represents the spectrums of a satellite image. NDVI image depends on the amount of plant coverage present, which permits rainfall to dissipate the energy before hitting on the terrain surface. It is particularly important to note that the Sentinel-2A, Level-2A satellite (which was launched on August 31, 2020) will offer high-resolution optical data (at resolutions of 10, 20, and 60 m) with atmospheric and radiometric correction. Using the European Space Agency’s Copernicus Access Hub (accessible on May 1, 2021) provides open access to ESA images technology and systems using the GIS application, we

13 Environmental Change Analysis Using Remote Sensing and GIS …

293

Fig. 13.3 Land use/land covers change (2000–2020)

were able to retrieve these images. For this experiment, it was decided to use NIR and green bands with a spatial resolution of 10 m (Fig. 13.5). The different NDVI classification of the vegetative cover has been given in the following table. The Upper Baitarani basin is experiencing different intensities of vegetation cover ranging from −0.32 to 0.79 (values of NDVI). The vegetative cover is classified into four groups viz. very low vegetation, sparse vegetation, moderately healthy

294

T. R. Patra et al.

Fig. 13.4 Land use/land cover, area (km2 ), (2000–2020)

vegetation, and healthy vegetation. It is envisaged that the share of healthy vegetation is lowest throughout the two decades (2000–2020) among the four vegetative classes. It is also observed that the share of the area under healthy vegetation is gradually increasing from 17.16% (2000) to 23.1% (2020). On the other hand, the highest share of vegetation type is varying in different periods of time. The sparse vegetation scores highest position acquiring 29.37% area (2000), followed by very low vegetation (28.57% area, 2010), and moderately healthy vegetation with 29.73% area in the year 2020 (Table 13.3).

13.6 Conclusion and Recommendation In this study, Remote Sensing and GIS has been integrated for quantifying and understanding the land use land cover (LULC) changes and vegetation change (NDVI) in the upper Baitarani basin of Keonjhar district, Odisha over two decades. The techniques and methodologies used in the study are simple and inexpensive. The LULC changes are determined here using multi-temporal satellite imagery of LANDSAT. In this study, it is observed that there is a significant change in the LULC in the study area. From 2010 to 2020, the areas under water bodies, forests, vegetative cover, agriculture, and irrigated land decreased drastically due to the growing mining activities and urban expansion. On the other hand, there is an increment in the area under forest and urban built-up land between the 2 years from 2010 to 2020 plantation, afforestation measures taken by the Government of Odisha have checked the depletion of forest cover. Industrialization and migration based on mining activities have accelerated the growth of urban built-up areas considerably. During the last 20 years, it is

13 Environmental Change Analysis Using Remote Sensing and GIS …

295

Fig. 13.5 Temporal NDVI variation (2000–2020)

also noticed that there is negative growth of agricultural, irrigated, and barren lands due to the encroachment by the urban built-up area. In the study, healthy vegetation scores always lowest NDVI value over the two decades which is not a very good indication of environmental sustainability. Besides this, the area with the highest NDVI values is not stable. It is changing in different periods of time. The areas covered with the highest NDVI value are located mainly in the west of the upper Baitarani

296

T. R. Patra et al.

Table 13.3 NDVI values for vegetative cover types Sl. No.

Classes

Values of NDVI

Percentage of area 2000

2010

2020

1

Very low vegetation

(−)0.03–0.3

26.95

28.57

23.95

2

Sparse vegetation

0.3–0.45

29.37

27.63

23.22

3

Moderately healthy vegetation

0.45–0.60

26.52

24.73

29.73

4

Healthy vegetation

0.60–0.79

17.16

19.07

23.10

Source Derived from NDVI map

basin and in fragmented in the way the other part of the basin. The low NDVI values are concentrated in the area covering the eastern margin and its surrounding basin. However, it can be concluded that as the plantation and afforestation measures have increased the area under vegetative cover, likewise the construction of check dams, rainwater harvesting in rural areas, wetland conservation, and water shade management can be taken as immediate measures for the reclamation of water bodies. Thus, LULC changes and the conservation of healthy vegetation should be closely monitored in the future for the sustainability of the environment.

References Al-doski J, Mansor SB, Shafri HZM (2013) Change detection process and techniques. Civ Environ Res 3(10):37–45 Attri P, Chaudhry S, Sharma S (2015) Remote sensing and GIS based approaches for LULC change detection—a review. Int J Curr Eng Technol 5(5):3126–3137 Bardhan A, Rao CM (2022) Hydrologic response estimation using different descriptors for upper Baitarani river basin. In: Advanced modelling and innovations in water resources engineering. Springer, Singapore, pp 219–233. https://doi.org/10.1007/978-981-16-4629-4_16 Belal AA, Moghanm FS (2011) Detecting urban growth using remote sensing and GIS techniques in Al gharbiya govrnorate, Egypt. Egyptian J Remote Sens Space Sci 14(2):73–79. https://doi. org/10.1016/j.ejrs.2011.09.001 Brown S (1997) Estimating biomass and biomass change of tropical forest: a primer, No. FAO-134: FAO Carmelo RF, Giuseppe M, Maurizio P (2012) Land cover classification and change detection analysis using multi-temporal remotely sensed imagery and landscape metrics. Eur J Remote Sens 45(1):1–18. https://doi.org/10.5721/EuJRS20124501 Chaudhuri G, Mishra NB (2016) Spatio-temporal dynamics of land cover and land surface temperature in Ganges-Brahmaputra delta: a comparative analysis between India and Bangladesh. Appl Geogr 68:68–83. https://doi.org/10.1016/j.apgeog.2016.01.002 Courage K, Jonah G, Hitomi M (2013) Monitoring urban spatial growth in Harare Metropolitan Province, Zimbabwe. Adv Remote Sens 2(4):322–333. https://doi.org/10.4236/ars.2013.24035 Dash PP, Kakkar R, Shreenivas V, Prakash PJ, Mythri DJ, Kumar KV, Sahai RMN (2015) Quantification of urban expansion using geospatial technology—a case study in Bangalore. Adv Remote Sens 4(04):330–342. https://doi.org/10.4236/ars.2015.44027

13 Environmental Change Analysis Using Remote Sensing and GIS …

297

Dwivedi RS, Sreenivas K, Ramana KV (2005) Cover: land use/land cover change analysis in part of Ethiopia using Landsat Thematic Mapper data. Int J Remote Sens 26(7):1285–1287. https:// doi.org/10.1080/01431160512331337763 Eastman JR, Fulk M (1993) Land sequence time series evaluation using standardized principal components. Photogram Eng Remote Sens 59(6):991–996. https://www.asprs.org/wp-content/ uploads/pers/1993journal/aug/1993_aug_1307-1312.pdf Gadrani L, Lominadze G, Tsitsagi MF (2018) Assessment of land use/land cover (LULC) change of Tbilisi and surrounding area using remote sensing (RS) and GIS. Ann Agrarian Sci 16(2):163– 169. https://doi.org/10.1016/j.aasci.2018.02.005 Gupta DM, Munshi MK (1985) Urban change detection and land-use mapping of Delhi. Int J Remote Sens 6(3–4):529–534. https://doi.org/10.1080/01431168508948474 Herold M, Goldstein NC, Clarke KC (2003) The spatiotemporal form of urban growth: measurement, analysis and modeling. Remote Sens Environ 86(3):286–302. https://doi.org/10.1016/ S0034-4257(03)00075-0 Kafi KM, Shafri HZM, Shariff ABM (2014) An analysis of LULC change detection using remotely sensed data; a case study of Bauchi City. IOP Conf Ser: Earth Environ Sci 20. https://doi.org/ 10.1088/1755-1315/20/1/012056 Karaku¸s CB (2019) The impact of land use/land cover (LULC) changes on land surface temperature in Sivas City Centre and its surroundings and assessment of Urban Heat Island. Asia-Pac J Atmos Sci 55(4):669–684. https://doi.org/10.1007/s13143-019-00109-w Lambin EF, Rounsevell MD, Geist HJ (2000) Are agricultural land-use models able to predict changes in land-use intensity? Agr Ecosyst Environ 82(1–3):321–331. https://doi.org/10.1016/ S0167-8809(00)00235-8 Mas JF (1999) Monitoring land-cover change: a comparison of change detection techniques. Int J Remote Sens 20(1):139–152. https://doi.org/10.1080/014311699213659 Mathur OP, Shukla P (eds) (2019) Urban planning: introduction to urban development. IGNOU, New Delhi, India Mustafa G, Tahsin Y, Reis S (2007) Using landsat data to determine land use/land cover changes in Samsun, Turkey. Environ Monit Assess 127(1): 155–167. https://doi.org/10.1007/s10661-0069270-1 Prakasam C (2010) Land use and land cover change detection through remote sensing approach: a case study of Kodaikanal Taluk, Tamil Nadu. Int J Geomat Geosci 1(2):46–55. https://doi.org/ 10.4236/gep.2020.88003 Reis S (2008) Analyzing land use/land cover changes using remote sensing and GIS in Rize, North-East Turkey. Sensors 8(10):6188–6202. https://doi.org/10.3390/s8106188 Saxena HM (2017) Environmental geography. Rawat Publication, New Delhi, India Shanmugam T, Rajagopalan B (2013) Urban change detection based on remote sensing and GIS study of salem revenue division, Salem District, Tamilnadu, India. Eur J Geogr 4(3):50–59. https://www.eurogeographyjournal.eu/articles/4.URBAN%20CHANGE% 20DETECTION%20BASED%20ON%20REMOTE%20SENSING_INDIA.pdf Sinha S, Sharma LK, Nathawat MS (2015) Improved land-use/land-cover classification of semiarid deciduous forest landscape using thermal remote sensing. Egypt J Remote Sens Space Sci 18(2):217–233. https://doi.org/10.1016/j.ejrs.2015.09.005 Singh A (1989) Digital change detection techniques using remotely-sensed data. Int J Remote Sens 10(6):989–1000. https://doi.org/10.1080/01431168908903939 Sreedhar Y, Nagaraju A, Krishna GM (2016) An appraisal of land use/land cover change scenario of Tummalapalle, Cuddapah region, India—a remote sensing and GIS perspective. Adv Remote Sens 5(4):232–245. https://doi.org/10.4236/ars.2016.54019 Vivekananda GN, Swathi R, Sujith AVLN (2021) Multi-temporal image analysis for LULC classification and change detection. Eur J Remote Sens 54(sup2):189–199. https://doi.org/10.1080/ 22797254.2020.1771215

Chapter 14

Mapping Urban Footprint Using Machine Learning and Public Domain Datasets Prosenjit Barman and Sk. Mustak

Abstract The urban footprint is termed as the physical cover of the urban built-up. In the past several decades urbanization has been accelerated due to rural–urban migration, economic growth, globalization, etc., and it is observed that over half of the world’s population now living in cities. Mostly, the unintentional urbanization causes impermeable surface which triggers several environmental challenges such as trash disposal, groundwater scarcity, heat island effect, and so on which need to be managed to support urban sustainability. The main objective of this study is to map urban footprint of Kolkata metropolitan area using machine learning (ML) algorithms (e.g., SVM, Random Forest) and public domain dataset. Landsat TM satellite data, Night time light data, and census data were employed in this study. Satellite imagery was used for mapping Lulc and reclassifying the built-up area into rural and urban built up using socio-economic data like census data. The robustness of the ML algorithms was tested based on classification accuracy and transferability assessment. In Lulc analysis band and feature stacked images give the high accuracy than the normal image and PCA image in three ML algorithms. SVM-Linear gave the high accuracy comparatively to another ML algorithm. The building footprint of KMA was extracted from top three high accuracy LULC map of different ML algorithm. The built-up area of KMA was validated using the test sample and for the validation of test sample these test sample uses in GHSL images. This finding will aid in the categorization of rural and urban areas and gives the idea of urban extent in the Kolkata metropolitan area. This study will help urban planners, local governments, and policymakers for the urban policy improvement and sustainable urban development planning. Keywords Urban footprint · Machine learning · Public domain dataset · Earth observation · Census

P. Barman · Sk. Mustak (B) Department of Geography, Central University of Punjab, Bathinda, Punjab, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 Sk. Mustak et al. (eds.), Advanced Remote Sensing for Urban and Landscape Ecology, Advances in Geographical and Environmental Sciences, https://doi.org/10.1007/978-981-99-3006-7_14

299

300

P. Barman and Sk. Mustak

14.1 Introduction Urbanization is one of the most important aspects of these contemporary globalization era (Wang et al. 2021). According to the United Nations in 2019 prospects more than half around 4.3 billion people of the world’s lives in urban areas. In this report was estimated that in 2050 the global population will reach around 9.8 billion. Among this population more than twice around 6.7 billion people will live in urban area (Henderson 2003). This highly population growth and related urbanization creates a challenge for the sustainable development. So, it is very important to understand the factors and dynamic spatiotemporal development of cities. In developing countries, the urbanization was grown very high. Developing country like India shares a very significant characteristics features of urbanization. In 2001 in India total population was 1027 million. Out of this population 285 million population (27.81%) lived in urban area while 742 million population lived in rural area and in 2011 the urban population became 31.16% (Jaysawal & Saha 2014; Census of India 2011). The urban area expanded day by day and its affect the surrounding valuable natural landscape like wetland, open space, and green space especially in urban area. The conversion of impervious surface impacts on ecosystem, biological diversity, climate, etc., which creates various negative effects like heat island. To monitor the impacts and support the sustainable development of city planners and policymakers needs the built up extent information (Xu 2007). According to Indian census 1981, the rural and urban area was defined by different characteristics. The urban area was defined as: (a) All statutory towns, i.e., all places with a municipal corporation, municipal board, cantonment board, notified town area, etc. were considered as urban area. (b) Other places which follow the below criteria consider as urban area. (i) A minimum population of 5,000; (ii) 75% of the male working population engaged in non-agricultural and allied activity; and (iii) A density of population of at least 400 per sq. km (or 1,000 per sq. mile) (Urban/Rural Definition, Indian District Database 1981). Esch et al. (n.d.) proposed a new data to map the global urban footprint. Urban footprint mapping is a challenging task because of unavailable data. Earth observation is the only datasets that provide the urban footprint information in the area based. The German TanDEM-X data was used in this study to classify the urban footprint. A convolutional Neural Network deep learning method was applied for the automated built-up area extraction using the medium-resolution datasets like 10 m resolution Sentinel-2 data. The built-up area was prepared in 5*5 pixel on sentinel 2 image and created the robustness of the model with high accuracy (Corbane et al. 2021). Hecar (2020) applied a new rule-based approach based on Delaunay triangulation among selected centroids of roads and dead-end streets. Road network geometry data was used to map urban footprint in this study. The method was applied in Wellington and compared with available authority data which gave the sufficient accuracy. The largescale urban extent mapping was prepared at an intermediate spatial resolution data

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

301

using a machine learning algorithm with combining the nightlight visible infrared imaging radiometer suits (VIIRS) and daytime moderate resolution Imaging Spectroradiometer (MODIS) data. Random Forest, Gradient boosting machine, Neural Network, and their ensemble were explored to map the urban extent. The result of applied machine learning shows that single ML algorithm gives the high accuracy than their ensemble (Liu et al. 2019). Mahmoud et al. (2022) proposed a method to detect the infringements built-up region that sprawl in agricultural, green space areas. The urban built-up area detection and its encroachments in the agriculture area were evaluated using the radar satellite technology comparing with optical Seattleite data. The Sentinel 1A and 1B combining images gave the high accuracy than the sentinel 2 images using the supervised classification. RF algorithm performs well than the classical algorithm to extract the urban footprint. Microwave remote sensing and optical remote sensing data were used to map the urban footprint in Saudi Arabia and compared their accuracy for the best performing data. SAR Microwave and MSI optical data were used to mapping urban footprint and microwave remote sensing data gave the highest accuracy than the optical remote sensing data (Bahrawi and Mohamed 2021). Xu (2007) proposed a technique for the extracting of urban built-up land features using the Landsat TM and ETM images and applied it to two cities in China. Three indices (NDBI, MNDWI, SAVI) were applied to detect the three major urban land use classes. Supervised classification, principal component analysis, and logic calculation were applied to detect built-up area. Peña (2012) performed an experimental study to map and measure the urban growth using Census data, urban land cover, and dissymmetric mapping. How the Census data performed to study the urban growth was analyzed in this study. Khavari et al. (2021) suggested a methodology to translate the high-resolution raster population data into vector-based population clusters to classify the urban/rural classification with high accuracy. The settlement clusters characterize differently indicating population, electrification, rate, and urban–rural categories. Urban mapping in large scale is a very significant challenge using remote sensing because socio-economic feature was not captured correctly. For that the urban scale mapping remote sensing and geolocation data were used to extend the urban mapping combining with night light data, vegetation cover, land surface temperature, population density, accessibility, and road network. Random Forest classification was used to classify the land use and it gave the high accuracy (Xia et al. 2019). Balk et al. (2019) presented gridded estimates of population of urban areas based on the population and settlement type, and other remotely sensed measures of built-up land are of global human settlement layer. Census data and remotely sensed data was cross-classified for the continuum of urban settlement area. Zeng et al. (2013) proposed a multi-criteria and hierarchical evolution system for building extraction from the remote sensing data. The results were validated by the three components like the matched rate, shape similarity, and positional accuracy. Rural built-up area extraction was performed to develop the rural area. The spectral residual method combining with deep neural network was applied to demarcate the rural built up from large-scale remote sensing datasets (Li et al. 2022).

302

P. Barman and Sk. Mustak

14.2 Study Area The city region of Kolkata is the third largest metropolitan area also known as Greater Kolkata. Four municipal corporations and 37 municipalities are included in this area and Kolkata is the main center. Kolkata, north 24 paragons, south 24 paragons, Nadia, Howrah, and Hooghly districts make up the entire city region (Fig. 14.1). This metropolitan region has a population of 14.11 million people, a total area of 1886.67 km2 , while population density 7480 people per square kilometer (Kolkata Metropolitan Development Authority).

14.3 Datasets and Methodology In this study for mapping the urban and rural built up in Kolkata metropolitan area Remote Sensing and socio-economic datasets were used. This dataset was discussed in given below.

14.3.1 Remote Sensing Data For the mapping of rural and urban built up the Land use/cover was prepared in Kolkata metropolitan area. Remote sensing Landsat 5 data of 2011 was used for that. Landsat 5 remote sensing data was collected from the Google Earth Engine web-based platform. Landsat 5 satellite image is characterized by Thematic Mapper (TM) censor. Landsat 5 satellite image band characteristics was mentioned in Table 14.1. The DMSP-OLS night time light data was collected from the National centers for Environmental Information website. The producing Night light data known as average lights was derived from the average visible band digital number (DN) of cloud free light detection multiplied by the percent frequency of light detection (Earth Observation Group—Defense Meteorological Satellite Program, Boulder | Ngdc.Noaa.Gov).

14.3.2 Socio-economic Data To classify the built-up area into rural and urban built-up area socio-economic data was collected. Socio-economic data means Census data of 2011 was downloaded from the Census of India website. District handbook was downloaded for the data prepared from Census of India.

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

303

Fig. 14.1 Shows the study area (Kolkata Metropolitan area)

14.3.3 Methods To classify the urban and rural built up in KMA first land use and land cover map was prepared. From this LULC map built-up area was classified using the census data. LULC napping and built-up area classification were performed using various methods which have discussed below (14.2).

304

P. Barman and Sk. Mustak

Table 14.1 Landsat5 bands characteristics

14.3.3.1

Bands

Spectral bands

B1

Blue

Resolution (meter) 30

B2

Green

30

B3

Red

30

B4

Near-infrared

30

B5

Near-infrared

30

B6

Thermal

B7

Mid-infrared

120 30

Image Pre-processing

First the Landsat 5 image was processed for the land use classification. For the processing of Landsat 5 image first split the image and normalize the seven split bands of Landsat 5 image. The normalization of every bands was done using the following formula.   Image − Imageminimum

 Normalization =  Imagemaximum − Imageminimum

(14.1)

After the normalization the normalized band was stacked and created a stack band set which has classified using the three different algorithms for the Lulc mapping.

14.3.3.2

Feature Extract

Feature extraction was done using the QGIS. Total of 14 feature extraction was done. The NDVI and NDBI were extracted using the following formula.

14.3.3.3

NDVI =

(Band4 − Band3) (Band4 + Band3)

(14.2)

NDBI =

(Band5 − Band4) (Band5 + Band4)

(14.3)

Gray-Level Co-occurrence Matrix (GLCM)

The GLCM features were extracted from six bands of the landsat5 image. The mean and variance features were selected for the Lulc mapping. The GLCM feature was extracted using the QGIS. The six bands of GLCM was extracted because of the similarity in resolution (30 m).

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

14.3.3.4

305

Principal Component Analysis (PCA)

Principal component analysis is a one kind of the statistical unit which was used in frequently data dimension reduction to the data decorrelation (Mudrová and Procházka 2005).

14.3.3.5

Stack Images

The stack image is a created new band set. Using this tool, a new multi-temporal, multi-spatial image was created. In this study the landsat5 bands and extracted features were stacked using QGIS. Also, after the PCA application, the PCA bands were stacked (Fig. 14.2).

Fig. 14.2 Methodological flowchart

306

P. Barman and Sk. Mustak

14.4 Result and Discussion Rural and urban built-up area classification is a very important task using remote sensing. For that built-up area was extracted to classify the land use and land cover in KMA using different machine learning algorithms.

14.4.1 Land Use and Land Cover (LULC) Land use and land cover of KMA was prepared in 2011 using different machine learning algorithm like Random Forest, SVM-Liner, and SVM-RBF. For the land use/cover mapping seven Lulc class was taken like water body, wetland, built-up, agriculture land, fallow land, and open land. The Lulc class was selected on the basis of some previous studies in KMA (Mazumder et al. 2021; Rahaman et al. 2018). The land use and land cover of KMA was analyzed below.

14.4.2 Training and Test Sample For the land use and land cover mapping to continue the algorithm training sample is very important. In this Lulc classification, a total of 508 samples of seven classes were collected. These 508 samples were splited into 70 (348) and 30 (60)% using QGIS. 70% of samples were considered as training samples and the rest 30% were considered as test or validation samples.

14.4.3 Land Use and Land Cover Using Random Forest Land use and land cover of KMA were prepared using Random Forest (RF) algorithm. Three remote sensing images were classified using the RF algorithm to gain the high accuracy for Lulc classification. Normalized Landsat 5 image was classified using the Rf algorithm (Fig. 14.3a). The classified image was tested the accuracy using the test samples and it gave the 89.92% overall accuracy while Kappa hat was 0.87 (Table 14.2). Normalization band of Landsat 5 image and extracted features were stacked in one image. The stack image was classified using Rf machine learning algorithm (Fig. 14.3b). This normalized band and extracted stack image was validated by the test sample and it gives the overall accuracy of 90.71% and Kappa hat 0.88 (Table 14.2). The band and feature stack image were compressed by the PCA and the total

Fig. 14.3 Land use and land cover of KMA using random forest (RF)

(a) Lulc using RF in Landsat 5image (left) (b) Lulc using RF in Landsat image and extracted feature stacked image (middle) (c) LULC using RF in band feature stack using PCA (Right)

14 Mapping Urban Footprint Using Machine Learning and Public Domain … 307

308 Table 14.2 Accuracy assessment of LULC using RF

P. Barman and Sk. Mustak

Sl. No.

Image

Overall accuracy

Kappa hat

1

Band classify

89.92

0.87

2

Band + feature

90.71

0.88

3

Band + features + PCA

85.15

0.76

band characteristics were combined into 4 bands. These four bands were stacked and the stack images were classified using the RF ML algorithm (Fig. 14.3c). The classified image was tested using the test sample for the accuracy analysis. This image gave the overall accuracy of 85.15% and Kappa hat 0.76 (Table 14.2). RF algorithm classification shows that the band and extracted feature stack image gives the highest accuracy than the others and PCA image gives the lowest accuracy. The highest accuracy images for the built-up class producer accuracy (PA) is 89.50% and user accuracy (UA) is 84.57%.

14.4.4 Land Use and Land Cover Using SVM-Linear Land use and land cover of KMA in 2011 were classified using SVM-linear algorithm. Three different characteristics of images were classified using the same algorithm and classify them. The classified images were tested by the test sample to take the accuracy level. First only on the Landsat 5 images SVM-Linear ML algorithm was applied to classify the Lulc (Fig. 14.4a). The classified image was tested, and it gave the overall accuracy 77.71% and Kappa hat was 0.70 (Table 14.3). Landsat 5 and extracted feature stack image was classified using the SVM-Linear algorithm on KMA (Fig. 14.4b). The SVM-Linear classified of the band and feature stack image was done the accuracy assessment and it gave the overall accuracy of 99.23% and Kappa hat was 0.99 (Table 14.3). The stack band and feature image were modified using the PCA method to compress the dimensionality of image. The PCA image was classified using the SVM-Linear method (Fig. 14.4c) and the accuracy assessment was done. It gave the overall accuracy of 58.73% and Kappa hat was 0.45 which is very less (Table 14.3). The highest accuracy was given by the Band and feature stack images than the other two different band characterized images. The highest accuracy image gives the producer accuracy of 99.46% and user accuracy is 99.25%.

Fig. 14.4 Land use and land cover using SVM-linear

(a) Lulc using SVM-Linear in landsat5 image (left) (b) Lulc using SVM-Linear in Landsat5 band and features stack image (middle) (c) Lulc using SVM-Linear in Landsat 5 band feature using PCA stack image (Right)

14 Mapping Urban Footprint Using Machine Learning and Public Domain … 309

310 Table 14.3 Accuracy assessment of Lulc using SVM-Linear

P. Barman and Sk. Mustak

Sl. No.

Image

Overall accuracy

1

Band

77.71

Kappa hat 0.70

2

Band + feature

99.23

0.99

3

Band + feature + PCA

58.73

0.45

14.4.5 Land Use and Land Cover Using SVM-RBF SVM_RBF machine learning algorithm was applied for the Lulc analysis of KMA in 2011. Three different characterized images of KMA were classified. SVM-RBF ML algorithm was applied on Landsat images to classify the Lulc in KMA (Fig. 14.5a) and it gives overall accuracy of 77.55% and Kappa hat 0.67 (Table 14.4). Band and feature stack image were classified by the SVM-RBF ML algorithm (Fig. 14.4b) and it gave the overall accuracy of 99.20% and Kappa hat was 0.98 (Table 14.4). Band and feature stack image were compressed the dimensionality using the PCA method and classify the PCA stack image using the SVM-RBF ML algorithm (Fig. 14.4c). This image gave the overall accuracy of 74.04% and Kappa hat 0.54 (Table 14.4). Using the SVM-RBF algorithm Lulc classify images the band and feature images give the high accuracy than the other two classified images. In the built-up class the highest accuracy Lulc classified image gave the producer accuracy of 99.46% and users accuracy of 99.25%.

14.4.6 Built-Up Classification The built-up area was classified in rural and urban built up in KMA. The rural and urban built-up were extracted from the highest accuracy LULC map of different ML algorithms of KMA. The band and feature stack images gave the highest accuracy of the three algorithms so the rural and urban built-up area was reclassified from these 3 band feature stack images. The LULC map was reclassified into rural and urban built-up. The urban built-up area was divided using the Census of India rural–urban classification definition. Total population above 5000, population density above 400 per sq. km, and non-agriculture worker was applied as a condition to classify the builtup area into rural and urban built-up. In the built-up classification map, four class was selected like water body, wet land, urban built up, and rural built-up. The built-up classification was done by three different algorithms like the built-up classification of KMA using Random Forest algorithm Built-up classification using SVM-linear and built-up area classification using SVM_RBF.

Fig. 14.5 Land use and land cover using SVM-RBF

(a) LULC using SVM-RBF in landsat5 image (left) Figure (b) Lulc using SVM-RBF in Landsat5 band and features stack image (middle) (c) Lulc using SVM-RBF in Landsat 5 band feature using PCA stack image (Right)

14 Mapping Urban Footprint Using Machine Learning and Public Domain … 311

312 Table 14.4 Accuracy assessment of Lulc using SVM-RBF

P. Barman and Sk. Mustak

Sl. No.

Image

Overall accuracy

1

Band

77.55

Kappa hat 0.67

2

Band + feature

99.20

0.98

3

Band + feature + PCA

74.04

0.57

14.4.7 Validation The built-up classification of KMA was validated by the test sample. The test sample of built-up classification was taken from the images by overlapping the administrative boundary of KMA (Fig. 14.7). The urban and rural built-up samples were taken from the urban and rural area based on the administrative unit of KMA, 2011. Total of 104 samples were collected for three different classes. The urban and rural built-up classification (Fig. 14.6) images were validated using these samples. These samples also use for the robustness or validation on GHSL_BUILT_C_MSZ image (Fig. 14.6a). In this GHSL images building height was differentiated. The building height below 6 m was considered as rural area and above 6 m was classified as urban area (Rural Area Development Plan Formulation and Implementation (RADPFI) Guidelines, 2016 Ministry of Panchayati Raj Government of India). The accuracy of urban and rural built-up classification map of different algorithms were mentioned in the Table 14.5. The GHSL image validation accuracy was also mentioned in Table 14.5.

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

313

(a) Settlement characteristics of KMA (upper left), (b) Built-up classification using Random Forest (upper right), (c) Built up classification using SVM-Linear (lower left), (d) Built up classification using SVM-RBF (lower right)

Fig. 14.6 Built-up the classification of KMA

314

P. Barman and Sk. Mustak

Fig. 14.7 Referenced map for the urban and rural built-up validation Table 14.5 Accuracy assessment of built-up area

Sl. No.

Image

Overall accuracy

Kappa hat

1

Random forest

89.50

0.75

2

SVM-linear

96.20

0.92

3

SVM-RBF

96.20

0.92

4

GHSL

71.94

0.57

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

315

14.5 Conclusion The built-up classification in rural and urban built-up area is a very challenging task. The urban extent mapping by the built-up area delineation was done using various remote sensing techniques but the classification of built-up area into rural and urban area using remote sensing is unavailable. So, to classify the urban and rural built-up area in KMA the census rural–urban differentiation condition was used. First the different Machine learning algorithm (RF, SVM-Linear, SVM-RBF) was applied for the Lulc analysis of KMA. Different ML algorithm was applied on different multispectral images of KMA in 2011. Then the different ML algorithms highest accuracy images (Landsat 5/Landsat 5 + feature/Landsat5 + feature (PCA)) were considered for the built-up classification. The Lulc of high accuracy images was reclassified by applying the condition of census rural–urban differentiation definition. In Lulc analysis the Landsat5 bands and extracted feature stack images were performed best accuracy in three different ML algorithms. PCA image was performed with low accuracy in this area in Lulc analysis. The band and feature stack images gave the overall accuracy in Lulc analysis using RF, SVM-Linear, and SVM-RBF, respectively, 90.71, 99.23, and 99.20 and Kappa hat was 0.88, 0.99, and 0.98. In the Lulc analysis, SVM-Linear in the band and feature stack image was performed with high accuracy. The high accuracy images were reclassified into rural and urban built-up area. The built-up area was validated by the test sample. These test samples also validate the GHSL image. The GHSL image gives the overall accuracy of 71.94 and Kappa hat 0.57 (Table 14.5). In this study, the Landsat 5 image was used for the Lulc classification and to classify the built-up area into urban and rural built-up area. The Kolkata metropolitan area is a highly densed and high populated area. In peri-urban area the built-up area was mixed built up developed. The Landsat 5 image was 30-m resolution data so in the peri-urban area the built-up area was not extracted well. In 2021 classification of built-up Sentinel data is available which is 10-m resolution data. It can extract the built-up area very well. So there has scope to classify the rural and urban built-up using census data. In 2021 built-up classification using census data is not possible because of the unavailable of census data now that’s the limitation of this study. GHSL _builtup_C_MSZ image is only available in 2018 so it can face some errors to validate the Built-up area in KMA. The rural–urban built-up classification is very necessary. This information can help policymakers, planners, and local govt. to take actions to develop the area sustainable way.

316

P. Barman and Sk. Mustak

References Bahrawi J, Mohamed NA (2021) Microwave versus optical remote sensing data in urban footprint mapping of the Coastal City of Jeddah, Saudi Arabia. J Indian Soc Remote Sens 49(10):2451– 2466. https://doi.org/10.1007/s12524-021-01403-9 Balk D, Montgomery MR, Engin H, Lin N, Major E, Jones B (2019) Urbanization in India: population and urban classification grids for 2011. Data 4(1):1–16. https://doi.org/10.3390/data40 10035 Census of India 2011 (n.d.) Corbane C, Syrris V, Sabo F, Politis P, Melchiorri M, Pesaresi M, Soille P, Kemper T (2021) Convolutional neural networks for global human settlements mapping from Sentinel-2 satellite imagery. Neural Comput Appl 33(12):6697–6720. https://doi.org/10.1007/s00521-020-05449-7 Earth Observation Group—Defense Meteorological Satellite Progam, Boulder | ngdc.noaa.gov (n.d.). https://ngdc.noaa.gov/eog/dmsp/downloadV4composites.html. Accessed 30 Sept 2022 Esch T, Taubenböck H, Felbier A, Roth A, Dech S (n.d.) The path to mapping the global urban footprint using TanDEM-X data Hacar M (2020) A rule-based approach for generating urban footprint maps: from road network to urban footprint 100–108. https://doi.org/10.26833/ijeg.623592 Henderson V (2003) The urbanization process and economic growth: the so-what question. J Econ Growth 8(1):47–71. https://doi.org/10.1023/A:1022860800744 Jaysawal N, Saha S (2014) Urbanization in India: An impact assessment. Int J App Sociol, 4(2):60– 65. (PDF) Urbanization in India: An impact assessment (n.d.). https://www.researchgate.net/pub lication/264383323_Urbanization_in_India_An_Impact_Assessment. Accessed 28 Sept 2022 Khavari B, Korkovelos A, Sahlberg A, Howells M, Fuso Nerini F (2021) Population cluster data to assess the urban-rural split and electrification in Sub-Saharan Africa. Sci Data, 8(1). https:// doi.org/10.1038/S41597-021-00897-9 Kolkata Metropolitan Development Authority (n.d.). https://kmda.wb.gov.in/. Accessed 29 Sept 2022 Li S, Fu S, Zheng D (2022) Rural built-up area extraction from remote sensing images using spectral residual methods with embedded deep neural network. Sustainability 14(3):1272. https://doi. org/10.3390/SU14031272 Liu X, de Sherbinin A, Zhan Y (2019) Mapping urban extent at large spatial scales using machine learning methods with VIIRS Nighttime light and MODIS daytime NDVI data. Remote Sens 11(10):1–18. https://doi.org/10.3390/rs11101247 Mahmoud N, Samir N, Fathi A & Mohamed W (2022) A proposed methodology for detecting the urban footprint in Egypt. IOP Conference Series: Earth Environ Sci 992(1):012008. https://doi. org/10.1088/1755-1315/992/1/012008 Mazumder S, Saha J, Nandi G, Naskar M, Gayen J, Datta D (2021) Long-term monitoring of cropland transformation in Kolkata Metropolitan Area, India using open-source geospatial technologies. SN Appl Sci 3(1):1–19. https://doi.org/10.1007/S42452-020-04064-4/FIGURES/ 6 Mudrová M, Procházka A (2005) Principal component analysis in image processing Rahaman M, Dutta S, Sahana M, Das DN (2018) Analysing urban sprawl and spatial expansion of Kolkata urban agglomeration using geospatial approach. Appl Challenges Geospat Technol Potential Future Trends 205–221. https://doi.org/10.1007/978-3-319-99882-4_12/COVER Rural Area Development Plan Formulation and Implementation (RADPFI) Guidelines, 2016 Ministry of Panchayati Raj Government of India (n.d.) Science E (2022) A proposed methodology for detecting the urban footprint in Egypt. https://doi. org/10.1088/1755-1315/992/1/012008 Urban/rural definition, Indian district database. (n.d.). http://vanneman.umd.edu/districts/codebook/ defurban.html. Accessed 28 Sept 2022

14 Mapping Urban Footprint Using Machine Learning and Public Domain …

317

Using Census Data, Urban Land-Cover Classification, and Dasymetric Mapping to Measure Urban Growth of the Lower Rio Grande Valley, Texas by Eric Nathaniel Peña A Thesis Presented to the Faculty of the USC Graduate School in Partial Fulfillment of the R (2012) December Wang H, Gong X, Wang B, Deng C, Cao Q (2021) Urban development analysis using built-up area maps based on multiple high-resolution satellite data. Int J Appl Earth Obs Geoinf 103:102500. https://doi.org/10.1016/J.JAG.2021.102500 Xia N, Cheng L, Li MC (2019) Mapping urban areas using a combination of remote sensing and geolocation data. Remote Sens 11(12). https://doi.org/10.3390/rs11121470 Xu H (2007) Extraction of urban built-up land features from landsat imagery using a thematicoriented index combination technique. Photogram Eng Remote Sens 73(12):1381–1391. https:/ /doi.org/10.14358/PERS.73.12.1381 Zeng C, Wang J, Lehrbass B (2013) An evaluation system for building footprint extraction from remotely sensed data. IEEE J Selected Topics Appl Earth Observat Remote Sens 6(3):1640– 1652. https://doi.org/10.1109/JSTARS.2013.2256882