116 52 18MB
English Pages 343 [329] Year 2009
Tropical Circulation Systems and Monsoons
Kshudiram Saha
Tropical Circulation Systems and Monsoons
123
Dr. Kshudiram Saha 4008 Beechwood Road University Park MD 20782 USA [email protected]
ISBN 978-3-642-03372-8 e-ISBN 978-3-642-03373-5 DOI 10.1007/978-3-642-03373-5 Springer Heidelberg Dordrecht London New York Library of Congress Control Number: 2009935567 © Springer-Verlag Berlin Heidelberg 2010 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, 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. Cover design: deblik, Berlin Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Dedicated with gratitude to my beloved wife, Usha
Preface
The present book is a sequel to the author’s first book entitled ‘The Earth’s Atmosphere – Its Physics and Dynamics’, published in 2008, and addresses the practical side of the subject with stress on proper analysis and interpretation of tropical data. It deals specifically with circulation systems that are excited by heat sources and sinks in the Earth-atmosphere system over the tropics and extratropics. Together, the two volumes are intended to form a fairly comprehensive course of study of tropical meteorology, in theory and practice. A few new ideas, especially on monsoons, though empirically derived from data and analysis, would appear to be intuitionally born out of nowhere and are put forward by the author in the hope that they will be verified by practicing meteorologists in the years to come. The book is divided into three parts: the first deals with some general aspects of tropical general circulations and their disturbances; the second with monsoons over seven regions of the tropics, and the third with monsoons over Central America and the extratropical belt of North America. Part I, in dealing with the General circulation of the atmosphere in the tropics and its large-scale perturbations, introduces the Hadley, Walker and Monsoon circulations in a unified system. It also introduces the synoptic-scale tropical disturbances, such as Easterly waves, monsoon troughs, depressions and cyclones and gives a detailed account of their formation, structure, development and movement over different parts of the globe. Part II is devoted wholly to Monsoons in different parts of the tropics. The regions covered are Southern Asia, Eastern Asia, Maritime Continent, Australia, North and South Africa, and South America. Part III deals with monsoon over Central America and adjoining Southwestern North America as well as monsoons over North America. Monsoon is one area where the book advances several new ideas and concepts. It offers an entirely new definition of monsoon which allows monsoon to be detected and identified over several hitherto-unsuspected regions of the globe. It unveils the wave structure of the monsoon and identifies the areas to be associated with organized clouding and rainfall. Among the other important issues discussed in the book are the origin of African wave disturbances which later in their life cycle develop into Atlantic hurricanes, Low-Level Jets over several parts of the globe, and monsoon lows and depressions over South America, and formation of tropical cyclones over the South Atlantic Ocean. vii
viii
Preface
Written in an easy-to-understand style, with minimal mathematics, the book is intended for use by students pursuing courses in atmospheric science at undergraduate and graduate levels in colleges and universities and by scientists of other disciplines interested in tropical meteorology. It is likely to appeal to a large circle of readers interested in studies of tropical monsoons. Practicing meteorologists may find it particularly useful and challenging while comparing the results of their analysis and diagnosis of tropical circulation systems with the empirical and conceptual models suggested in the book. Inspite of the utmost care taken in the preparation of the book, it is likely that there have been errors and omissions. The author will be thankful if such lapses are brought to his notice. University Park, Maryland June 15, 2009
Kshudiram Saha
Acknowledgments
The author is grateful to his family, especially his daughters Manjushri and Suranjana, for helping him throughout the preparation of this book. Manjushri typed the long list of references. Suranjana who is co-author of many of the papers referred to in the book helped not only with insertion of nearly 150 diagrams in the book, but also along with her husband Dr. Huug van den Dool provided all the logistic support and technical facilities. Without their wholehearted support, it would have been impossible to complete this work. Huug’s review of the first draft of most of the chapters of the book was very helpful in improving the manuscript. The author’s special thanks are due to the National Centers for Environmental Prediction (NCEP) of the National Weather Service, USA, for several of their data and analysis products incorporated in the book. He expresses his deep gratitude to the numerous authors, publishers and learned Societies, who permitted him to reproduce diagrams and excerpts from their published work. Finally, the author would like to thank all the publishing staff at Springer-Verlag, involved in the book project, for their unfailing courtesy, active cooperation, and helpful attitude.
ix
Contents
Part I
Tropical Circulation Systems – A Survey
1 Large-Scale Tropical Circulations – Some General Aspects . . . . 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Tropical Circulation as Part of the Global General Circulation – The Tradewinds . . . . . . . . . . . . . . . . . 1.3 Poleward Boundary of the Tropical Circulation . . . . . . . . 1.4 Heat Sources and Sinks . . . . . . . . . . . . . . . . . . . . . 1.4.1 Definition of Heat Sources/Sinks . . . . . . . . . . . 1.4.2 Diabatic/Adiabatic Heat Sources/Sinks . . . . . . . . 1.5 Some Physical and Dynamical Constraints and Conservation Laws . . . . . . . . . . . . . . . . . . . . . 1.5.1 Direct and Indirect Circulations . . . . . . . . . . . . 1.5.2 Energy Transformations . . . . . . . . . . . . . . . . 1.5.3 Energy Transfer Process – Carnot’s Cycle . . . . . . 1.5.4 Conditional Instability and Convection . . . . . . . . 1.5.5 Cellular Structure – Shallow and Deep Convection . 1.5.6 Coriolis Control-Variation with Latitude . . . . . . . 1.5.7 Conservation Laws . . . . . . . . . . . . . . . . . . 1.6 Equatorial Circulations . . . . . . . . . . . . . . . . . . . . . 1.6.1 Circulation with Heat Sources and Sinks Placed Alternately Along the Equator – Walker Circulations 1.7 Meridional Circulation with Heat Source at the Equator and Heat Sinks in Higher Latitudes – The Hadley Circulations 1.8 Seasonal Migration of the Equatorial Heat Source . . . . . . . 1.8.1 Origin of Monsoon . . . . . . . . . . . . . . . . . . 1.8.2 The Wave Structure . . . . . . . . . . . . . . . . . . 1.8.3 Forcing for the Seasonal Movement of the Equatorial Heat Source . . . . . . . . . . . . . . . . 1.8.4 Intraseasonal Oscillation of Monsoon . . . . . . . . 1.9 Co-existence of Monsoon, Hadley and Walker Circulations – Inclined Troughs . . . . . . . . . . . . . . . . 1.10 Definition of Monsoon . . . . . . . . . . . . . . . . . . . . .
. .
3 3
. . . . .
4 6 7 7 8
. . . . . . . . .
9 9 10 11 11 12 12 13 14
.
14
. . . .
15 17 17 18
. .
19 21
. .
22 24 xi
xii
Contents
1.11
1.12
Global and Regional Distribution of Monsoons . . 1.11.1 Tropical Monsoons . . . . . . . . . . . . 1.11.2 Extratropical Monsoons . . . . . . . . . . 1.11.3 Zonal and Meridional Anomalies . . . . . Co-existence of Monsoon with Desert Circulation .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
2 Tropical Disturbances (Quasi-stationary Waves, Easterly/Westerly Waves, Lows and Depressions, Cyclonic Storms, and Meso-Scale Disturbances) . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Quasi-stationary Waves . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Quasi-stationary Wave in Wind Field . . . . . . . . . . 2.2.2 Quasi-stationary Wave in Temperature Fields . . . . . 2.2.3 Structure of the Quasi-stationary Wave in Circulations . 2.3 Traveling Easterly (E’ly) Waves . . . . . . . . . . . . . . . . . 2.3.1 Easterly Waves in Tropical North Atlantic . . . . . . . 2.3.2 Easterly Waves in Tropical North Pacific . . . . . . . . 2.3.3 Easterly Waves in the Indian Ocean Region . . . . . . 2.4 Development of Waves . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Meaning of Development . . . . . . . . . . . . . . . . 2.4.2 Development of a Quasi-stationary Monsoon Trough into a Depression . . . . . . . . . . . . . . . . 2.4.3 Development of a Depression into a Cyclonic Storm/Tropical Cyclone/Hurricane/Typhoon . . . . . . 2.5 Meso-Scale Disturbances and Severe Local Storms in the Tropics 2.5.1 General Considerations – Source of Energy of the Storm . . . . . . . . . . . . . . . . . . . . . . . . 2.5.2 Thunderstorms . . . . . . . . . . . . . . . . . . . . . 2.5.3 Hailstorms . . . . . . . . . . . . . . . . . . . . . . . . 2.5.4 Tornadoes . . . . . . . . . . . . . . . . . . . . . . . . 3 Tropical Cyclones/Hurricanes/Typhoons – Their Structure and Properties . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Observed Structure of a Tropical Cyclone . . . . . . . . 3.2.1 Wind Structure . . . . . . . . . . . . . . . . . 3.2.2 Radial and Tangential Components of the Wind 3.2.3 Vertical Motion in a Mean Typhoon . . . . . . 3.2.4 Pressure Distribution . . . . . . . . . . . . . . 3.2.5 Temperature Distribution . . . . . . . . . . . . 3.3 The Eye and the Eye-Wall . . . . . . . . . . . . . . . . 3.3.1 General Considerations – Formation of the Hurricane Eye . . . . . . . . . . . . . . . . . . 3.3.2 Circulation Inside the Hurricane Eye – Evidence of Meso-Scale Vortices . . . . 3.3.3 Concentric Multiple Eye-Walls . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
25 25 26 27 29
33 33 34 35 36 37 39 39 40 41 43 43 45 48 50 50 51 55 57
. . . . . . . . .
61 61 61 63 64 67 68 69 71
. . . .
71
. . . . . . . .
73 75
Contents
3.4
3.5
3.6
Part II
xiii
Spiral Bands Around the Eye-Wall . . . . . . . 3.4.1 Structure . . . . . . . . . . . . . . . . 3.4.2 Origin and Direction of Propagation . Storm Surge . . . . . . . . . . . . . . . . . . 3.5.1 Introduction . . . . . . . . . . . . . . 3.5.2 Some General Aspects of Storm Surge 3.5.3 Mathematical Models of Storm Surge Prediction of Cyclone Track and Intensity . . . 3.6.1 Early Models – The Steering Concept 3.6.2 Current Models . . . . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
77 77 78 78 78 79 80 82 83 83
Tropical Monsoons over Continents and Oceans
4 Monsoon over Southern Asia (Comprising Pakistan, India, Bangladesh, Myanmar and Countries of Southeastern Asia) and Adjoining Indian Ocean (Region – I) . . . . . . . . . . . 4.1 Introduction – Physical Features and Climate . . . . . . . . . . 4.2 The Winter Season (December–February) . . . . . . . . . . . . 4.2.1 General Climatic Conditions . . . . . . . . . . . . . . 4.2.2 Disturbances of the Winter Season . . . . . . . . . . . 4.3 The Transition Season (March–May) . . . . . . . . . . . . . . 4.3.1 Western Disturbances . . . . . . . . . . . . . . . . . . 4.3.2 ‘Heat Lows’ over Land and ‘Cold Highs’ over Ocean . 4.3.3 Severe Local Storms . . . . . . . . . . . . . . . . . . 4.3.4 Developments over the Equatorial Indian Ocean . . . . 4.4 Advance of Summer Monsoon to the Indian Subcontinent – General Remarks . . . . . . . . . . . . . . . . . 4.4.1 Advance over the Indian Ocean (April–June) – Stage 1 4.4.2 Onset over the Indian Subcontinent (June–July) – Stage 2 . . . . . . . . . . . . . . . . . . 4.4.3 Advance to Western Himalayas (July–August) – Stage 3 4.4.4 Source of Moisture for Monsoon Rainfall . . . . . . . 4.5 Disturbances of the Summer Monsoon during the Onset Phase . 4.5.1 Onset Vortex over the Arabian Sea and the Bay of Bengal . . . . . . . . . . . . . . . . . . . . . . . . 4.5.2 Monsoon Depressions and Cyclonic Storms . . . . . . 4.5.3 Interaction of Monsoon with W’ly Waves . . . . . . . 4.6 Rainfall over the Indian Subcontinent during the Onset Phase . 4.7 Summer Monsoon – Withdrawal Phase (September–November) 4.7.1 Dates of Withdrawal of Monsoon . . . . . . . . . . . . 4.7.2 Retreating Monsoon Rain over Tamil Nadu . . . . . . 4.7.3 Disturbances of the Withdrawal Phase . . . . . . . . . 4.8 Variability of the Indian Summer Monsoon Rainfall . . . . . . . 4.8.1 Interannual Variability . . . . . . . . . . . . . . . . . 4.8.2 Factors Likely Responsible for Interannual Variability . 4.8.3 Intraseasonal Variability . . . . . . . . . . . . . . . . .
89 89 91 91 92 93 94 94 94 95 98 102 104 105 106 107 107 107 108 110 111 111 111 113 115 115 117 121
xiv
Contents
5 Monsoon over Eastern Asia (Including China, Japan, and Korea) and Adjoining Western Pacific Ocean . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Physical Features and Climate . . . . . . . . . . . . . . . . . 5.3 The Winter Season over Eastern Asia (November–March) . . 5.3.1 Temperature, Pressure, and Wind . . . . . . . . . . . 5.3.2 Quasi-stationary Wave in Westerlies – Its Interaction with Traveling Waves – Cold Surges . . . 5.3.3 Winter Rainfall over Eastern Asia . . . . . . . . . . 5.4 Airmass Transformations and Cyclogenesis over the Oceans . 5.4.1 Cyclonic Disturbances over Eastern Asia and Neighboring Ocean . . . . . . . . . . . . . . . . 5.5 Transition Period (April) . . . . . . . . . . . . . . . . . . . . 5.5.1 Development of ‘Heat Low’ over Eastern Asia . . . . 5.6 Origin of Monsoon over Eastern Asia . . . . . . . . . . . . . 5.7 Seasonal March of the Summer Monsoon . . . . . . . . . . . 5.8 Stationary States and Jumps . . . . . . . . . . . . . . . . . . 5.9 Meteorological Developments Associated with the Jump to Central China . . . . . . . . . . . . . . . . . . . . . . . . 5.9.1 Tibetan Plateau Monsoon . . . . . . . . . . . . . . . 5.9.2 The Meiyu (Plum Rain) Front over China . . . . . . 5.10 Jump of East Asian Monsoon to Extratropical Latitudes . . . 5.10.1 Evidence of Jump in Climatological Fields . . . . . . 5.10.2 Zonal Anomaly in Seasonal Variations . . . . . . . . 5.10.3 Climatological Rainfall over Eastern Asia During July . . . . . . . . . . . . . . . . . . . . . . 5.11 Monsoon over Japan . . . . . . . . . . . . . . . . . . . . . . 5.11.1 Geographical Location and Climate . . . . . . . . . 5.11.2 The Baiu Front – Its Seasonal Movement and Activity 5.12 Monsoon over Korea . . . . . . . . . . . . . . . . . . . . . . 5.12.1 Historical Background . . . . . . . . . . . . . . . . 5.12.2 Physical Features and Climate . . . . . . . . . . . . 5.12.3 Winter Monsoon over Korea . . . . . . . . . . . . . 5.12.4 Summer Monsoon over Korea – Changma Season . . 5.12.5 Korea’s Climatic Zones (After McCune, 1941) . . . . 6 Meteorology of the Maritime Continent (Region – III) (Comprising Philippines, Indonesia and Equatorial Western Pacific Ocean) . . . . . . . . . . . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . 6.2 Climate of the Maritime Continent . . . . . . . . . 6.2.1 Pressure . . . . . . . . . . . . . . . . . . 6.2.2 Temperature . . . . . . . . . . . . . . . . 6.2.3 Relative Humidity and Cloudiness . . . . 6.2.4 Rainfall . . . . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . .
123 123 124 125 125
. . .
128 130 130
. . . . . .
130 133 134 134 136 137
. . . . . .
137 138 139 142 143 145
. . . . . . . . . .
146 147 147 147 150 150 150 150 151 152
. . . . . . .
155 155 156 156 157 158 158
Contents
6.3
xv
. . . . . .
. . . . . .
. . . . . .
. . . . . .
159 159 160 160 165 168
7 Monsoon over Australia (Region – IV) . . . . . . . . . . . . . 7.1 Introduction – Location and Physical Features . . . . . . . 7.2 Early Studies . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Climate of Australia and Surrounding Oceans . . . . . . . 7.3.1 Ocean Surface Temperature (SST, C) . . . . . . . 7.3.2 Air Temperatures . . . . . . . . . . . . . . . . . 7.3.3 Atmospheric Pressure (Isobaric Height) . . . . . 7.3.4 Wind and Circulation . . . . . . . . . . . . . . . 7.4 Monsoon over Australia . . . . . . . . . . . . . . . . . . 7.4.1 Onset of Monsoon . . . . . . . . . . . . . . . . . 7.4.2 Co-existence of Monsoon and Hadley Circulations – Interhemispheric Movement . . . . 7.4.3 Summer Monsoon Rainfall over Australia . . . . 7.5 Annual Rainfall of Australia and Its Seasonal Variability . 7.5.1 Annual Rainfall . . . . . . . . . . . . . . . . . . 7.5.2 Seasonal Variability . . . . . . . . . . . . . . . . 7.6 Variability of Australian Rainfall with ENSO . . . . . . . 7.7 Tropical Disturbances in the Australian Region – Depressions and Cyclones . . . . . . . . . . . . . . . . . 7.8 Tropical-Midlatitude Interaction in the Australian Region . 7.8.1 Northerly and Southerly Bursters . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
171 171 172 173 173 174 176 177 182 182
. . . . . .
. . . . . .
. . . . . .
183 185 187 187 188 189
. . . . . . . . .
189 191 192
6.4 6.5
Factors Affecting the Climate of the Maritime Continent 6.3.1 Geographical Location and Topography . . . . 6.3.2 Ocean Currents . . . . . . . . . . . . . . . . . 6.3.3 Equatorial Trough, the ITCZ and Monsoons . . The Maritime Continent – A Heat Source . . . . . . . . The Maritime Continent and the ENSO . . . . . . . . .
8 Monsoon over Africa (Region – V) . . . . . . . . . . . . . . . . . 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Physical Features and Environment . . . . . . . . . . . . . . 8.3 Climates of Africa and Surrounding Oceans . . . . . . . . . . 8.3.1 Sea Surface Temperature and Wind . . . . . . . . . . 8.3.2 Air Temperature . . . . . . . . . . . . . . . . . . . . 8.3.3 Isobaric Height (gpm) . . . . . . . . . . . . . . . . . 8.3.4 Wind and Circulation . . . . . . . . . . . . . . . . . 8.3.5 Rainfall over Africa . . . . . . . . . . . . . . . . . . 8.4 Equatorial Westerlies over Africa . . . . . . . . . . . . . . . 8.5 The Equatorial Trough over North Africa – Its Zonal Anomaly 8.6 Structure of the Circulation Associated with the Equatorial Trough . . . . . . . . . . . . . . . . . . . . . . . 8.6.1 Zonal Circulation . . . . . . . . . . . . . . . . . . . 8.6.2 Meridional Circulation . . . . . . . . . . . . . . . . 8.7 Origin of African Wave Disturbances . . . . . . . . . . . . . 8.7.1 Early Studies . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . .
195 195 196 198 198 198 199 201 205 207 209
. . . . .
210 211 211 212 212
xvi
Contents
. . . .
213 217 219 219
. . . .
220
9 Monsoon over South America (Region – VI) . . . . . . . . . . . . . 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Physical Features and Environment . . . . . . . . . . . . . . . 9.2.1 Physical Dimension of the Continent . . . . . . . . . . 9.2.2 Topography . . . . . . . . . . . . . . . . . . . . . . . 9.2.3 Oceanic Environment and Its Influence on Climate . . 9.3 Climatological Features . . . . . . . . . . . . . . . . . . . . . 9.3.1 Air Temperature and Pressure . . . . . . . . . . . . . . 9.3.2 Atmospheric Circulation – Monsoon . . . . . . . . . . 9.3.3 Co-existence of Monsoon and Hadley Circulations . . 9.3.4 Rainfall over South America . . . . . . . . . . . . . . 9.4 Quasi-stationary Waves and Their Associated Weather . . . . . 9.4.1 Weather Phenomena Related to the Northern Boundary 9.4.2 Weather Phenomena Associated with the Southern Boundary . . . . . . . . . . . . . . . . . . . 9.5 Tropical Disturbances over South America . . . . . . . . . . . 9.5.1 Types of Disturbances . . . . . . . . . . . . . . . . . . 9.5.2 Monsoon Lows and Depressions . . . . . . . . . . . . 9.5.3 Upper-Tropospheric Cyclonic Vortices . . . . . . . . . 9.6 A Tropical Cyclone over the South Atlantic Ocean . . . . . . . 9.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 9.6.2 Formation of the Initial Vortex – Interaction with W’ly Waves . . . . . . . . . . . . . . . . . . . . 9.6.3 Structure, Movement and Development of the Vortex .
223 223 224 224 225 225 227 227 230 233 235 236 238
8.8 8.9
8.7.2 Influence of Midlatitude Forcing . . . . . . . . 8.7.3 Sudan – The Breeding Ground . . . . . . . . . 8.7.4 Role of Topography . . . . . . . . . . . . . . . Structure, Development and Movement of the Waves . . Interaction of South African Monsoon with Midlatitude Waves of the Southern Hemisphere . . . . . . . . . . .
. . . .
. . . .
. . . .
240 241 241 242 243 248 248 250 251
Part III Extratropical Monsoons 10
Monsoon over Central America and Adjoining Southwestern North America (Region – VII) . . . . . . . . . . . . . 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Heat Sources and Sinks and Their Seasonal Movement . . . . . 10.3 The Climate of Central America and Adjoining North America . 10.3.1 Surface Temperatures and Winds . . . . . . . . . . . . 10.3.2 Upper Air Temperatures . . . . . . . . . . . . . . . . 10.3.3 Upper Air Height (gpm) . . . . . . . . . . . . . . . . . 10.3.4 Upper Air Wind Field and Circulation . . . . . . . . . 10.4 Rainfall over Central America and Adjoining Areas . . . . . . . 10.4.1 Annual Rainfall over Mexico . . . . . . . . . . . . . .
255 255 256 257 257 259 260 262 264 267
Contents
xvii
10.4.2 10.5
11
Source of Moisture for Rainfall over the Arizona-Sonoran Desert . . . . . . . . . . . . . . . . . Some Characteristic Features of Weather over Central America . 10.5.1 Weather Associated with W’ly Waves . . . . . . . . . 10.5.2 Weather Associated with ‘Northers’ . . . . . . . . . . 10.5.3 Land and Sea Breezes on the Pacific Coast of Mexico . 10.5.4 Temporales of the Caribbean Sea and the Gulf of Mexico . . . . . . . . . . . . . . . . . . . . . . . . 10.5.5 Hurricanes and Tropical Storms . . . . . . . . . . . .
Extratropical Monsoon over North America . . . . . . . . . . . . . 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Climatological Background of North American Monsoon . . . . 11.2.1 Physical Features of the Land . . . . . . . . . . . . . . 11.2.2 Semi-permanent High and Low Pressure Systems over Oceans . . . . . . . . . . . . . . . . . . 11.3 The Seasonal Movement of Heat Sources and Sinks . . . . . . . 11.4 Seasonal Circulations – Monsoons . . . . . . . . . . . . . . . . 11.4.1 The Winter Monsoon (December–February) . . . . . . 11.4.2 The Spring Transition Season (March–May) . . . . . . 11.4.3 The Summer Monsoon (June–August) . . . . . . . . . 11.4.4 The Autumn Transition Season (September–November) 11.5 Interaction of Monsoons with W’ly Wave Disturbances . . . . . 11.6 Some Characteristic Features of East Coast Monsoon . . . . . . 11.6.1 Seasonal Variations and Reversals . . . . . . . . . . . 11.6.2 Monsoonal Characteristics of the East Coast Region . . 11.7 Role of the Appalachian Mountain Range – Leeside Cyclogenesis – Northeast Storms . . . . . . . . . . . . . . . . 11.8 Interaction of Monsoon with Storms and Hurricanes . . . . . .
268 270 270 270 271 272 272 275 275 275 276 277 278 279 279 281 281 286 287 287 287 288 288 289
Appendix: Meanings of Uncommon Words/Terms Used in the Book . .
291
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
299
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
313
Guide to Numbering Figures and Equations
Figures are numbered serially chapterwise. For example, Fig. 5.12 is diagram No. 12 of Chapter 5. Equations are numbered serially, sectionwise and chapterwise. For example, Eq. (5.12.4) is Equation No. 4 in Section 12 of Chapter 5.
xix
Part I
Tropical Circulation Systems – A Survey
Chapter 1
Large-Scale Tropical Circulations – Some General Aspects
1.1 Introduction Meteorology in general and tropical meteorology in particular have made tremendous strides during the last half a century or so after the introduction of computer technology and space satellites. Undeniably, we now know much more about our atmosphere and its behaviour than ever before. It is now well established that tropical circulation forms an integral part of the global general circulation and that there is a continual exchange of heat, momentum and moisture between the tropics and the rest of the atmosphere. Yet, tropical circulation has some distinctive characteristics of its own which need to be identified and studied independently in a more comprehensive manner. At present, uncertainty prevails in several areas of interest and there are many dark or twilight areas. In a seminal paper, Charney and Shukla (1981) had observed that quite unlike the midlatitude weather systems the predictability of which was limited by short period baroclinic wave activity, tropical weather systems which are determined by circulations between long-period heat sources and sinks are more predictable. Surface and boundary layer characteristics, such as ocean surface temperature, seasonal ground heating and cooling, have much longer lifespans and as such amenable to prediction over longer periods of time. It is proposed to show in the present text that tropical circulation systems are basically forced by boundary layer heat sources and sinks, on different time and space scales, though wave activity arising from flow instability plays a significant role in weather-forming processes. However, several questions regarding tropical circulations remain unresolved. For example, with a heat source centered at the equator and flanked by heat sinks in both zonal and meridional directions, what sort of circulations and weather patterns would evolve over the equatorial region? In this area, we haven’t gone far beyond what we learnt from the excellent theoretical work of Matsuno (1966), Gill (1980, 1982) and others. An uncertainty shrouds the existence or non-existence of double equatorial troughs in some oceans and continents. For that matter, what is the origin of the SW Pacific Convergence Zone (SPCZ), or the SW Atlantic Convergence Zone (SACZ)? Are they to be regarded as equatorial, tropical, subtropical or extratropical convergence zones? Monsoon is probably one area where our knowledge K. Saha, Tropical Circulation Systems and Monsoons, C Springer-Verlag Berlin Heidelberg 2010 DOI 10.1007/978-3-642-03373-5_1,
3
4
1 Large-Scale Tropical Circulations – Some General Aspects
and understanding lack precision. Firstly, its definition and then the manner of its advance and retreat over different parts of the globe! In the past, monsoon has been defined almost exclusively in terms of either rainfall or seasonal reversal of the prevailing surface wind. Little is known of the structure of monsoon and the manner of its advance and retreat in different areas. Notwithstanding commendable progress made in studies of monsoon (for a recent review of literature, see Webster et al., 1998) over the globe in general, we yet do not have a clear idea of its structure and how it advances or retreats over the Arabian Sea or the Bay of Bengal, for example, during the northern summer. What brings up the monsoon current over the Arabian Sea which remains relatively cold compared to its surrounding land areas during summer? Then, again, what brings about the well-known intraseasonal oscillation in monsoon rainfall and other meteorological parameters? In some parts of the tropics, the equatorial trough of low pressure which changes its orientation with season shows strong correlation with the genesis of tropical cyclones/hurricanes/typhoons. What is the connection between the two? Another area where we lack precise information is how large mountain systems, such as the Andes in South America, the Rockies in North America, the Mountains of East Africa and the Great Himalaya complex with Tibetan Plateau in Asia affect circulation systems over the respective continents. Our knowledge of the effect of ocean currents and ocean surface temperatures on equatorial circulation is also inadequate. Here we have not gone far beyond what we learnt from the pioneering studies of Walker (1924), Bjerknes (1966, 1969) and others. There are many such areas where we need further and better information from observations and their analysis and diagnosis, along with theoretical and dynamical studies, than heretofore. In a lecture delivered during a seminar in 1979, Charney had observed that ‘Data by themselves are not sufficient, nor is mind. We need a combination of both’. A summary of this lecture was prepared by J. Shukla (1979) from notes taken by him. To-day, equipped with much better network of surface and upper-air observations and improved methods of data analysis, we have much better opportunity to have a fresh look into some of the above-mentioned outstanding problems of tropical meteorology and apply our mind to them than ever before. Frankly, that is the main objective of this book.
1.2 Tropical Circulation as Part of the Global General Circulation – The Tradewinds From time immemorial, mariners sailing over oceans in low latitudes for trading purposes encountered a set of highly-steady seasonally-reversing winds which they called the tradewinds and which they used to great advantage in moving from place to place. In late 17th century, Edmund Halley (1686) was the first to make a detailed study of the tradewinds with the data then available and hypothesized that the observed
1.2
Tropical Circulation as Part of the Global General Circulation – The Tradewinds
5
winds at the surface were part of a direct thermally-driven circulation between a heat source and a heat sink, which reversed its direction between winter and summer But it was Hadley (1735) who offered an explanation for the cause of the tradewinds as well as their observed reversal of direction on the basis of differential heating between the equator and the poles and the rotation of the earth. He argued that a general equatorward drift of the tradewinds at low levels required a compensating poleward drift at high levels in order to prevent an undue accumulation of mass near the equator. Further, a general westward drag by the tradewinds near the earth’s surface at low latitudes due to the rotation of the earth required a compensating eastward drag by the westerlies in high latitudes so as to prevent a general slowing down of the earth. It was found later that the general westward or eastward drift of the wind was consistent with the principle of conservation of absolute angular momentum of the earth. A parcel of air moving equatorward from high latitudes in order to conserve the angular momentum of its original latitude would acquire an increasingly westward drift, while a poleward-moving parcel would acquire an increasingly eastward drift. This was due to the fact that the earth’s surface moved faster at the equator than at higher latitudes. Hadley’s idealized single-cell circulation model held ground and went unchallenged for nearly a century and it was once thought that Hadley’s model was representative of mean meridional circulation over all parts of the globe at all times of the year. However, later observations called for a modification of Hadley’s idealized single-cell model. The new observations revealed the presence of a well-marked high pressure belt over the subtropics and a low pressure belt further poleward near 60◦ latitude, which suggested a meridional pressure gradient and a poleward drift of air, instead of an equatorward drift near surface, and a compensating equatorward drift at some height, over the midlatitudes. Further, the westerly wind over the midlatitudes were found to be baroclinically unstable and characterized by large-scale eddy motion. Amongst the early attempts to modify Hadley’s original scheme were those of Thomson (1857) and Ferrel (1859) who introduced a shallow indirect cell, characterized by a poleward flow near the surface and equatorward flow at some height, over the midlatitudes, within the framework of the idealized single-cell Hadley circulation model. Further modifications to the meridional circulation model were made in the light of later observations. That tropical circulation forms a distinct integral part of a three-cell global general circulation model was first spelt out by Rossby (1947) and others almost two centuries after Hadley (1735) had proposed his single-cell poleto-equator zonally-averaged annual-mean general circulation model. The model proposed by Rossby (1947) and others, is in Fig. 1.1. Rossby’s three-cell meridional circulation model shows a direct circulation cell over the tropical belt with rising motion over the equatorial region and sinking motion over the subtropical belt, an indirect circulation cell over the midlatitudes and a direct circulation cell over the polar latitudes, with a polar front located at a latitude of about 60◦ . The Rossby model has, by and large, stood the test of time
6
1 Large-Scale Tropical Circulations – Some General Aspects
Fig. 1.1 Schematic of a three-cell meridional circulation model proposed by Rossby (1947) and others
and found general acceptance by the scientific community to this day. In this model, the classical Hadley circulation has been shown to have its rightful place over the tropical belt only, as against the original model with a pole-to-equator circulation. In the Rossby model, the tradewinds blowing towards the equator pick up increasing amounts of heat and moisture from the warm ocean surface and deliver them in rising currents over the equatorial belt to form cloud and precipitation and then turn around in the upper troposphere to blow as anti-trades towards the subtropical belt where they subside to low levels to blow as low-level trades or tradewinds again. This constitutes the direct kinetic energy-producing tradewind circulation cell of low latitudes shown in the Rossby model.
1.3 Poleward Boundary of the Tropical Circulation In the past, there appears to have been some controversy and divergence of opinion regarding the definition of tropics, particularly its poleward boundary. Climatologists divide climates on the earth’s surface into three distinct latitudinal belts, calling that which lies within the first 30◦ parallels of latitude, the tropics or the hot climate zone, as distinct from the other 30-degree belts which lie poleward with temperate and frigid climates. Defined in this way, the tropics is the hottest part of the earth’s surface with the sun always shining overhead somewhere over the region. Some people even put the poleward boundary of the tropics at the solstices, that is at 231/2◦ of latitude north and south of the equator. However, meteorologists, by and large, are aware that the poleward boundary of the tropical circulation is not determined by any fixed latitude, since it meanders about its annual mean location considerably with longitude and season. To them, the ridge of the global subtropical high pressure belt at surface, which divides the easterly tradewinds from the midlatitude westerly winds, would appear to be a more reasonable and acceptable choice. That this last is the correct view is also suggested by a three-dimensional examination of the structure of the tropical atmosphere vis-à-vis that of the higher
1.4
Heat Sources and Sinks
7
latitudes. Besides surface temperatures, the vertical distribution of temperature over the two belts is such as to create a fairly sharp thermal and circulation divide between the tropics and the midlatitudes, which creates a strong horizontal temperature gradient across a narrow boundary zone to drive a westerly wind maximum known as the subtropical jet (STJ) along the poleward boundary of the tropics. One of the earliest to find the existence of this jet was Palmen (1951) who studied the wintertime mean meridional circulation on the earth’s surface. Palmen’s finding has been substantiated by several other studies (e.g., Krishnamurti, 1961; Wallace and Hobbs, 1977; Galvin, 2007). According to observations, the boundary moves equatorward of its annual-mean latitude during the winter and poleward during the summer in a hemisphere. So, the tropics of our concern in the present text will be the latitudinal belt between the subtropical belts of the two hemispheres, bounded in each hemisphere by the ridge of the subtropical high pressure at surface and the subtropical jet in the upper air. The tropical circulation, however, is not isolated. In some parts of the globe, a tropical circulation system may extend to high extratropical latitudes, or an extratropical circulation system extend to very low tropical latitudes, and there is continual interaction and exchange of heat, mass, and momentum between the two circulations.
1.4 Heat Sources and Sinks Before proceeding further, we need to define what we mean by heat sources and sinks and state how we can identify them in the atmosphere.
1.4.1 Definition of Heat Sources/Sinks At any given time of the year, when the atmosphere over a place is continually heated (cooled) relative to its surrounding, the differential heating creates a heat source (sink) over the place. We define a heat source or a heat sink by the following criteria: dH/dt = 0,
∇ 2 dH/dt < 0,
Heat source
(1.4.1)
dH/dt = 0,
∇ 2 dH/dt > 0,
Heat sink
(1.4.2)
Where H denotes the steady-state heat content of the air at a given place at time t, and ∇ is a Del operator. Here, H = ρcp T, where ρ is density, cp is specific heat at constant pressure p, and T is temperature determined by local heat balance. Thus, to maintain a steady-state, a heat source must give up its excess heat to the environment, and a heat sink must make up its deficit heat by acquiring heat from the environment.
8
1 Large-Scale Tropical Circulations – Some General Aspects
1.4.2 Diabatic/Adiabatic Heat Sources/Sinks 1.4.2.1 Diabatic Heating/Cooling A diabatic heating or cooling process is one in which a working sample of the atmosphere is free to exchange heat with its environment. There are three important processes of diabatic heating or cooling in the atmosphere in the tropics. These are: (i) Absorption or emission of short- and long-wave radiation; (ii) Latent heat released by condensation of water vapour, or lost through evaporation of water, and mixing; and (iii) Sensible and latent heat gained from, or lost to, a boundary surface through turbulence and convection.
1.4.2.2 Adiabatic Heating/Cooling In an adiabatic process, the air sample has to work in a closed system and is barred from sharing its heat with the environment. This means that any change that occurs in its temperature is due to its own expansion or compression. In this process, air with upward velocity cools by expansion to a lower pressure and that with downward velocity warms up by compression to a higher pressure. 1.4.2.3 Temperature Change in the Atmosphere Due to Condensation Heating Both diabatic and adiabatic processes are important in producing temperature change at a point in the atmosphere. The most important diabatic heat source in a tropical disturbance is the release of latent heat of condensation of water vapour in the atmosphere which amounts to nearly 2.5 × 106 J kg–1 and this value increases with temperature. The temperature change produced by this process may be computed by using the approximate relation (Anthes, 1982) (∂T/∂t)cond = (L/cp )C,
(1.4.3)
where C is the local condensation rate (mass of water vapour condensed per unit mass of air per unit time) and L is the latent heat of condensation of water vapour. The local condensation rate in a saturated updraft, for example in a mature thunderstorm, may be computed by using the approximate relation C˜ ∼ −ω∂qs /∂p
(1.4.4)
where ∂qs /∂p represents the quantity of water released between two pressure surfaces assuming a saturated adiabatic lapse rate of temperature. Anthes (loc. cit.) finds that for a saturated adiabat through, say, 24◦ C, this term would yield about
1.5
Some Physical and Dynamical Constraints and Conservation Laws
9
5 g kg–1 of water in the 200 mb-layer between 800 and 600 mb. For an updraft velocity of 1000 mb/h, the value of C is 25 g kg–1 h–1 . Substitution of this value of C in (1.4.3) yields a condensation heating rate of about 1500 K (day)–1 . Of course, such an enormous heating rate is never observed in the atmosphere. This is because the diabatic heating is almost totally compensated by adiabatic expansion and cooling, so that actual change of temperature at any level is only a small imbalance between diabatic heating and adiabatic cooling. It is only near the lower boundary where vertical motions are constrained to be small that diabatic heating produces large changes of temperature. 1.4.2.4 Identification of Heat Sources and Sinks Heat sources and sinks are created in the atmosphere by both diabatic and adiabatic processes and may be identified generally by their effects on the distribution of temperature and pressure at surface and aloft. At the lower boundary of the atmosphere, a diabatic heat source (sink) may be identified with a ‘heat low’ (‘cold high’). In the atmosphere above, a diabatic heat source is created by condensation of water vapour and a diabatic heat sink by the reverse process of evaporation of drops, or by radiative cooling. In the atmosphere, in the absence of precipitation, convection generally leads to adiabatic cooling and subsidence to adiabatic warming. The processes lead to a ‘cold low’ and ‘warm high’ respectively. This means that a ‘cold low’ is produced when warm air rises and expands adiabatically by rising to a lower pressure, and a ‘warm high’ when a sample of cold air is compressed and heated adiabatically by subsidence to a higher pressure, i.e., to lower elevations.
1.5 Some Physical and Dynamical Constraints and Conservation Laws Circulations forced by heat sources and sinks are, however, required to comply with certain physical and dynamical constraints and conservation laws. Some of these are stated below.
1.5.1 Direct and Indirect Circulations When a heat source over a place gets continually heated up, it tends to maintain itself as a source by transferring the excess heat to a heat sink in the neighborhood. Likewise, when a heat sink continually gets colder, it tends to maintain itself as a heat sink by transferring its excess cold to a neighboring heat source. The heat exchanges are assumed to take place via a direct kinetic energy-producing vertical circulation in which warm air rises and cold air sinks. But due to adiabatic processes in the atmosphere, the rising air cools and the sinking air warms up. So, the transfer
10
1 Large-Scale Tropical Circulations – Some General Aspects
process requires an independent source of energy to provide for the kinetic energy of the direct circulation. This source must be an indirect circulation which would generate sufficient available potential energy to provide for the kinetic energy of the direct circulation.
1.5.2 Energy Transformations Since the total energy of the circulations in a closed frictionless domain must remain constant, it follows that the required available potential energy (both zonal and eddy, as defined by Eqs. (1.5.1) and (1.5.2) below), must be generated in one part of a system to provide for the kinetic energy of the other part, in accordance with the following relationships (after Krishnamurti and Surgi, 1987): Generation of zonal available potential energy (Pz ) γ [ < T > ][ < Q > ]dm (1.5.1) G(Pz ) = m
Generation of eddy available potential energy (Pe ) γ [ < T Q > + < T >∗ < Q >∗ ]dm G(Pe ) =
(1.5.2)
m
Conversion from zonal available potential energy (Pz ) to zonal kinetic energy (Kz ) < Pz .Kz >= − [ < ω > ] [ < α > ] dm (1.5.3) m
Conversion from eddy available potential energy (Pe ) to eddy kinetic energy (Ke )
[ < ω α > + < ω >∗ < α >∗ ]dm
< Pe .Ke >= − m
where γ is a stability parameter, and the symbols used with meanings are: ω vertical p-velocity, where p is pressure, T temperature, Q total diabatic heating per unit mass (m) of air, α specific volume, Deviation from zonal mean, Deviation from horizontal area average, m integration over mass, m, zonal average, [] time average, and ∗ deviation from time average.
(1.5.4)
1.5
Some Physical and Dynamical Constraints and Conservation Laws
11
1.5.3 Energy Transfer Process – Carnot’s Cycle In physics, Carnot’s heat engine works in a closed system and the process of heat transfer between a source and a sink is in a cycle which is strictly reversible. In the open atmosphere, the principle of reversibility clearly does not hold, since the working substance which is a part of the environment is exposed to the environment with which it can exchange heat. Yet, the general principle of the Carnot’s heat engine has been found to be useful in interpreting the atmospheric heat transfer processes. Let us see how it works. In the first stage, the working substance which is in contact with a heat source which in the atmospheric case is latent heat released by condensation draws a certain quantity of heat from the source by rising and expanding isothermally to a somewhat lower pressure; in the second stage, it expands adiabatically by rising to a much lower pressure and temperature; in the third stage, it descends isothermally to a somewhat higher pressure and delivers a certain quantity of heat to the heat sink with which the working substance is now placed in contact, from where an adiabatic compression to a higher pressure and temperature during the fourth stage will restore the working substance to its original pressure and temperature. Thus, the working of the Carnot’s cycle in the atmosphere represents a heat transfer process in which cold air is raised adiabatically to become colder and warm air lowered adiabatically to become warmer, a process which generates available potential energy by an indirect circulation.
1.5.4 Conditional Instability and Convection The tropical atmosphere most often remains in a state of conditional instability in the sense that it is unstable (∂θ e /∂z < 0) in the lower troposphere below about 700 hPa, but stable (∂θ e /∂z > 0) above, where θ e is the equivalent potential temperature of the air which may be defined as ‘the temperature attained by a parcel of air which is raised adiabatically from its existing level to a level where all its moisture is condensed out and the latent heat of condensation added to it, and then brought down dry-adiabatically to a standard level, usually 1000 mb.’ However, it does not follow from this stability condition that it will automatically lead to convective overturning. In the tropics, moisture is usually confined to a shallow boundary layer in contact with the earth’s surface. So, convective overturning is only possible when this low level moisture is lifted to higher levels by synoptic-scale convergence. It is not surprising, therefore, that most of the tropical disturbances such as depressions and cyclones develop over oceans where low level convergence is able to lift the boundary layer moisture to higher levels for large-scale release of the latent heat of condensation which will potentially lead to explosive growth of the disturbance. Such developments may also occur over land areas where there
12
1 Large-Scale Tropical Circulations – Some General Aspects
is a copious supply of moisture from neighboring oceans and strong lapse rate of temperature may develop during afternoon-evening hours. In all cases, cyclogenesis must lift the low-level moisture to higher levels for the latent instability energy of the atmosphere to be released.
1.5.5 Cellular Structure – Shallow and Deep Convection It is well-known from laboratory experiments as well as theoretical and observational studies that when the atmosphere is thermodynamically and/or hydrodynamically unstable, it breaks down into convective cells of different dimensions depending upon the depth of the layer involved and the amount of moisture present. These convective cells may appear in the sky in the form of clouds of different horizontal and vertical extents. For example, when instability is present in a shallow layer above the earth’s surface and there is a limited supply of moisture, we may see only small-scale fair weather cumulus type of clouds. On the other hand, when instability involves a deeper layer of the atmosphere and there is deep moisture convergence at low levels with divergence above, clouds may appear in different layers and some of the cloud cells may grow to great heights. The cellular structure of the atmosphere then becomes quite clearly visible. However, when more than one layer is involved, the cells must arrange themselves vertically so as to secure sufficient vertical compensation to ensure a small pressure change that is actually observed at the earth’s surface. In a developing system, however, there is continual rearrangement of the cells which may result in a large pressure change. In the tropics, the cellular structure of atmospheric circulation is ubiquitous. However, the actual structure of the cells in any case depends upon the configuration and dimensions of heat sources and sinks and their orientation. It must also satisfy the requirements of energy conservation and transformations and physical and dynamical constraints discussed earlier in this section.
1.5.6 Coriolis Control-Variation with Latitude The relationship between pressure and wind is largely controlled by the Coriolis force, apart from friction and gravity. In higher latitudes where Coriolis control is strong, pressure rapidly adjusts to the wind field, but in the tropics where the Coriolis control is weak, the wind tends to adjust to the pressure field. So, in the absence of friction or any other force, when there is a pressure gradient across the equator where the Coriolis control vanishes, there can be a direct cross-equatorial airflow down the pressure gradient. However, away from the equator, the wind becomes increasingly quasi-geostrophic.
1.5
Some Physical and Dynamical Constraints and Conservation Laws
13
1.5.7 Conservation Laws In steady state, tropical circulation is required to satisfy certain conservation laws. These include: (a) The Law of conservation of heat energy, and (b) The Law of conservation of potential vorticity. (a) Conservation of Heat Energy: As mentioned earlier, condensation heating is one of the powerful diabatic heat sources in the atmosphere but its effect is almost totally offset by adiabatic processes. The balance between the two processes is governed by the thermodynamic energy equation in the form ∂T/∂t = −(u∂T/∂x + v∂T/∂y) + σ ω + (1/cp )δQ/dt
(1.5.5)
Where T is temperature, u, v are the components of the wind along the coordinate axes x, y respectively along the pressure surface p, ω is the vertical p-velocity, cp is the specific heat at constant pressure, δQ/dt is the rate of diabatic heating, and σ is the static stability parameter given by the relation, σ (= κT/p–∂T/∂p), where κ = R/cp , and R is Gas constant. The balance is reached when the left-hand side vanishes. In the tropics, the third term on the right hand side of (1.5.5) balances largely with the second term, while in the midlatitudes, the first term becomes important as well. (b) Conservation of Potential Vorticity: The principle of conservation of potential vorticity stated in the form (1.5.6) (ζ +f )/(δθ/δp) = Constant
(1.5.6)
(Where ζ is relative vorticity, f is Coriolis parameter, θ is potential temperature, and p is pressure) is a powerful constraint on tropical circulation though the equation was derived on the assumptions that the atmosphere was frictionless, adiabatic, and barotropic. Obviously, the assumptions are not quite realistic, since frictional forces are always at work in the earth’s boundary layer, and adiabatic processes may be more or less compensated by diabatic processes, and baroclinicity cannot be ruled out from the tropical atmosphere. Yet, the principle of conservation of potential vorticity is useful in interpreting the distribution of vorticity in airflows negotiating large mountains. For example, during northern winter, the cold NE-ly winds with negative relative vorticity over the Tibetan Plateau, on descending to the Plains of Northern India, develops a narrow zone of positive relative vorticity before resuming the normal negative relative vorticity over Central India. This change appears to be in conformity with the principle of conservation of potential vorticity. The principle appears to explain similar changes in relative vorticity in several other mountainous regions of the globe.
14
1 Large-Scale Tropical Circulations – Some General Aspects
1.6 Equatorial Circulations In the annual mean, the equatorial region of the earth receives the maximum solar radiation and may act as a perennial heat source. However, owing to distribution of continents and oceans and differential heating between them, heat sources and sinks are created along the equator. Heat sources and sinks are also created over equatorial oceans by powerful warm and cold ocean currents, especially between their western and eastern parts.
1.6.1 Circulation with Heat Sources and Sinks Placed Alternately Along the Equator – Walker Circulations A general problem of this type was first addressed by Matsuno (1966) theoretically by using a one layer homogeneous divergent barotropic model and later by Gill (1982) and others. In the first part of his treatment, Matsuno studied the different types of wave motions like Rossby and inertio-gravity oscillations that are excited in the equatorial atmosphere when he placed mass sources and sinks alternately along the equator, but in the second part he addresses the problem of forced stationary circulation and obtained the following interesting results: (i) In latitudes somewhat away from the equator, the surface tends to be raised where mass is added, and lowered where mass is extracted; (ii) In the immediate vicinity of the equator, however, the deviations of the surface elevation is less than that in the higher latitudes in magnitude. Consequently, high and low pressure cells tend to be divided into two parts, one on each side of the equator; (iii) Strong zonal flow is created along the equator when mass flows from source to sink. The zonal flow along the equator is intensified by convergence of flow from higher latitude circulations and weakened by divergence of equatorial flow to higher latitudes; (iv) In the higher latitude region, the velocity fields are in geostrophic balance. When he applied the same boundary conditions to the two-level model of the atmosphere in which heat sources and sinks are placed alternately along the equator, the differential heating produces low pressure and high pressure respectively. The wind blows geostrophically in the high latitude region. The convergence or divergence of such winds along the equator induces vertical motions. These vertical motions counteract to the imposed heat sources and sinks, and their effects are strongest along the equator. Consequently, the warm air associated with the heat source is split into two parts, by the cold air belt located at the equator. Similarly, the cold air associated with the heat sink is split into two parts by the warm air belt located at the equator. Matsuno’s results in the case of the forced stationary circulation are depicted in Fig. 1.2(a, b, c). In Fig. 1.2, mass sources and sinks are shown in terms of high (H) and low (L) pressure systems, along with the induced vertical circulations in the lower panel
1.7
Meridional Circulation with Heat Source at the Equator and Heat Sinks
15
Fig. 1.2 Forced stationary horizontal (b) and vertical (c) circulations caused by imposition of mass sources (+) and mass sinks (–) alternately along the equator (a). In (b), pressure replaces mass; H for source, L for sink, CV denotes convergence, DV divergence (adapted from Matsuno, 1966)
(c) which has come to be known as the Walker Circulation. The arrows show the directions of air motion. It is easy to see from Fig. 1.2(c) that the regions of low-level convergence and upper-level divergence are those of dense clouding and heavy precipitation (as shown by hatching below the bottom line in the lower panel) and where they diverge at low levels and converge aloft are relatively cloud-free areas with little or no precipitation. According to observations, heavy clouding and precipitation occur along the equator in some preferred regions, such as equatorial Eastern Indian Ocean, Western Pacific Ocean, the Amazon basin of South America, and some parts of equatorial Africa.
1.7 Meridional Circulation with Heat Source at the Equator and Heat Sinks in Higher Latitudes – The Hadley Circulations The classical Hadley circulation cell visualizes a zonally-symmetric annual-mean meridional circulation between a heat source with its cyclonic circulation centered at the equator and a heat sink characterized by anticyclonic circulation centered at
16
1 Large-Scale Tropical Circulations – Some General Aspects
the ridge of the subtropical high pressure with rising motion at the equator and subsidence over the subtropical belt. Tradewinds diverging from the high pressure belt were assumed to travel equatorward, converge at the equator and rise in penetrative convection producing cloud and rain and thereby releasing latent heat of condensation of water vapour carried by the tradewinds. At the equator, the rising currents are assumed to diverge in the upper troposphere and flow poleward as anti-trades to give up heat and subside over the tropical belt in order to flow back again towards the equator as tradewinds in some kind of a meridional-vertical circulation. It was this mean meridional circulation which was assumed to transfer sensible and latent heat from the equator poleward. However, when applied to the real atmosphere, the idealized Hadley circulation cell model as described in the preceding para, faces several difficulties. Observation shows that the equatorial tropopause with a temperature of about –80◦ C at a height of about 16 km above sea level is the coldest place in the tropical atmosphere and as such a transfer of heat from the equator poleward in the upper troposphere against a temperature gradient is not possible, as envisaged in the classical model. The inadequacy of the classical single-cell model in this regard was first pointed out by Fletcher (1945) who during a flight across the equatorial eastern Indian Ocean during the Second World War found two well-organized zonally-oriented cloud bands, one on each side of the equator, with little or no cloud in between over the equator. To overcome the difficulty of explaining the observation with the classical Hadley circulation model, Fletcher proposed a revised model with a small-scale meridionalvertical circulation cell interposed between the equator and the Hadley cell. The equatorial cell was supposed to have subsidence at the equator and rising motion where the Hadley cell circulation would converge into the equatorial circulation at some distance away from the equator. We call this convergence zone, the Tropical Convergence Zone (TCZ). The Fletcher model was supported by Rossby (1947) and several others. But it soon became apparent that the Fletcher model which visualized a single equatorial vertical circulation cell may also have problems in explaining some observed phenomena in the atmosphere. One of these concerns the formation of different types of clouds and release of condensation heating in different layers of the atmosphere. In small-scale cumulus-type clouds, maximum heating is likely to be small and confined mostly to a lower level. In large cumulonimbus-type of clouds, maximum heat is released by condensation in the upper troposphere. In the vertical distribution of diabatic heating in the mean tropical atmosphere, the level of maximum heating appears at about 400 mb (Holton, 1979). This fact appears to suggest that more than one layer of convective clouds may be involved in releasing latent heat of condensation in the atmosphere and that the upper layer may, in fact, contribute a greater amount of diabatic heat to the warming of the equatorial troposphere than the lower layer. In the present text, we, therefore, suggest a further revision to the Classical Hadley cell model, by dividing the single equatorial cell into two cells, a Lower Equatorial (LE) and an Upper Equatorial (UE), as shown in Fig. 1.3.
1.8
Seasonal Migration of the Equatorial Heat Source
17
Fig. 1.3 Schematic showing the suggested revised model of the Hadley circulation with two equatorial cells (UE, Upper-Equatorial) and (LE, Lower Equatorial) and the location of the TCZ; Symbols DV means Divergence, CV Convergence. Arrow shows the direction of air motion
1.8 Seasonal Migration of the Equatorial Heat Source 1.8.1 Origin of Monsoon A northward movement of the equatorial heat source from the equator is accompanied by change in the structure of the associated Hadley circulations described in the preceding section. The SE’ly tradewinds, which were earlier confined to the southern hemisphere, cross the equator directly and turn into SW/W’ly tradewinds. So the movement of the equatorial heat source brings about a latitudinal expansion of the belt of the W’ly tradewinds to the south of its trough with influx of cool, moist airmass from across the equator. On converging into the circulation around the heat source on its equatorward side at low levels, the convergence produces the well-known Intertropical Convergence Zone (ITCZ) where the converging winds rise in penetrative convection, precipitate, and diverge in the upper troposphere. A branch of the diverging currents returns equatorward as NE-ly antitrades to sink over the region of the heat sink where on sinking it joins the low-level diverging current which converges into the ITCZ in order to complete a vertical circulation which has come to be known as the Monsoon circulation, while the other branch of the diverging upper currents moving poleward sinks over higher latitudes to form the upper branch of the Hadley circulation. However, within the framework of the Hadley circulation of the northern hemisphere, there is a secondary vertical circulation associated with the TCZ where the converging low-level E/NE tradewinds rise in subdued convection, often producing clouding and light rain. The structure and properties of the monsoon circulation appear in clear perspective when we consider a general case of horizontal circulation around the equatorial heat source, interacting with the tradewind circulations at low levels, as shown schematically in Fig. 1.4, for summer in: (a) Northern Hemisphere, and (b) Southern Hemisphere.
18
1 Large-Scale Tropical Circulations – Some General Aspects
Fig. 1.4 Schematic showing the structure of the horizontal circulation around the equatorial heat source (EQ.TR.) and the locations of the ITCZ and the TCZ where the tradewinds (thick bold lines with arrows) of the two hemispheres converge: (a) Northern Hemisphere (N.H.), (b) Southern hemisphere (S.H.)
1.8.2 The Wave Structure The circulations depicted in Fig. 1.4 reveal the following general features: 1. The equatorial heat source as shown has two troughs of low pressure, one zonally-oriented (marked EQ.TR.) and the other meridionally-oriented (unmarked); they separate out four heat sinks, two on each side; 2. The winds diverging from each heat sink appear to converge into the circulation around the heat source, forming two distinct convergence zones with a zone of divergence in between on each side of the EQ.TR: these are the tradewinds; (a) the cold NE trades on the poleward side representing the Hadley circulation, and the cold SW trades on the equatorward side representing the Monsoon circulation in the Northern Hemisphere (N.H.); (b) the corresponding flows in the Southern Hemisphere (S.H.) are the cold NW Monsoon current on the equatorward side, and the cold SE Monsoon current on the poleward side; 3. The tradewind circulation on either side of the EQ.TR appears to have the structure of a wave characterized by two convergence zones separated by a zone of divergence; a wave associated with the TCZ on the poleward side, and a wave associated with the ITCZ on the equatorward side of the EQ.TR;
1.8
Seasonal Migration of the Equatorial Heat Source
19
4. Over most parts of the globe, the average zonal wavelength of the monsoon wave appears to be about 2000–2500 km; 5. In the waves, it is the convergence zones where penetrative convection and precipitation occur, with relatively clear, or less cloudy, conditions in the divergence zone in between; 6. The Monsoon and the Hadley circulations appear to co-exist with the equatorial heat source circulation at all times and in all monsoon climes. They may interact with traveling waves in forming tropical disturbances.
1.8.3 Forcing for the Seasonal Movement of the Equatorial Heat Source It is well-known that in the summer hemisphere, given the same solar radiation, temperature rises and pressure falls over both land and ocean, but the changes occur much more rapidly and with greater amplitude over land than over neighboring ocean due to much lower heat capacity of the land compared to that of the ocean. Conversely, for the same reason, in the winter hemisphere, temperature drops and pressure rises much faster and with greater amplitude over the land than over the neighboring ocean. An example of this difference in pressure tendency between land and ocean is presented in Fig. 1.5, which shows the latitudinal distribution of mean sea level pressure along longitude 70◦ E over the Arabian Sea and adjoining Western India during January (DJF) and July (JJA).
Fig. 1.5 Latitudinal distribution of mean sea level pressure along 70◦ E over the Arabian Sea sector during January (DJF) and July (JJA)
20
1 Large-Scale Tropical Circulations – Some General Aspects
So, it is the gradient of pressure tendency (also called isallobaric gradient) between the land and the ocean that appears to force the equatorial trough of low pressure over the ocean to move towards the heat source over the neighboring land in the summer hemisphere with a velocity given approximately by the well-known kinematic relation (Petterssen, 1956) C = −∇(∂p/∂t)/∇ 2 p
(1.8.1)
where C is the velocity vector, p is pressure, t is time, ∇ is Vector Del operator, (∂p/∂t) is pressure tendency, the denominator is the curvature of the pressure field in the trough, and the isallobaric gradient is in a direction at right angle to the axis of the trough of the equatorial heat source. In a study of the onset, advance and withdrawal of summer monsoon over the Indian Subcontinent, Saha and Saha (1980) tested the validity of this hypothesis qualitatively by applying it to the case of advance and withdrawal of summer monsoon along a narrow longitudinal belt parallel to the West Coast of India (about 73◦ E). Since pressure and height are related hydrostatically, the tendencies of mean monthly height of the 850 mb surface at eight stations along the meridian were calculated using the mean height data available from Ramage and Raman (1972). Results are shown in Fig. 1.6.
Fig. 1.6 Values of 850 mb mean monthly height tendency (gpm/month) in different months at Minicoy (MNC), Bangalore (BNG), Bombay (BOM), Ahmedabad (AHM), Jodhpur (JDP), Delhi (DLH), Lahore (LHR), and Peshawar (PWR). Height data are taken from Ramage and Raman (1972). Arrows show the direction of the height tendency gradient during periods of advance and withdrawal of monsoon
1.8
Seasonal Migration of the Equatorial Heat Source
21
In regions where monsoon is interhemispheric, the forcing for cross-equatorial flow appears to be provided by the isallobaric gradient between the heat source on one side and the heat sink on the other. The crossing appears to open a floodgate for cold, humid airmass of the winter hemisphere to rush into the summer hemisphere to start the process of advance of summer monsoon in that hemisphere. A reversal in the direction of the isallobaric gradient is what causes the retreat of the monsoon. The isallobaric gradient is clearly northward (indicated by a long arrow in Fig. 1.6) during the period April–June, and southward during the period September–November.
1.8.4 Intraseasonal Oscillation of Monsoon The advance and retreat of monsoon as a result of the movement of the equatorial heat source and its associated heat sinks would produce an intraseasonal oscillation in usual meteorological fields of the region. Figure 1.7 illustrates the rationale behind occurrence of intraseasonal oscillation on this account in the field of atmospheric pressure. This intraseasonal oscillation appears to result from the superposition of the monsoon perturbation upon the seasonal oscillation of pressure over the region. An example of intraseasonal oscillation in mean monthly maximum and minimum temperatures at Delhi (India) is shown in Fig. 1.8.
Fig. 1.7 Illustrating the formation of intraseasonal oscillation in atmospheric pressure (P´ denotes perturbation pressure) during advance and retreat of monsoon: Symbols used: H – High, L – Low, C – Cold, W – Warm. Other symbols have their usual meanings
22
1 Large-Scale Tropical Circulations – Some General Aspects
Fig. 1.8 Distribution of mean monthly maximum (continuous line) and minimum (dashed line) temperatures at Delhi
Intraseasonal oscillation is observed in other meteorological parameters as well; for example, in temperature, humidity, rainfall, etc. It is well-known that people living in monsoon regions experience prolonged rainy spells twice during the year, once during advance and a second time during retreat of monsoon. Also, as for temperature, it is not always cool and humid during the monsoon season. During certain periods within the season, temperature rises and weather becomes unbearably hot and humid, at least twice during the season, once during advance and a second time during retreat. For example, at New Delhi, with the arrival of monsoon towards the end of June, the daytime air temperature may rapidly drop from about 46 to 40◦ C or even lower, but it does not remain at that reduced value for long, since it rises again at least twice, the first in August when the monsoon wave moves up to the Western Himalayas, and a second time in October during the retreat of the trough. An intraseasonal oscillation of this kind occurs in other airmass properties as well, such as sunshine hours, air quality, physical comfort level, etc. Similar intraseasonal oscillations are believed to be occurring in monsoons in other regions as well.
1.9 Co-existence of Monsoon, Hadley and Walker Circulations – Inclined Troughs Over most parts of the tropics, the equatorial troughs of low pressure are seldom zonally or meridionally oriented. They are inclined to latitudes or longitudes. There may be several reasons for the observed inclination, but the main reason appears to be the general inclination of the coastline. In summer, when a powerful heat low develops over the land, its sphere of influence extends over the neighboring ocean and tends to draw an oceanic trough of low pressure which may come within its sphere of influence. So, the oceanic trough gets inclined to the heated land. In winter, when a cold high pressure develops over the land, the oceanic trough will move away from the cold land. Thus, in general, the trough will be inclined towards the nearest land during summer and away from it during winter.
1.9
Co-existence of Monsoon, Hadley and Walker Circulations – Inclined Troughs
23
Inclined troughs are found over all the global oceans, on both sides of the equator, during both summer and winter. The areas which are particularly noted for occurrence of inclined troughs are the following: 1. 2. 3. 4.
The western and the Eastern parts of the Pacific and the Atlantic Oceans; Eastern Arabian Sea and the Bay of Bengal; A wide area of the southwestern Indian Ocean close to the coast of Madagascar; A wide area of the southeastern Indian Ocean close to the Indonesian Islands and the coast of northwestern Australia; 5. A long stretch of the southwestern Pacific Ocean, extending southeastward from New Guinea across the Coral Sea.
Fig. 1.9 Streamlines showing the circulations around an NW–SE oriented inclined trough during northern summer: (a) a plan view; (b) vertical circulation in a zonal section through the center of the trough
24
1 Large-Scale Tropical Circulations – Some General Aspects
There is re-organization of the associated circulation cells when a traveling wave interacts with a quasi-stationary inclined equatorial trough. The TCZ is then activated by additional convergence at low levels and divergence at high levels which favor development. Figure 1.9 shows schematically (a) a plan view of a NW–SE oriented inclined trough, and (b) the likely zonal-vertical circulation across the trough in the N.H. An inclined trough is more than of academic interest, for it appears to mark out the regions noted for development of tropical cyclones. But, what is the connection between the trough and the cyclone and how is development favored by an inclined trough? This aspect of the question will be looked into further in the next chapter after we have introduced the easterly and the westerly waves which initiate the development.
1.10 Definition of Monsoon In the past, the word ‘monsoon’ has been defined exclusively in terms of either seasonal rainfall (Rao, 1976), or reversal of the direction of the prevailing surface wind between summer and winter (Ramage, 1971). For people wholly dependent upon rainfall for water resources, food production, etc., especially in developing countries of Southeast Asia, Africa and South America, a year with subnormal rainfall means to them poor monsoon and that with above-normal rainfall good monsoon. To these people, the word ‘monsoon’ is almost synonymous with rainfall. The Arabs first noted the seasonal reversal of the surface wind direction while sailing over the Arabian Sea more than a thousand years ago and coined the word ‘Mawsim’ for the phenomenon. This definition has also continued to this day. Ramage (1971) defines monsoon by the following criteria: (1) The prevailing wind direction shifts by at least 120◦ between January and July; (2) The average frequency of prevailing wind directions in January and July exceeds 40%; (3) The mean resultant winds in at least one of the months exceed 3 m s–1 ; and (4) Less than one cyclone-anticyclone alternation occurs every two years in either month in a 5◦ latitude-longitude rectangle. Definitions on similar lines, seeking to identify monsoon by either rainfall, or the seasonal reversal of the prevailing surface wind, have been advanced by several workers. None of the other characteristic features of the monsoon are ever mentioned in these definitions. It was Edmund Halley (1686) of England who appears to have been the first to conceive monsoon as a seasonally-reversing tradewind circulation when he wrote in his celebrated paper in the Philosophical Transactions of the Royal Society of London, the following:
1.11
Global and Regional Distribution of Monsoons
25
But as the cool and dense air, by reason of its greater gravity, presses on the hot and rarefied, it is demonstrable, that this latter must ascend in a continued stream, as fast as it rarefies, and that being ascended, it must disperse itself to preserve the equilibrium; that is, by a contrary current, the upper air must move from those parts where the heat is greatest: so by a kind of circulation, the north-east tradewind below will be attended with a south-westerly above, and the south-easterly with a north-west wind above. And that this is more than a mere conjecture, the almost instantaneous change of the wind to the opposite point, which is frequently found in passing the limits of the trade winds, seems to assure us; but that which above all confirms this hypothesis, is the phenomenon of monsoons, by this means most easily solved, and without it hardly explicable.
Our analysis of monsoon as a divergent tradewind circulation which converges into the seasonally-migrating equatorial heat source, producing a wave on the equatorward side of its trough prompts us to offer the following holistic, comprehensive, and yet simple, alternative definition of tropical monsoon: “Monsoon is a large-scale perturbation of the tradewind circulation associated with the seasonal movement of the equatorial heat source, which converges into the circulation around the heat source on its equatorward side at low levels, producing a wave with two zones of precipitation and a zone of clearance in between, and a host of other characteristic changes in airmass properties, during its advance and retreat.”
The foregoing definition brings within its fold not only the observed rainfall and reversal of wind direction, but also a host of other characteristic changes in prevailing meteorological conditions, such as changes in air temperature, cloudiness, etc. It also produces an intraseasonal oscillation in meteorological parameters during advance and retreat because of its wave structure. It is believed that the definition given here can be used as a dependable criterion for identifying and tracking monsoon over any part of the globe at any time of the year. Our definition implies co-existence of monsoon with Hadley and, at some locations, Walker-type east-west circulations. However, in offering the above definition of monsoon, we should not lose sight of some other regions of the globe, especially the extratropics, where such thermal contrast may arise between a large continent and a neighboring ocean and force a seasonal movement of the oceanic trough of low pressure between land and ocean. Of course, in such regions, the prevailing winds and eastward-propagating baroclinic wave disturbances may often interfere with the weak monsoon circulation that may develop and make its identification difficult.
1.11 Global and Regional Distribution of Monsoons 1.11.1 Tropical Monsoons Since tropical monsoon constitutes a perturbation of the tradewind circulation that converges into the ITCZ, the seasonal movement of the ITCZ offers a practical means to identify the leading edge of the monsoon over a region at any time of the year (e.g., Riehl, 1954; Saha, 1978; Walisser and Gautier, 1993; Saha et al., 1998).
26
1 Large-Scale Tropical Circulations – Some General Aspects
In other words, the region swept out by the ITCZ in the course of its movement between summer and winter constitutes the domain of the tropical monsoon in that region. A definition of a monsoon region on similar lines based on the seasonal movement of the ITCZ was also advanced by Asnani (1993). In general, tropical monsoons are interhemispheric in character and usually follow the seasonal movement of heat lows over the land from one hemisphere to the other. Three continental sectors stand out in this respect, viz., Asia, Africa, and the Americas. In the Asia sector, the movement is between Australia and Asia; in the African sector, between Northern and Southern Africa; and in the American sector between South America, and Central and North America. The monsoon regions in these sectors are as follows: Asia–Australia sector Region I Indian Sub-continent and adjoining SE Asia Region II Eastern Asia including China, Korea and Japan Region III Maritime Continent including Indonesia and Philippines Region IV Australia Africa sector Region V North and South Africa American sector Region VI South America Region VII Central America and adjoining States of Southwestern North America In general, the domains of oceanic monsoons are much narrower, and confined to latitudes closer to the equator. However, they are much wider over the western and the eastern parts of the Atlantic and the Pacific Oceans where they tend to join up with the continental monsoons along inclined troughs.
1.11.2 Extratropical Monsoons Monsoonal-type atmospheric circulations between continents and oceans are also observed in some extratropical regions of the globe where seasonal contrasts in temperatures between land and ocean and differential heating between them bring about seasonal movements of the circulation systems and associated rainbelts across the coastline and to higher latitudes. They are also observed over some extratropical mountain plateau regions where the circulations and the associated rainbelts are affected by seasonal reversals of temperature and pressure between the Plateau and the neighboring Lowlands or oceans. However, it is often difficult to see the full impact of such monsoons in the face of interference from strong planetary wind systems and frequent movement of baroclinic wave disturbances across the region. The concept of a monsoon in extratropical latitudes is not new, though it has been opposed by some meteorologists (e.g., Ramage, 1971). It was advocated by several
1.11
Global and Regional Distribution of Monsoons
27
workers (e.g., Alisov, 1954; Khromov, 1957 and others), though the steadiness of the prevailing surface wind, which was cited to be one of the main characteristics of a monsoon circulation, was found to be somewhat low on account of the interference from disturbances (Klein, 1957). It is proposed to show in Part 3 of this text that the idea of an extratropical monsoon is not just a fantasy, but a stark reality. It is driven by differential heating between a heat source and a heat sink in much the same way as in a tropical monsoon, but in a different environment. A tropical monsoon results from the seasonal movement of an oceanic heat source called the equatorial heat source from one hemisphere to the other and its movement is limited to the tropical belt. For this monsoon, the subtropical high pressure cells with their anticyclonic circulations and diverging tradewinds serve as heat sinks. On the other hand, for an extratropical monsoon, the same subtropical high pressure cells would act as heat sinks but heat sources would lie over the extratropical belt. The exact mechanism of this process will be elaborated in Chaps. 5 and 11 with a few case studies. Figure 1.10 shows the approximate domains of tropical and extratropical monsoons over different parts of the globe including the oceans.
Fig. 1.10 Global distribution of principal monsoon regions (stippled areas) bounded by the January and July locations of the ITCZ: January (full line); July (dashed line). Regions marked I to VII are areas of tropical continental monsoons; VIII and IX are domains of Extratropical monsoon; SPCZ denotes Southwest Pacific Convergence Zone, and SACZ the Southwest Atlantic Convergence Zone
1.11.3 Zonal and Meridional Anomalies As shown in Fig. 1.10, large zonal and meridional anomalies exist in the distribution of monsoons over the globe. These anomalies arise from the following:
28
1 Large-Scale Tropical Circulations – Some General Aspects
(a) Geographical distribution of continents and oceans; (b) Influence of ocean currents – oceanic monsoons; and (c) Effects of large-scale orographic barriers. (a) Geographical Distribution of Continents and Oceans: Whatever may be the geological origin of continents, their locations in a certain alignment on the earth’s surface, appears to be very important for the seasonal movement of monsoons. This happens in three major continental sectors, viz., Asia-Australia, North and South Africa, and North and South America. A quick glance at the world map shows that in each sector, the landmasses are oriented more or less in a general NW–SE direction, and monsoons in general follow this direction in their interhemispheric movement following the seasonal movement of the sun. (b) Influence of Ocean Currents – Oceanic Monsoons: Over all the world oceans, the subtropical anticyclonic gyres and the coastal ocean currents exercise large control on the distribution of ocean surface temperature. Their seasonal shift gives rise to oceanic monsoons and in this process ocean currents play a very major role. For example, in the Pacific Ocean, the seasonal location of the ITCZ is affected by the following major ocean currents: (1) The poleward-flowing warm Kuroshio Current in the northwest; (2) the equatorward-flowing cold California Current in the northeast; and (3) the equatorward-flowing cold Peruvian or Humbolt Current in the southeast. On account of the cross-equatorial flow of the Peruvian Current, the equatorial eastern Pacific Ocean remains cold almost throughout the year. Similarly, in the Atlantic Ocean, the corresponding ocean currents are: (1) the poleward-flowing warm Gulfstream in the northwest; (2) the equatorward-flowing cold Canaries current in the northeast; and (3) the equatorward-flowing cold Benguela current in the southeast. The equatorward-flowing cold Benguela Current keeps the temperature of the equatorial eastern Atlantic Ocean low almost throughout the year. On account of these cross-equatorial flows of cold ocean currents, the ITCZ seldom appears south of the equator over these oceans. The situation over the Indian Ocean is somewhat different. Here, in the western Indian Ocean, because of the dominant influence of the cold Somali current during northern summer, a zonal anomaly of surface temperature is maintained between the western and the eastern parts of the equatorial ocean almost throughout the year. In the Southern Indian Ocean, the equator-flowing cold West Australian current keeps the surface temperature of the eastern part relatively cold, while in the western part the poleward-flowing warm Agulhas current maintains a warm ocean surface. These movements of the ocean currents are reflected in the equatorial distribution of ocean surface temperature shown in Fig. 1.11. In some of the oceans where equatorial belt remains cold, the ITCZ is displaced from the equator to the relatively warmer parts of the ocean or to the nearest warm continent. For example, the SW Pacific Convergence Zone which runs from equatorial New Guinea eastsoutheastward across the Coral Sea area may be looked upon as the displaced ITCZ of the southern Pacific Ocean. Similarly, in the western part of the equatorial Indian Ocean which remains relatively cold throughout the year, the ITCZ is displaced to about 15◦ S during northern winter and to about 25◦ N during
1.12
Co-existence of Monsoon with Desert Circulation
29
Fig. 1.11 Equatorial distribution of mean ocean surface temperature during February (continuous line), and August (dashed line) (after Defant, 1961)
northern summer. Similar displacements may also be observed over other parts of the tropical oceans and continents. (c) Effects of Mountain Barriers: Large mountain systems, such as the Himalayas of Asia, the Mountains of East Africa, the Rockies of North America, the Andes of South America, and the Great Dividing Range of Australia, all appear to affect the circulations in their respective areas by relocating the ITCZ and other troughs and ridges of pressure through their mechanical and thermal effects on circulation. In accordance with the principle of conservation of potential vorticity, the blocking of a W’ly flow without horizontal shear approaching a north-south oriented mountain range will produce a trough of low pressure in the run-up to the base of the mountain, then a high pressure as it climbs up to the top, and then a low pressure as it descends on the leeside, till it resumes its zonal flow. In the case of an E’ly flow approaching such a mountain range, there is a difference. The flow senses the presence of the mountain barrier from a long distance and accordingly adjusts the pressure in such a way that will enable it to get over the top and emerge as an E’ly current on the leeside. This it does by first producing a low pressure trough and then a high pressure ridge of such intensity as will enable it to cross the mountain and emerge on the other side as an E’ly current. The ITCZ or the TCZ, as the case may be, is relocated whenever the airflow around the equatorial trough has to negotiate these high mountain barriers.
1.12 Co-existence of Monsoon with Desert Circulation It is by no means an accident that the world’s principal monsoons co-exist with subtropical deserts on their poleward sides. In the northern hemisphere, the great central Asian desert, the great Saharan desert in Africa, the Mojave-Sonoan desert
30
1 Large-Scale Tropical Circulations – Some General Aspects
in Central America and adjoining southwestern North America, all lie on the poleward sides of monsoons in the respective regions. Similar co-existence of monsoons with deserts on their poleward sides is also observed in the southern hemisphere in all the continents; for example, the Great Gibson and Western Deserts in Australia, the Namib-Kalahari desert in Southern Africa and the Patagonia desert in South America all lie on the poleward sides of their respective monsoons. So, a coexistence of monsoons and desert circulations appears to be a global phenomenon which has drawn the attention of meteorologists for several decades. Ramage (1966) observed a seesaw type inverse relationship between surface pressure anomaly in the ‘heat low’ over the Thar Desert in Pakistan and monsoon rainfall anomaly over the neighboring Arabian Sea off Bombay. He observed that a fall (rise) of surface pressure in Thar Desert over Pakistan was correlated with increase (decrease) of rainfall over the neighboring Arabian Sea. In a theoretical study of the dynamics of deserts and recurrent drought in the Sahel at the southern periphery of the Saharan desert, Charney (1975) observed that to maintain the radiative equilibrium of the atmosphere over a surface with high albedo, such as a sandy soil surface, against radiative heat loss, air must descend and that it is this descent or adiabatic subsidence and warming of the air that leads to continued dryness and maintenance of the desert. He also recognized the contribution of the descending branch of the Hadley circulation to the desertification process but expressed the view that the impact of the radiative heat loss was greater than subsidence warming.
Fig. 1.12 Mean vertical circulations (resultant streamlines) along the Greenwich meridian at 12 GMT during August, showing the monsoon and desert circulations STF marks the location of the subtropical front between the desert and the Mediterranean Sea circulations (after Saha and Saha, 2001b)
1.12
Co-existence of Monsoon with Desert Circulation
31
Blake et al. (1983) who carried out a detailed observational study of heat balance of the atmosphere over a part of the Saudi Arabian desert (Rub-al-Khali) in May 1979 found that there was strong subsidence in the middle and upper troposphere which often descended to very low levels at night. Strong convection over the desert surface during daytime was limited to the lower troposphere only and there was strong outflux of sensible heat in the lower and middle troposphere from the desert to the monsoon over the adjacent Arabian Sea. Saha and Saha (2001b), using NCEP reanalysis data, computed the mean vertical circulation along the Greenwich meridian over the western part of the Saharan desert at 12 GMT during August and showed how the monsoon and the Hadley circulations co-exist with the desert circulations and are linked to one another. Their results are presented in Fig. 1.12. Figure 1.12 shows that in the meridional circulation, while the monsoon winds converge into the ITCZ along about 12◦ N at low levels producing strong penetrative convection which diverges in the upper troposphere both equatorward as well as poleward, there is strong sinking motion in the Hadley circulation over the desert. In fact, the sinking motion over the desert area is so strong that it prevents the thermally-direct low-level desert convection from rising beyond 700 mb. Along the northern boundary of the desert heat low which extends beyond latitude 30◦ N, cool, moist winds diverging from a high pressure area over the neighboring Mediterranean Sea converge into the heat low circulation, forming a subtropical front (STF), the presence of which was first reported by Soliman (1958). In the present text, we have designated the STF as the TCZ
Chapter 2
Tropical Disturbances (Quasi-stationary Waves, Easterly/Westerly Waves, Lows and Depressions, Cyclonic Storms, and Meso-Scale Disturbances)
2.1 Introduction It is widely believed that weather in the tropics is largely seasonal and monotonously of one type; hot and humid during summer and cold and dry during winter. In reality, it is not always quite so, for, the monotony is often broken by different types of tropical disturbances ranging from synoptic-scale wave disturbances that form in the tradewinds that converge into the circulations around the equatorial trough of low pressure, and by severe local storms originating in meso-scale and small-scale disturbances. The synoptic-scale wave disturbances which have a mean horizontal scale of about 1000–3000 km and a Lagrangian time scale of 3–7 days usually move westward with a speed of about 5–8 m s–1 . Some of these disturbances, under favorable conditions, develop into depressions and cyclones. They are mostly thermally driven and powered by the latent heat of condensation of water vapour carried by the tradewinds. For this reason, the most intense of these disturbances form and develop over the warmer parts of the oceans where there is almost unlimited supply of heat and moisture from the underlying surface for their development. They are called cyclones in the Indian Ocean, hurricanes in the Atlantic and typhoons in the Pacific Oceans, though they are fundamentally the same. Out of the large number of lows and depressions that form in the tropics, a limited number only develop into cyclones or severe cyclones and, out of these, only a small percentage develop an ‘eye’ at the center. The severe cyclones are associated with extremely high winds and torrential rain. Some of them cause storm surges which inundate many coastal belts and cause heavy loss of life and property. Many of them breed deadly tornadoes which knock down many standing structures and cause heavy loss of life and property. Tornadoes have drawn the attention of meteorologists from early times and we have an extensive literature on the subject, dealing with their formation, structure, development and movement. Table 2.1 gives a list of the synoptic-scale disturbances classified on the basis of their scale and tangential wind speed, adopted by most meteorological services. Most of the above-mentioned disturbances originate in quasi-stationary waves when they interact with traveling disturbances. Embedded within the abovementioned large- or synoptic-scale disturbances, there are several types of mesoor small-scale disturbances of shorter duration which are very violent in nature. Just K. Saha, Tropical Circulation Systems and Monsoons, C Springer-Verlag Berlin Heidelberg 2010 DOI 10.1007/978-3-642-03373-5_2,
33
34
2 Tropical Disturbances Table 2.1 Classification of synoptic-scale tropical disturbances Classification
Horizontal scale (km)
Range of wind speed (m s–1 )
Low/trough Depression Deep depression Cyclonic storm/tropical storm Tropical cyclone (severe cyclonic storm/hurricane/typhoon)
2000–2500 1500–2500 1500–2000 1000–1500 980 965–979 945–964 920–944 65
Fig. 3.1 NOAA satellite view of hurricane ‘ANDREW’ over the Gulf of Mexico, 23 August 1992
view of the system. In fact, our current knowledge of the three-dimensional structure of meteorological conditions inside tropical cyclones has been derived mainly from radiosonde, radar and aircraft data. Since 1936, an expanding network of radiosonde observatories has been operating in different parts of the globe. They routinely record pressure, temperature and humidity at different levels of the atmosphere. Similarly, several countries exposed to the dangers of these storms have set up radar stations to observe and measure various storm parameters. But, perhaps, the most comprehensive and reliable sets of data regarding the three-dimensional structure of tropical cyclones have come from aircraft probes of these storms. The meteorological community owes a great deal of gratitude to the valiant pilots of the hurricane research aircraft of the United States of America
3.2
Observed Structure of a Tropical Cyclone
63
Weather Bureau (now National Weather Service) who, since the forties of the twentieth century, at great risk to their lives, penetrated deep into hundreds of these dangerous storms including the eye and collected data at various flight levels and at various distances from the center for each individual cyclone. Detailed reports of these flights are available with the National Hurricane Research Laboratory and the National Hurricane Center at Miami (Florida) and constitute, perhaps, our best source of data for determining the structure of cyclones. The valuable reports have been used in two ways. A few workers constructed the structure of individual cyclones. For example, Hawkins and Imbembo (1976) reported the wind and temperature structure of hurricane ‘INEZ’ that occurred in September 1966. Sheets (1980) reported the structure of wind and pressure distributions inside hurricane ‘ANITA’ in September 1977. Several workers (e.g., Hughes, 1952; Gray and Shea, 1973; Frank, 1977; Gray, 1979) used a compounding technique in which data from several cyclones were stratified according to the size, intensity, season and quadrant of each cyclone and many other parameters. The compounding technique reduced random errors, and, by averaging, eliminated important asymmetries between individual cyclones. The analysis of such averaged data provides, perhaps, a most complete three-dimensional composite picture of the large-scale structure of a tropical cyclone. A brief description of the radial and vertical distributions of wind, pressure, temperature and several other parameters in some individual as well as composite cyclones is provided in the pages that follow. The information furnished is representative of conditions in respect of hurricanes in North Atlantic Ocean and typhoons in Western North Pacific Ocean.
3.2.1 Wind Structure Hawkins and Imbembo (1976) present the horizontal and vertical distributions of several important wind parameters in respect of hurricane ‘INEZ’. Their streamline and isotach analyses of low-level winds at about 950 hPa in this hurricane are presented in Figs. 3.2 and 3.3 respectively. Figure 3.3 shows the wind speeds (isotachs) at different locations around the eye of the hurricane INEZ. The analyses of the wind field presented in Figs. 3.2 and 3.3 reveal, inter alia, the following: (1) Considerable degree of asymmetry in the direction and speed of the flow outside the eye-wall; (2) The spiralling winds starting from great distances appear to be converging towards the center and meet in concentric circles along the eye-wall; (3) At the center of the eye, there appears to be little or no wind.
64
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.2 Analyses of low-level winds (streamlines) at 950 hPa in hurricane ‘INEZ’ (Hawkins and Imbembo, 1976)
3.2.2 Radial and Tangential Components of the Wind 3.2.2.1 Radial Wind A vertical cross-section of the radial component of the wind (m s–1 ) in western Atlantic composite hurricane (Gray, 1979) is presented in Fig. 3.4. The cross-section shows a layer of inflow of environmental air at low levels and a layer of outflow aloft, the two flows being separated by a surface which slopes outward with height from about 800 mb at the boundary of the eye wall to about 300 mb at the outer boundary of the hurricane at a distance of about 1000 km from the center. The strongest inflow occurs in the layer between about 950 and 850 mb, while the strongest outflow occurs between about 300 and 100 hPa. A layer of inflow appears to occur above the tropopause level.
3.2
Observed Structure of a Tropical Cyclone
65
Fig. 3.3 Analyses of the wind field (isotachs) around the eye of the hurricane ‘INEZ’ (Hawkins and Imbembo, 1976) The axisymmetric eyewall is shaded
3.2.2.2 Tangential Wind Figure 3.5 shows the zonal-vertical distribution of tangential winds (m s–1 ) at different distances from the center of the eye in a Pacific composite typhoon, after Frank (1977). Figure 3.5 shows cyclonic flow in the lower troposphere and anticyclonic flow in the upper troposphere. Similar features of the distribution of tangential winds are also revealed by a vertical cross-section for the tangential wind speed (m s–1 ) in hurricane INEZ studied by Hawkins and Imbembo (1976). Their report shows that at the radial boundaries of the eye wall the tangential wind attains hurricane speeds (>100 Kts). It also shows that the maximum speed inside the wall is higher in the rear of the storm than in front by about 10 knots in a westward-moving system.
66
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.4 Vertical cross-section of radial winds (m s–1 ) in the western Atlantic composite hurricane (Gray, 1979) Negative sign means inward flow
The distribution shows that starting from a distance of about 40 miles where the speed is about 50 knots the speed continually increases till it reaches the highest value of about 120 knots in the eye wall and then it drops rapidly to very low values inside the eye of the storm. Within the eye-wall, a weak vertical wind shear is maintained between about 900 and 400 mb by intense cumulus convection which transports cyclonic momentum upward. Outside the eye wall, the speed of the tangential wind decreases with distance from the center. Attempts have been made to estimate the tangential and radial components of the wind in a hurricane by making use of the law of conservation of angular momentum, given by the relations vλ r + fr2 /2 = Const
(3.2.1)
∂vr /∂r + vr /r = 0
(3.2.2)
where vλ , vr are the tangential and the radial components of the wind respectively, r is the radial distance from the center and f is the Coriolis parameter (taken constant in space). Since (3.2.1) would lead to infinite speeds at the center, Deppermann (1947) proposed a so-called Rankine vortex with velocity distribution, vλ /r = Constant, for the central area. The radial distribution of the tangential wind speed, vλ (r), outside the radius of the maximum wind, R0 , is often represented by the empirical relation
3.2
Observed Structure of a Tropical Cyclone
67
Fig. 3.5 Vertical distributions of tangential winds (m s–1 ) at different distances from the center in a Pacific composite typhoon (Frank, 1977). Positive denotes cyclonic
vλ (r) = vλ (R0 )(R0 /r)x ,
R0 < r < r0
(3.2.3)
where r0 is at the outer edge of the disturbed area of the hurricane at a radial distance of about 1000 km or so. The exponent x in (3.2.3) varies from 0.5 to 0.7 near the center of different storms (Miller, 1967). Sheets (1980) computed values of vλ (r) in hurricane ‘ANITA’ using (3.2.3) for values of x = 0.5 and 0.6.
3.2.3 Vertical Motion in a Mean Typhoon Frank (1977) computed vertical motion (mb day–1 ) in a mean Pacific typhoon, which is shown in Fig. 3.6. The radial-vertical section in Fig. 3.6 shows penetrative convection with a maximum of 400 mb day–1 (though it may be much stronger in an individual typhoon) in the eye-wall involving the whole troposphere, which appears to extend to about 4◦ latitude circle from the center. Beyond that distance, while light upward motion continues in the upper troposphere (above about 400 mb) up to a distance of at least 1000 km and replaced by downward motion farther away, alternate zones of subsidence and convection are found in the lower troposphere up to a distance of about 1000 km from the center. This brings us to an important question: Why is
68
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.6 A cross-section showing vertical motion (mb day–1 ) in mean typhoon (Frank, 1977)
this difference in the structure of vertical motion between the lower and the upper tropospheres? Perhaps, the answer lies with the stability properties of the two layers and the intensity of moist convection at low levels. The upper troposphere is statically stable, so a weak low-level convection is unable to pierce through the stability barrier above and its effect remains confined to the lower troposphere only with alternate zones of upward and downward motion as in a gravity wave motion. On the other hand, strong convergence in the eye-wall leads to penetrative convection involving both the lower and the upper tropospheres.
3.2.4 Pressure Distribution Barograph traces at shore stations lying in the path of a hurricane often give a measure of pressure fall in the core of a hurricane if the rate of movement of the storm center is known. Riehl (1954) quoted the barograph trace marking the passage of a hurricane in New Orleans, Lousiana, USA in September 1947, which is shown in Fig. 3.7.
3.2
Observed Structure of a Tropical Cyclone
69
Fig. 3.7 Barograph record marking the passage of a hurricane in New Orleans, Lousiana, USA in September 1947 (after Riehl, 1954). Pressure is in inches of mercury
It shows that the pressure fell extremely rapidly in the three hours before arrival of the eye. It rose equally rapidly after passage of the eye.
3.2.5 Temperature Distribution Hurricane ‘INEZ’ was a rather small-scale hurricane, though it was well-developed. It was found to possess a temperature structure very similar to that in many largescale hurricanes, as shown in Fig. 3.8. In Fig. 3.8 which shows the radial distribution of temperature anomalies (deviation from mean annual tropical atmosphere) in hurricane ‘INEZ’, a most striking feature of the temperature distribution is the appearance of two concentrated cores of warm air inside the eye wall, one in the upper troposphere with the maximum warm anomaly (deviation from mean annual tropical atmosphere) above the environment of about 16◦ C at 250 mb and the other in the lower troposphere with the maximum warm anomaly above the environment of about 11◦ C at 600 mb. The warm anomaly is minimal at about 500 mb between the maxima above and below. A strong temperature gradient exists across the eye wall at all levels, especially in the lower troposphere. The warm anomaly in the eye wall where heavy precipitation occurs is about half of what is inside the eye. In the environment outside the eye wall, alternate layers of cold and warm anomalies appear in the vertical. For the same radial-vertical section inside the hurricane ‘INEZ’, Hawkins and Imbembo also worked out the vertical distribution of equivalent potential temperature, θ e (not shown here), which showed that, by and large, the same type of thermal structure and vertical stability prevailed in the field of the hurricane as exhibited by the three well-known tropical soundings, presented in Fig. 3.9 (taken from Anthes, 1982, with permission of Am. Meteor. Soc.)
70
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.8 Radial-vertical distribution of temperature anomalies (◦ C) (from mean annual tropical atmosphere) through the center in a NW–SE direction in hurricane ‘Inez’ on 28 September 1966 (Hawkins and Imbembo, 1976)
Since the formation of an eye is crucial to the generation of extremely low pressures and high winds in a hurricane, it is interesting to consider how these low pressures found in the eye are created. Observations show that at the top of a hurricane, the horizontal pressure gradient almost disappears. This means that a lowering of pressure at surface is directly proportional to an increase in the temperature of the air column inside the eyewall. Malkus and Riehl (1960) showed that the warmest possible column of air rising from surface moist adiabatically with a value
3.3
The Eye and the Eye-Wall
71
Fig. 3.9 Vertical profiles of equivalent potential temperature, θ e for mean September tropical sounding, mean hurricane within 100 n miles of the center, and mean eye of storms with pressures between 939 and 949 mb
of θ e = 350 K at surface would yield a pressure of 1000 mb at surface. They showed that under moist adiabatic conditions, a change in equivalent potential temperature would be related to a change in mean sea level pressure, ps , by the empirical relation (3.2.4) ps = −2.5θe The relation (3.2.4) assumes that the whole atmospheric column has the same. Equivalent potential temperature as at the surface; an assumption which is seldom true in the real atmosphere.
3.3 The Eye and the Eye-Wall 3.3.1 General Considerations – Formation of the Hurricane Eye While it is likely that a high value of equivalent potential temperature and deep moist convection may explain a certain amount of lowering of pressure in the eye wall, the same cannot be said for the formation of the lowest pressure at the center
72
3
Tropical Cyclones/Hurricanes/Typhoons
of the hurricane eye which is found to be relatively dry and cloudless and much warmer than the eye wall. According to Gray and Shea (1973), the temperature excess over the environment in the eye is almost double of that in the eye-wall, an observation which is well supported by hurricane ‘INEZ’ data (Fig. 3.8). It is due to the temperature difference that pressure is considerably lower in the eye than in the eye-wall or the environment outside the eye-wall. Thus, the eye plays a fundamental role in the thermodynamics of the hurricane circulation by producing the warmest core with the lowest pressure at the center and the strongest hurricane winds and heaviest rainfall in the eye-wall. What then is responsible for the triggering and formation of a hurricane eye? Numerical studies (e.g., Eliassen, 1952, 1959; Kuo, 1959; Krishnamurti, 1961; Estoque, 1962; Willoughby, 1979; Smith, 1980) by computing the radial – vertical circulations in a symmetric hurricane using plausible assumptions in regard to friction, condensation heating, etc., have hinted at downward motion in the eye to explain the observed lowering of surface pressure ins ide the eye-wall. However, these studies leave many questions unanswered, especially that relating to the role of heat released by condensation in the evolution of the circulation, especially how it leads to such low pressure inside the eye. It is evident that a unified, comprehensive theory of development of a hurricane eye starting from a pre-existing feeble wave disturbance is yet to come. Our conceptual model of development of a cyclone from a pre-existing monsoon trough, presented in Sects. 2.4.2 and 2.4.3, emphasizes the role of condensation heating in producing the subsidence warming at the center by an indirect circulation which is required for supporting the energy-producing direct circulation between the eye-wall and its outer environment. The degree of development depends on the extent of the subsidence warming in the eye produced by condensation in the eyewall. The presence of such warm areas, one in the upper and the other in the lower troposphere in the eye region was well demonstrated by the temperature profiles in hurricane INEZ (Fig. 3.8). Initially, a trough of low pressure has only one sector of large-scale condensation at the location of the ITCZ and a subdued one at the TCZ. The heat generated at these convergence zones through condensation is, perhaps, just about what is required to maintain the trough in its existing form; it cannot lead to further development. It requires docking with an easterly or W’ly wave to initiate the development process by producing two active zones of penetrative convection, one on either side of the trough axis and a zone of subsidence warming and lowering of pressure in between. Thus, development proceeds by stages as the zone of condensation heating extends around the center of the low pressure trough by the tangential winds and it is only when it encircles the center and the disturbance becomes quasi-axi-symmetric that it attains maximum development and gets ready to form an eye. However, observations show that once formed, an ‘eye’ may not last long. Its continuance is subject to a delicate balance amongst the processes which create it. In some cases, with little variation in dimension and intensity, it may last for hours and days, or form a new after a collapse, especially when the system is over a warm ocean with favorable environmental conditions. In others, it may appear only momentarily and then vanish from the scene.
3.3
The Eye and the Eye-Wall
73
3.3.2 Circulation Inside the Hurricane Eye – Evidence of Meso-Scale Vortices Recent studies by Montgomery et al. (2006) and Aberson et al. (2006) and several others based on an observational analysis of a unique set of data inside hurricane ‘Isabel’ over the western part of North Atlantic, 12–14 September 2003, appear to throw light on several aspects of the low-level circulation inside its eye and its thermodynamics. The storm-relative data from multiple dropwindsondes, several in-situ visual and radar observations from aircraft flying at different levels below about 4 km, as well as high-resolution satellite imagery at different radial distances from the center and different heights inside and outside the eyewall during the period (12–14 September) were composited and the analysis was based on these composites. Figure 3.10 (the upper panel) is a satellite view of the hurricane eye at 1711 UTC on 13 September 2003, which reveals the presence of four meso-scale vortices near the inner edge of the hurricane and a pentagonal structure of the eye-wall. Isabel was a category-5 hurricane on the Saffir-Simpson scale during the period, 12–14 September 2003, with an azimuthally-averaged tangential wind speed maximum of 76 m s–1 located at an altitude of about 1 km above the ocean surface at a radial distance of about 42 km from the center. However, an extreme tangential wind speed maximum of 107 m s–1 and a vertical wind maximum of 25 m s–1 were observed by a dropwindsonde at an altitude of about 1.4 km a.s.l. at 40 km radial distance at 1752 UTC on 13 September 2003. It was also found that the tangential wind speed fluctuated greatly azimuthally at this radial distance. It is likely that these extreme wind speeds and their azimuthal fluctuations in the horizontal are related to the mesovortices that were observed at these locations. According to Montgomery et al. (loc. cit.), though the high radial and vertical resolutions of nearly 30,700 data points (approximately half of which were from dropwindsondes) instill high confidence in the retrieved axisymmetric features in the eyewall region, greater uncertainty exists about the axisymmetric features within the 30 km-radius of the center where data sampling was limited. The lower panel (b) of Fig. 3.10 shows a schematic representation of the likely low-level horizontal circulation inside the eye of hurricane Isabel with four identified mesovortices near the inner edge of its eyewall. Panel (c) shows a sketch of the likely radial-vertical secondary circulation inside the hurricane eye. Montgomery et al. (loc. cit.) used the data of the storm-relative tangential and radial winds as well as the equivalent potential temperature (θ e ), absolute angular momentum and transverse secondary circulation (vector) composited from the GPS dropwindsonde and the flight-level data between 1600 and 2300 UTC on 13 September 2003 to derive the radius-height mean structure of the circulation in the eye and the eyewall region. Some of their important findings are: (a) The tangential wind speed has a maximum at a radial distance of 40–50 km from the eye center at altitude 2–3 km a.s.l.; (b) The base of the eyewall appears lifted to a height of about 1 km, above the ocean surface, thereby allowing outside oceanic air to flow heavily into the eye
74
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.10 (a) A visible Moderate Resolution Imaging Spectroradiometer (MODIS) image (from NASA’s Aqua Satellite) of the eye of hurricane Isabel at 1711 UTC on 13 September 2003, showing the four whirls and vortices near the eye-wall and the pentangular structure of the eye-wall (after Montgomery et al., 2006 and Aberson et al., 2006). The lower panel (b) gives a schematic representation of the likely low-level horizontal circulation inside the eye with mesovortices near the eye-wall. The lower panel (c) is a schematic of the radial-vertical secondary circulation inside the eye. Shaded areas are zones of windspeed maxima. L denotes Low, V vortex
region to a radius of about 15 km. A low-level (0–250 m) radial inflow speed maximum of 20 m s–1 is located at a radial distance of 25 km from the center; (c) The equivalent potential temperature has a value of about 370 K at the eye center (0–15 km radius), 360 K at the nominal eyewall (40–50 km radius), 355 K in the outer core (about 200 km radius) and about 350 K in the ambient environment (300–1000 km radius) at the sea surface. Within the first 2 km above the ocean
3.3
(d)
(e)
(f)
(g)
The Eye and the Eye-Wall
75
surface, it varies little with height within the eye and the eyewall, but outside the eyewall it decreases steadily with height; The low-level inflowing air converges into the circulation which diverges out from the central region of the eye and the convergent currents flow into the base of the eyewall and rise in penetrative convection with heavy condensation inside the eyewall. However, condensation heating leads to strong divergence above about 1.5 km level, one branch turning inward towards the center of the eye at a height of about 2 km and above and the other turning upward and outward; The inward-turning branch of the diverging current above 2 km would appear to converge at the center of the eye and subside to lower levels only to diverge out near sea level towards the eyewall in some kind of a vertical secondary circulation; The secondary vertical circulation between the center of the eye and the eyewall adjusts itself suitably to accommodate a mesovortex if one happens to be present; As pointed out earlier, potential energy generated by subsidence of the warmer air at the center and lifting of the colder air in the eyewall in the secondary circulation would appear to be what provides for the kinetic energy of the hurricane circulation in which the warmer air rises in the eyewall and the colder air subsides in the outer environment.
3.3.3 Concentric Multiple Eye-Walls A tropical cyclone is a continually evolving system in which more than one eye-wall can form at different radial distances from the center, especially in intense cyclones. McNoldy (2004) reports a case of three concentric eye-walls which were observed in a hurricane named ‘Juliette’ in the eastern North Pacific during the period 23–27 September 2001. From details of observations by satellite as well as reconnaissance aircraft, McNoldy presents the radial distributions of tangential velocity and relative vorticity in the northwest quadrant of the hurricane at a height of about 3 km on 25 and 26 September, shown in Fig. 3.11. The hurricane developed off the coast of southern Mexico and moved almost parallel to the coastline to enter Baja California later in the month. The northwest quadrant was chosen for its higher resolution and better quality data on both days. The relevant features are the two peaks in tangential velocity at 9 and 58 km (with corresponding peaks in relative vorticity at 7 and 55 km) on 25 September and the three peaks in tangential velocity at 11, 56, and 90 km (with corresponding peaks in relative vorticity at 9, 54, and 82 km) on 26 September. In this simplified radial perspective provided by Fig. 3.11, the relative vorticity is produced solely by the radial gradient of the tangential wind. The steepness of the vorticity curves may play a role in the formation of moats. From the diagram, it is clear that the sharpest drop-off of relative vorticity occurs on the outer side of the innermost wall, that on outer side of the second and the third eyewalls being much less prominent.
76
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.11 Radial profiles of (top) tangential velocity and (bottom) relative vorticity through the northwest quadrant of Juliette on 25 and 26 September 2001 at 1819–1849 and 1722–1755 UTC, respectively (after McNoldy, 2004)
Once the initial eye-wall is formed, the positive relative vorticity falls steeply beyond the radius of maximum tangential velocity, establishing some kind of a moat beyond which deep convection is free to organize a new concentric eye-wall. Sometimes in just a few hours, the new outer eye-wall will dominate and the inner eye-wall will dissipate. The outer eye-wall will then move in to replace the inner eye-wall. The process may continue at various radial distances to produce more than one concentric eye-wall. Of course, it is the innermost eye-wall which will dominate, the intensity of the others falling off sharply with radial distance. As McNodly states: ‘Concentric eyewalls are ephemeral; once formed, they typically are not maintained for much longer than 12 h As the new outer wall forms, the original inner eyewall usually lacks the necessary inflow to maintain itself and
3.4
Spiral Bands Around the Eye-Wall
77
it gradually dissipates. In time, the dynamic processes cause the outer eyewall to contract, and the process can repeat itself.’ Black and Willoughby (1992) had called the process an eyewall replacement cycle. Observations of multiple concentric eyewalls in hurricanes are rather scarce, but from a study of western North Pacific tropical cyclones during 1969–1971, Willoughby et al. (1982) estimated that approximately 53% of intense tropical cyclones (windspeeds greater than 65 m s–1 ) exhibit concentric eyewalls. More often than not, individual convective cloud cells at the new eyewalls may move in bands spiralling inward to converge into the innermost eye-wall.
3.4 Spiral Bands Around the Eye-Wall 3.4.1 Structure Almost all mature tropical cyclones exhibit a number of (usually two or three) welldeveloped cloud bands spiralling inward into the central cloud mass (see Fig. 3.1). What are these cloud bands and how do they form? While there is considerable uncertainty about their origin, Senn et al. (1957) has shown that the geometrical structure of many of these bands follows a modified logarithmic structure Ln(r − r0 ) = A + Bλ
(3.4.1)
where r is radius, λ azimuth, r0 the radius of a limiting circle with center coincident with the cyclone center, and A is a constant that specifies the angular origin (orientation) of the spiral, The constant B determines the crossing angle α between the spiral path and the circle about the center according to the relation tan α = B(1 − r0 /r)
(3.4.2)
The constant B is evaluated by noting α at an infinite distance from the center, so that (r0 /r) tends to 0, and B = tan α. The spiral bands consist of large convective cloud cells associated with smallscale cumulus convection. The individual cells move with the mean wind in the layer in which they are embedded within the spiral bands which move at a much slower rate. The cloud cells which move through the bands have typical life spans of 20–40 min and last as long as they experience rising motion within the band in convectively unstable air. New cells form at the upwind (inner) side of the band, travel through the band, and dissipate at the downwind (outer) side of the band. The cells yield heavy precipitation while they move through the band. Some of the individual cloud cells within the band appear to move inward towards the center of the cyclone, while others move outward. However, the number of those that move inward appears to predominate. A schematic model based on several observational studies of the structure of spiral bands indicates that a meso-scale trough of low pressure of amplitude about
78
3
Tropical Cyclones/Hurricanes/Typhoons
1–2 mb occurs at the leading edge of the spiral band with heaviest rainfall occurring about a quarter wavelength behind the trough, the width of the spiral band being half a wavelength. This structure is consistent with the phase relationship of a shallow water gravity wave. The mean low-level winds show convergence into the trough of low pressure in front of the band and divergence in the rear. At the approach of a band, the mean wind veers by 10–20◦ and then backs towards the original direction after the passage of the band. Because of strong convection in the band, the surface wind often becomes highly gusty during its passage. Gentry (1964) analyzed many aircraft-observed profiles of wind, temperature, humidity and pressure (D-values) along and perpendicular to spiral bands and found that the vertical velocity and temperature perturbations in the bands are positively correlated, with warm updrafts and cold downdrafts, thereby producing kinetic energy within the bands which in magnitude was almost comparable to the kinetic energy produced by the large-scale circulation associated with the cyclone. It is likely, however, that much of the perturbation kinetic energy produced locally in the bands is offset by local dissipation due to a high degree of turbulence in the vicinity of intense convection.
3.4.2 Origin and Direction of Propagation Observational evidence regarding the origin and direction of propagation of spiral bands varies, although the predominant evidence supports an outward propagation of the band relative to a stationary storm center (propagation velocity of the band after the velocity of the storm has been subtracted). According to Senn and Hiser (1959) and several others, the bands originate near the eye and propagate radially outward. Fletcher (1945) hypothesized that bands of clouds from the intertropical convergence zone were drawn into the cyclone’s circulation and became the spiral bands. Wexler (1947) suggested that cloud streets which are common in the tradewind atmosphere coil into the cyclone. Tepper (1958) proposed that gravity waves were generated near the eye-wall and propagated outward to become spiral bands. A number of theoretical and simulation studies (e.g., Abdullah, 1966; Kurihara, 1976; Diercks and Anthes, 1976; Willoughby, 1977, 1978a,b) have also been carried out to explain the generation, structure and propagation of hurricane spiral bands. A detailed review of some of these studies has been presented by Anthes (1982).
3.5 Storm Surge 3.5.1 Introduction Cyclones disturb the ocean as well as the atmosphere. Out over the open ocean, the extremely low pressure under a cyclone raises the water level producing a mound, the height of which varies with the pressure anomaly. The high winds drag the ocean
3.5
Storm Surge
79
surface producing stresses and ocean currents and also mighty waves which are harmful to ocean-going vessels which might happen to be in its vicinity. But, by far, the greatest damage is done when it approaches an inhabited coastal belt and produces what is known as a storm surge at the coast. The water level may then build up to several meters which can inundate a whole coastal region inflicting death and destruction to life and property. It is estimated that the storm surge produced by the Bangladesh cyclone in November 1970 in the delta region of Ganga-Brahmaputra killed as many as 300,000 people apart from flooding extensive low-lying areas. The storm surge problem is, however, complicated, because it depends not only upon the characteristics of the cyclone but also upon the vertical profile of the ocean floor near the coast, such as the depth and slope of the sea-bottom, orientation of the coastline to the storm track, presence of a bay, peninsula, island or estuary, etc, in the path of the storm. Further, the possible effect of an astronomical tide on the rise of sea level also needs to be taken into consideration. The superposition of an astronomical high tide on the effects of the cyclone can magnify storm surge to dangerous levels. A recent example of the extent of damage and destruction that a storm surge can inflict upon a coastal community is that associated with the category-5 hurricane, Katrina, which on August 29, 2005, while crossing the coast of Louisiana from the Gulf of Mexico near Buras-Triumph broke three imposing concrete levees which protected the city of New Orleans and flooded the city killing at least 1420 people and causing a material loss to the tune of $75 billion dollars. It was one of the costliest hurricanes so far in American history.
3.5.2 Some General Aspects of Storm Surge The total rise of water level in a storm surge is produced by the superposition of several scales of wave motion in the ocean. The large-scale motion of period of several hours may be of the scale of the cyclone itself and measured by the radius of the maximum wind or any such length. The pressure drop associated with the cyclone and the wind stress averaged over horizontal scales of several kilometers and period of an hour or so are the forcing functions for the large-scale oceanic waves. Typical surge amplitudes associated with this scale may range from 2 to 6 m. Then, there are smaller scale fluctuations of water level produced by individual waves the amplitudes of which may vary from 1 to 10 m and the period of which may be a few seconds only. But, by far, the greatest rise of water level and for a somewhat longer duration is caused when the effects of local ocean-bottom and coastal topography as well as astronomical tides are superimposed on the responses due to large-scale and small-scale ocean waves. Because of the sensitivity of storm surge to all these varying factors, including the occasional tsunamis, it is difficult to present a standard model of the phenomenon. The simple picture described above is, however, strongly modified by coastal topography and configurations. A bay, for example, may double the amplitude of the storm surge. In some bay region, the ocean bottom slopes up to shallow waters a long distance before the coastline and a cyclone heading towards such a coastline
80
3
Tropical Cyclones/Hurricanes/Typhoons
produces disastrous storm surge effects. Such bay regions can be found over several parts of the globe, for example in the Ganga-Brahmaputra delta region of IndiaBangladesh at the northern boundary of the Bay of Bengal, or the Mississippi delta region of USA off the coast of Louisiana at the northern boundary of the Gulf of Mexico, where extensive storm surge flooding occurs whenever a severe cyclone crosses these coasts.
3.5.3 Mathematical Models of Storm Surge Since storm surge involves nonlinear interactions between the atmosphere, the ocean and bottom topography, it is impossible to devise a mathematical model which can be solved analytically. Only highly simplified numerical models may be tried to determine the approximate oceanic response under actual cyclone conditions. In a pioneering study, Jelesnianski (1965, 1966, 1967) developed a general numerical model of the storm surge and has used it to test various physical hypotheses regarding the effects of (a) pressure anomaly; (b) wind stress, and (c) intensity, size, speed and direction of motion of the cyclone in producing the storm surge. Jelesnianski’s storm surge models consider only the barotropic mode. The linearized equations of his model (Jelesnianski, 1966) consist of predictive equations for the depth-weighted velocity components U and V, and the continuity equation for h, the departure in height of the surface water from the undisturbed level. The equations are: ∂U/∂t = −gD(x,y)∂h/∂x + fV + {D(x,y)/ρ}∂Pa /∂x + ρ −1 τx (x,y,t)
(3.5.1)
∂V/∂∂t = −gD(x,y)∂h/∂y − fU + {D(x,y)/ρ}∂Pa /∂y + ρ −1 τy (x,y,t)
(3.5.2)
∂h/∂t = −(∂U/∂x + ∂V/∂y)
(3.5.3)
where D(x, y) is the initial (undisturbed) depth of the fluid, τ is the surface wind stress per unit mass, and Pa is the atmospheric pressure at the surface. Equations (3.5.1), (3.5.2), and (3.5.3) describe the linearized response of a shallow layer of the fluid to the forcings of atmospheric pressure gradients and surface wind stress, τ (x, y, t), on a rotating earth. Besides forced modes, the system allows inertia-gravity waves. A typical ocean basin is resolved as shown in Fig. 3.12. Finite-difference technique is adopted to solve the Eqs. (3.5.1), (3.5.2), and (3.5.3). The model is forced by specifying Pa (x, y, t) and τ (x, y, t) from the field of either an actual cyclone or a model cyclone. In Jelesnianski’s simulations, the pressure and stress fields are specified by empirically determined functions of space. Storm parameters included in the functions are the intensity (maximum wind), size (radius of the maximum wind), and an inflow (cross-isobar) angle. From the wind distributions, the surface eddy stress τs was estimated using the bulk aerodynamic relation
3.5
Storm Surge
81
Fig. 3.12 A typical one-dimensional ocean depth profile near a coast. A finite-difference grid with variable horizontal resolution S is shown. The curved bottom topography is approximated by a series of steps with constant depth Di (i = 1, 2, 3,. . .). Higher resolution may be introduced near the coast (Jelesnianski, 1966)
τ s = cD ρa |Va |Va where cD is the drag co-efficient and the subscript ‘a’ refers to the air. In some preliminary experiments, Jelesnianski (1965) investigated the ocean’s response to a variety of prescribed storm conditions, such as (a) a circular wind stress only, (b) a wind stress with a radial (inward) component introduced by a 30◦ cross-isobar angle, (c) an asymmetric wind stress associated with a moving storm, and (d) an atmospheric pressure gradient only. Two interesting results emerged from these preliminary experiments. The first was a 60% increase in the maximum surge at the coast by inward flow of water when a radial component to the stress was introduced. The second was a three times increase in the static height ‘h’ of the mound of water under the storm at the coast when the barometric pressure gradient alone was used to force the ocean (the static height is the height of the mound of water under the cyclone above the undisturbed level when it is far out over the ocean and is given by the relation, h = Pa /ρg, where ρ is the density of water). It turned out from the preliminary experiments that the two major forces driving the storm surge are the wind stress and the atmospheric pressure gradient. To test the relative importance of the two forcings, the model war run with each force separately and then with the combined forces. The results showed that the effect of the wind stress was approximately double of that due to the pressure gradient. This would be clear from Fig. 3.13 which shows the maximum coastal surge for the
82
3
Tropical Cyclones/Hurricanes/Typhoons
Fig. 3.13 Maximum coastal surge for forces associated with wind stress and pressure anomaly alone and combined forces as a function of speed of storm crossing coast at an angle of 90◦ . The dots are values from actual computer runs (Jelesnianski, 1966)
three combinations of forces when the storm crosses the coast at an angle of 90◦ with different speeds. Note that the peak storm surge obtained from the combined forces is not a simple sum of the surges obtained from the forces separately. Also, the maximum storm surge does not vary linearly with the storm speed, but reaches a maximum at an approximate speed of 30–40 knots. Jelesnianski (1966) simulated storm surge effects on the same coastal topography for many different model storms and constructed nomograms showing the quantitative effects of storm speed, intensity and direction of approach on the characteristics of the storm surge. The reader interested in further details of these effects and the nomograms may consult his original papers.
3.6 Prediction of Cyclone Track and Intensity From very early times, tropical cyclones have drawn the attention of meteorologists because of their power to destroy life and property wherever they occur. A demand for accurate forecast of these storms was, therefore, inevitable. Prior to the days of regular meteorological observations, the only guidance available in this regard was observations of movement of high clouds and sea swells in coastal areas. However, the deductions made from such observations were often misleading. It was found that plumes of high clouds diverging from the upper part of a cyclone did not necessarily move in the direction of motion of the storm to provide early warning of
3.6
Prediction of Cyclone Track and Intensity
83
its arrival. Further, sea swells were not exclusive to tropical cyclones. They could be caused by other agencies as well, such as underwater volcanoes and earthquakes (tsunamis). For example, the recent tsunami (high tidal wave) in the Indian Ocean area which killed nearly 175,000 people in the littoral countries originated from an underwater earthquake near the coast of Sumatra in Indonesia on 26 December 2004, not from a tropical cyclone.
3.6.1 Early Models – The Steering Concept With the availability of upper-air observations in the nineteen fifties and sixties, one of the methods used to predict cyclone motion with considerable success utilized what is known as the steering principle. According to this principle, small-scale cyclones tend to move with the large-scale currents in which they are embedded. Initially, a single level wind was used as the steering current but later it was realized that a layer-mean wind might be a better steering current than a single-level wind. The realization gave rise to other models. However, the basic concept of the steering current survived and found a place amongst the latest models of storm track forecasting. Since 1950s and 1960s, a number of objective methods have been developed to predict cyclone motion. These are either statistical or dynamical methods. Statistical methods seek to relate predicted movement to a number of cyclone parameters in an empirical way. The dynamical methods, on the other hand, use some form of the equations of motion to predict numerically the motion of the cyclone from an observed initial state of the atmosphere. In some objective methods, the output from a dynamical model is used in a statistical model. These are called hybrid models. Neumann and Pelissier (1981a,b) described a total of seven operational models which were in use at the National Hurricane Center (NHC) in USA for prediction of tropical cyclone motion in the North Atlantic Ocean area (Table 3.2).
3.6.2 Current Models During the eighties and nineties, there were continued efforts to improve the hurricane track forecast models in the light of experience gained during the earlier years. In 1988, SANBAR was replaced by the Quasi-Lagrangian Model (QLM) which had higher horizontal and vertical resolution and a larger model domain (Mathur, 1991). The QLM produced skillful forecasts during 1989–1993. Meanwhile, in 1992, the Geophysical Fluid Dynamics Laboratory (GFDL) developed a model which was adapted for real-time hurricane track forecasting. It included moving nested grids and a sophisticated initialization scheme. The horizontal grid spacing on the inner mesh was about 20 km, which was about half of that in QLM. The GFDL model outperformed all other models in terms of average track error when it was tried during the 1992–1993 hurricane seasons and replaced the QLM in 1995.
84
3
Tropical Cyclones/Hurricanes/Typhoons
Table 3.2 Operational models for the prediction of tropical cyclone motion over the North Atlantic area (after Neumann and Pelissier, 1981a,b) Model
Model type
Description (reference)
HURRAN
Statistical
CLIPER
Statistical
NHC-67
Statistical
NHC-72
Statistical
NHC-73
Statistical
SANBAR
Dynamical
MFM
Dynamical
Analog model based on tracks of all Atlantic tropical cyclones since 1886 (Hope and Neumann, 1970) Regression equation model, utilizing predictors derived from climatology and persistence (Neumann, 1972) Regression equation model utilizing predictors derived from climatology, persistence and observed geopotential height data (Miller and Chase, 1966) Regression equation model utilizing predictors derived from output of CLIPER model and observed geopotential height data (Neumann et al., 1972) Regression equation model utilizing predictors derived from output of CLIPER model, observed and numerically forecast geopotential height data (Neumann and Lawrence, 1975) Barotropic model based on pressure-weighted wind field averaged through troposphere and represented on a 154 km (at 22.5◦ N) space-grid (Sanders and Burpee, 1968, Sanders et al., 1975) Movable Fine Mesh (MFM) baroclinic model having 10 levels in the vertical and 60 km grid spacing in the horizontal (Hovermale and Livezey, 1977)
Since 1989, after withdrawal of SANBAR, a couple of other barotropic models have been tried. One of these was the VICBAR (Vic Ooyama Barotropic model) which has been run in NHC in an experimental mode in real time. The main advantages of VICBAR relative to SANBAR was that it obtained lateral boundary conditions from the National Meteorological Center (now National Centers for Environmental Prediction, NCEP) global forecast model and used a much more accurate numerical method. Another model which used the barotropic steering concept is the Beta and Advection Model (BAM). In this model, the track forecast is obtained by following a trajectory in the vertically averaged horizontal wind from the NCEP aviation global forecast model, after applying a correction for vortex drift. The BAM model which uses a layer-mean wind for steering has been used for shallow, medium and deep layers. All three versions are in use at NHC.
3.6
Prediction of Cyclone Track and Intensity
85
The large-scale regional and global models from NCEP have long been used as an aid in hurricane forecasting by providing forecasts of the hurricane environment. The skill of the global forecast models have improved over the hyears (Kalnay et al., 1994). This improvement can be attributed, inter alia, to improved model physics, increased horizontal and vertical resolution, and better data assimilation techniques. Beginning in 1992, a tropical cyclone bogussing system was implemented in the NMC operational global model for track forecasting (Lord, 1991). Currently, for cyclone track forecasting in the Atlantic, a number of models are in operation in NHC. These include: The simple analog model (HURRAN). CLIPER, the statistical-dynamical models (NHC-90), the barotropic model (VICBAR), BAM (3 versions), the GFDL regional baroclinic model, and the aviation model. All of these models, except CLIPER, use information from the aviation model. A version of the NHC-90 model, called UK-90, uses information from the UK Meteorological Office global forecast model. There were also efforts during this period to improve hurricane intensity forecasting. Experience showed that the intensity of a cyclone was very much influenced by environmental conditions, such as sea surface temperatures, upper-level flow patterns, temperatures near the tropopause, etc. These ideas were incorporated in a Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic basin which has been run in NHC on an experimental basis since 1990. Intensity prediction has also been attempted in the GFDL model, beginning in 1992 (Kurihara et al., 1993). Thus, a beginning has been made to use three-dimensional global circulation models to predict both track and intensity of hurricanes.
Part II
Tropical Monsoons over Continents and Oceans
Chapter 4
Monsoon over Southern Asia (Comprising Pakistan, India, Bangladesh, Myanmar and Countries of Southeastern Asia) and Adjoining Indian Ocean (Region – I)
4.1 Introduction – Physical Features and Climate The world’s largest and most powerful monsoon circulation develops over the region of Southern Asia and its adjoining Indian Ocean. There are several reasons for this development: The first is a most favorable land-sea distribution. The geographical location of the region (see a physical map in Fig. 4.1) with the Tropic of Cancer passing through almost the middle of the landmass of the region and the Tropic of Capricorn through the middle of the Southern Indian Ocean (indicated by thin dashed lines) appears to provide an ideal setting for a heat source to develop over the land and a heat sink over the ocean during northern summer and vice versa during northern winter. A second reason is orography. The lofty Himalaya Mountains and the Tibetan Plateau standing along the northern boundary of the region and rising steeply from the plains of northern India to peak heights of 8–9 km above sea level and then an average Plateau height of 4–6 km above mean sea level not only protect it from the icy cold winds of Central Asia during the winter but also help shape the structure of a well-defined summer monsoon circulation over the region and provide a natural barrier to the moisture-laden onshore winds from the Indian Ocean blowing into the region. The high mountain ranges all along the northwestern, northern and the eastern boundaries of the region as well as the coastal mountain ranges, such as the Western Ghats Mountain of peninsular India and the Arakan Yoma and the Tennaserim Ranges of Myanmar (erstwhile Burma) play important roles in shaping the structure of the monsoon circulation and the distribution of monsoon rains over the region. The geographical location of Southeastern Asia, sandwiched between the Bay of Bengal and the South China Sea, presents a complex array of narrow coastal mountain ranges and broad inland high plateaux with heights ranging from 1.5 to 2.5 km above sea level. Some of these ranges which are mostly north-south oriented rise steeply above the plains of Myanmar, Thailand and Malaysia. It will be shown later in this chapter that the geographical locations of these mountain ranges and high plateaux, both along the northern and the eastern borders of the Indian Subcontinent play crucial roles not only during advance and retreat of monsoon, but also in controlling the distribution of monsoon rains over the subcontinent. K. Saha, Tropical Circulation Systems and Monsoons, C Springer-Verlag Berlin Heidelberg 2010 DOI 10.1007/978-3-642-03373-5_4,
89
90
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.1 Physical map of Southern Asia and adjoining Indian Ocean and its littoral countries
In the western sector, a chain of semi-arid and desert lands extends from Pakistan and Iran southwestward to Saudi Arabia and beyond to East Africa across Somalia, Kenya, Tanzania, Zimbabwe to Mozambique and South Africa against the backdrop of the high East African Mountains with heights ranging between 2 and 3 km asl. It will be shown in the present text that these varied land features along the western boundary of the Indian Ocean play crucial roles in advance and withdrawal of monsoon as well as in shaping the structure of the circulation and distribution of monsoon rains over the Western Indian Ocean and its littoral countries during both winter and summer. The seasons over the region which exercise great influence on climatic conditions are as follows: The winter season characterized by extremely cold and dry conditions over the northern parts of the region appears to cover a period of 3 months from December to February. Then comes a period of approximately 3 months, March to May, when the land surface warms up rapidly to very high temperatures and ‘heat lows’ develop over the land. This is the period of transition from winter to summer monsoon. Summer monsoon proper arrives towards the end of May or early June and covers the whole region including the northernmost parts of the Western Himalayas by end of August. It starts withdrawing from the northwestern part of the region from early September, but the process of withdrawal is not completed till end of November.
4.2
The Winter Season (December–February)
91
Thus, the total period of summer monsoon over the region, from its first onset to final withdrawal, is almost 6 months. It, therefore, stands to reason that in the present text, the period of summer (wet) monsoon be divided into two parts: the onset phase (June to August) and the withdrawal phase (September–November), in order to conform to ground realities. The transition from summer to winter monsoon conditions varies from place to place but is usually about a month. The winter monsoon is of shorter duration. It is important to bear in mind the great diversity of the seasonal weather and climate over the Indian Subcontinent arising largely from its topography and latitudinal extent from about 5 to 35◦ N. For example, in winter (December–February), while the northern parts of the subcontinent (north of about 25◦ N) shiver in nearfreezing temperatures at times, the Indian peninsula and a large part of SE Asia enjoys cool, salubrious climate. Conversely, during the SW monsoon season, cool, humid weather of the south and the southeast stands in sharp contrast to the hot, semi-arid desert climate of the NW India and Pakistan.
4.2 The Winter Season (December–February) 4.2.1 General Climatic Conditions In winter, the surface temperatures are generally low over the subcontinent and decrease with latitude. The lofty Himalayan mountain ranges along the northern boundary separate the region from Central Asia and protect it from the icy cold temperatures of that region. Relatively warmer temperatures prevail over the Indian Ocean, with the warmest temperatures being over the equatorial region where the ITCZ is located along about 15◦ S over the western Indian Ocean, but is much closer to the equator over the eastern part. In keeping with the above-mentioned temperature distribution, m.s.l. pressure is generally high over the cold Indian Subcontinent and low over the warm equatorial ocean. An interesting aspect of the pressure distribution is that the three land segments of Southern Asia, viz., the Arabian Peninsula, the Indian Subcontinent, and Southeastern Asia, develop their independent high pressure cells, with low pressure troughs located in between over the somewhat warmer oceans; one over the northern Arabian Sea and the other over the northern Bay of Bengal. The orographic influence of the Himalaya mountain ranges and other coastal mountain systems, such as the Western Ghats of India and the Arakan Yoma of Myanmar is also significant. Generally, across any of these mountain ranges of appreciable height, a trough of low pressure tends to form on the windward as well as the lee sides of the mountain range with a ridge of high pressure on the mountain itself. Accordingly, troughs of low pressure are observed to form along the foothills of the Himalayas; one over Pakistan in Western Himalayas and the other over West Bengal and Bangladesh in the eastern part of the subcontinent.
92
4 Monsoon over Southern Asia and Adjoining Indian Ocean
The prevailing quasi-stationary pressure distribution drives anticyclonic circulation around the high pressure cells and cyclonic circulations around the troughs of low pressure. As part of the anticyclonic circulations, low-level airflow is generally northeasterly over the northern Indian Ocean. However, near the equator, the airflow is cross-equatorial in the western part of the ocean (west of about 70◦ E), and more or less westerly in the eastern part. The equatorial westerlies in the eastern part of the ocean divide the warm low pressure over the area into two cells, one north and the other south of the equator, each having its own trough, viz., the North Equatorial Trough (NET) and the South Equatorial Trough (SET). According to rainfall map prepared by the India Meteorological Department (1943) (not shown), there is little rainfall over the subcontinent during the season, except areas affected by traveling disturbances.
4.2.2 Disturbances of the Winter Season 4.2.2.1 Western Disturbances (W.D.) These are large-scale wave disturbances which form in midlatitude-subtropical baroclinic westerlies and usually travel from west to east. During northern winter, they travel from the Eastern-Mediterranean region and enter the Indian region from November onward. Their arrival is heralded by appearance of first high clouds and then medium and low clouds with rain or snowfall in the mountains of western Himalayas. With advance of the season, their track continually shifts equatorward and more and more of the subcontinent are affected by them and experience rain or snowfall. In December, western disturbances cross the mountain ranges of Iran and Afghanistan and arrive over NW Pakistan and adjoining western Himalayas in a diffused state with distorted structure, but on arrival over the plains of Pakistan, some of them interact with the pre-existing orographically-maintained trough of low pressure and regain their frontal structure and identity. The interaction leads to strengthening of the wave disturbance. Those developing large amplitude may draw moisture from the Arabian Sea and strengthen further. The rejuvenated disturbance (which is popularly called a ‘Western Disturbance’ in the subcontinent) then generally moves eastward, often in an occluded state, affecting southern Himalayas and adjoining plains of northern India. Some of them travel as far as Bangladesh and Assam and the mountains of eastern Himalayas where they usually break up and disappear. While passing over the plains of northern India, they draw moisture from the Bay of Bengal and are often accompanied by thundery rain and squally conditions. The average lifetime of a western disturbance is 4–6 days and there may be 6–8 disturbances per month at the peak of the season. Farmers welcome winter rain, for it helps in the cultivation of winter crops, such as wheat, etc. There is also heavy snowfall in the western Himalayas, the depth of snow occasionally exceeding 15 metres and causing dangerous avalanches. In the wake of some of these disturbances, dense fog may appear over large tracts of northern India and persist for days, interfering with road, rail and aerial communications. During the passage
4.3
The Transition Season (March–May)
93
of these disturbances, the subtropical westerly jetstream may often reach a speed of 60–75 m s–1 at about 200 hPa over the latitude belt 25–30◦ N.
4.2.2.2 Easterly Waves and Cyclonic Storms of the Northern Indian Ocean A surge in the low-level E/NE-ly tradewinds that converge into the equatorial trough of low pressure over South China Sea often give rise to easterly waves which travel westward. Arriving over the Bay of Bengal, some of them develop into westward-propagating depressions and cyclones. Their formation and movement are clearly seen in day-to-day satellite cloud imagery. Large cloud clusters and rainfall associated with them are observed to move westward a few degrees north of the equator. During their movement, they may strike the coast of Sri Lanka and southern Tamilnadu. Once in a while, they may cross the peninsula and emerge into the Arabian Sea. But such occasions are rare. But not all surges in the tradewind easterlies develop into waves and cyclonic disturbances. Freeman (1948) draws an analogy between these surges and supersonic gas flows and concludes that surges lead to hydraulic jumps in the easterly flow and formation of cloud lines which move downstream over long disturbances.
4.2.2.3 Easterly Waves and Cyclones of the Southern Indian Ocean Westward-propagating easterly waves also form south of the equator when a surge in the low-level SE-ly tradewinds, or the monsoon northwesterlies, converges into the circulation around the SET and increases its cyclonic vorticity. The enhanced vorticity gives rise to westward-propagating cyclonic disturbances. After formation, these disturbances usually move westward as lows or depressions. However, a few of them which survive and enter the western Indian Ocean (west of about 80◦ E) recurve southwestward and develop into tropical storms and cyclones. Several islands of the region including Madagascar and Mauritius are badly hit by these cyclones almost every year. Beyond these islands they recurve further around the subtropical high pressure belt. Some of them are drawn into the circulation of the midlatitude wave disturbances which sweep across South Africa and move eastward across the Southern Indian Ocean towards Australia – New Zealand.
4.3 The Transition Season (March–May) Several important developments take place in the weather over the Subcontinent and the adjoining Indian Ocean during the transition season. These include: (1) Passage of WDs across the northern part of the Subcontinent; and (2) Development of powerful ‘heat lows’ over land and relatively ‘cold highs’ over neighboring oceans.
94
4 Monsoon over Southern Asia and Adjoining Indian Ocean
The equatorial Indian Ocean is, perhaps, one region which becomes most active during this period when summer monsoon withdraws from the southern hemisphere and enters the northern hemisphere. Varieties of interesting phenomena are observed over the equatorial region about this time.
4.3.1 Western Disturbances Western disturbances, described in Sect. 4.2.2.1, continue to be active over the northern part of the subcontinent during the transition season, though their track gradually shifts poleward with advance of the season. During their eastward passage, they draw additional moisture from the Arabian Sea as well as the Bay of Bengal and cause increased convective rainfall on the plains of northern India and extensive snowfall on the mountains.
4.3.2 ‘Heat Lows’ over Land and ‘Cold Highs’ over Ocean During March and April, ‘heat lows’ form over the Indian Subcontinent as well as other land areas bordering the Northern Indian Ocean, such as Somalia and Saudi Arabia in the west, and Malaysia, Thailand and Myanmar in the east, while the adjoining oceans, the Arabian Sea and the Bay of Bengal continue to be under ‘cold highs’. By mid-May, the heat lows over the land areas develop further and move further north to take up their northernmost summer locations. The heat low over the Indian Peninsula also deepens further with its trough at surface oriented in a direction almost paralleling the east coast of the peninsula. While low level convergence and convection occur in the boundary layer of these heat lows, strong subsidence and divergence prevails in the upper troposphere above them. With adiabatic cooling of the rising air and subsidence warming aloft, the result is a stable stratification in the vertical in these heat lows at this time. The scenario changes when, under the influence of the prevailing pressure distribution, low-level winds diverging from the neighboring oceanic high pressure cell inject cool, moist air into the heat low circulation over the land along the coastal belts of Southeastern India and Bangladesh. The injection of moisture at low levels in this manner makes the atmosphere over the region, especially the coastal belts, conditionally unstable.
4.3.3 Severe Local Storms Over central and eastern parts of the Subcontinent where a lot of latent instability energy is stored in the atmosphere due to influx of cool, moist air from the Bay of Bengal at low levels, widespread thunderstorms and hailstorms occur whenever a Western Disturbance affects the region. Locally, in Bengal, these storms are known
4.3
The Transition Season (March–May)
95
Fig. 4.2 Illustrating the formation of a Kal-Baisakhi: (a) A plan view showing locations of lowlevel troughs (double-dashed), (b) a vertical section showing superposition of a W’ly jetstream (J) on low-level trough. DV – divergence, CV – convergence
as ‘Kal-Baisakhis’ or deadly storms of the month of Baisakh (April–May). They are accompanied by thunder and lightning, squally winds, and heavy downpours. Many of them breed deadly tornadoes which inflict heavy loss of life and property. Figure 4.2 shows schematically a typical synoptic situation favorable for occurrence of a Kal-Baisakhi over eastern India and adjoining Bangladesh. The left panel of Fig. 4.2 shows the heat low troughs (Tr), to which cool, moist air from the Bay of Bengal converges and rises in strong penetrative convection when a upper-air W’ly trough approaches the region and has its divergent area associated with upper-level Jetstream (J) superimposed upon the pre-existing trough with lowlevel moisture convergence (right panel).
4.3.4 Developments over the Equatorial Indian Ocean During transition season, summer monsoon withdraws from the southern hemisphere and enters the northern hemisphere. The movement is marked by several interesting events over the equatorial Indian Ocean, including the following: (a) Cross-equatorial movement of Monsoon Circulation; (b) Formation of Equatorial Westerlies, Double Equatorial Troughs and Cloud Bands; (c) Increased Cyclonic and Anticyclonic Activity over the Equatorial Zone (a) Cross-Equatorial Movement of Monsoon Circulation: With increased warming of the earth’s surface and falling of pressure to the north of the equator and cooling and rising of pressure to the south and differential rate of heating between land and ocean, a gradient of pressure tendency develops between the two sides of
96
4 Monsoon over Southern Asia and Adjoining Indian Ocean
the equator, which enables the monsoon current of the southern hemisphere to cross the equator and move into the northern hemisphere to start the process of its advance towards the Indian Subcontinent. An example of this type of forcing for equatorial crossing and relocation of the monsoon current from the winter to the summer hemisphere may be seen in Fig. 4.3 which shows the distribution of mean sea level pressure at 12 GMT on 7 April 2008 when monsoon appeared to have crossed the equator in 2008 in the Western Indian Ocean. According to the isobaric field shown, a pressure gradient exists in the vicinity of the East African coastline not only between the two sides of the equator but also along the equator from west to east. Also, north of the equator, a pressure gradient exists between the ocean and the land across the coast of Somalia. These pressure gradients appear to provide a safe passage for a parcel of air approaching the equator from the south to move along a path indicated by a thick continuous line with arrow in Fig. 4.3. The new locations of the cold sector of the monsoon and related ITCZ (dashed) north of the equator are also shown. (b) Formation of Equatorial Westerlies, Double Equatorial Troughs: The low level winds and circulation over the equatorial Indian Ocean during the transition season show a broad band of equatorial westerlies between two troughs of low pressure, one in each hemisphere. Tradewinds diverging from the oceanic high pressures converge into these equatorial troughs forming penetrative convection and Cloud
Fig. 4.3 Map showing MSLP (mb) at 12 Z, 7 April 2008 over the Indian Ocean and littoral countries Thick continuous line shows how the cool humid monsoon current crosses the equator, with ITCZ (dashed) located to its east (NCEP Reanalysis)
4.3
The Transition Season (March–May)
97
Fig. 4.4 Wind vectors and streamlines over the equatorial Indian Ocean at 10 m above sea surface and ocean surface temperatures (◦ C) at 12 GMT, 14 April 2008. Thick continuous lines show the locations of the troughs (adapted from NCEP Reanalysis)
Bands, one on either side of the equator, during a short period of equatorial crossing of monsoon. Just how this happens is demonstrated by Figs. 4.4 and 4.5 which shows winds and streamlines at 10 m above mean sea level and 850 hPa at 12 GMT on 14 April 2008, approximately a week after equatorial crossing. The double equatorial convergence zones are characterized by formation of double cloud bands, one on each side of the equator, the presence of which is revealed in satellite cloud imagery (see Fig. 4.6). Double equatorial troughs and associated cloud bands are persistent features of the circulations over the equatorial Indian Ocean during the transition season. However, they are short-lived and usually observed around the time of equatorial crossing only. (c) Increased Cyclonic Activity over the Equatorial Eastern Indian Ocean: The spring transition season appears to be the period of maximum cyclonic activity over the equatorial eastern Indian Ocean as well as Arabian Sea (see Fig. 4.7) The reason why the equatorial eastern Indian Ocean is so cyclogenetic during the transition season is to be sought, inter alia, in the warm ocean surface temperature of the Bay of Bengal. Further, most of the cyclonic storms develop around monsoon troughs which interact with traveling disturbances such as E’ly and W’ly waves. After development, the cyclonic disturbances move northwestward and later recurve
98
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.5 Wind vectors and streamlines at 850 hPa showing the locations of two equatorial troughs, one on either side of the equator at 12 Z, 14 April 2008
northward and then northeastward to strike the coasts of West Bengal, Bangladesh and Myanmar. However, a few of them may continue their northwestward movement to strike the coast of Andhra Pradesh before recurving. A few may even cross the peninsula to emerge over the Arabian Sea to affect the west coast of India. A few cyclones may also form and develop over the Arabian Sea itself.
4.4 Advance of Summer Monsoon to the Indian Subcontinent – General Remarks The India Meteorological Department (1943) has traditionally used certain criteria associated with rainfall to determine the normal date of onset of summer monsoon at a place. Ananthakrishnan et al. (1968) who proceeded on that basis and studied the onset of SW monsoon over Kerala found that the synoptic conditions over the Arabian Sea associated with the onset were, in their own words, as follows: (i) A disturbance in the Arabian Sea/Bay of Bengal. The most common initial form of the disturbance is a trough of low pressure in southeast Arabian Sea;
4.4
Advance of Summer Monsoon to the Indian Subcontinent – General Remarks
99
Fig. 4.6 Satellite Cloud Imagery showing double cloud bands, one on each of the equator, over the Equatorial Indian Ocean, at 12 GMT, 14 April 2008
(ii) Reports from ships and island stations in the South Arabian Sea, of heavy convection, squally weather and rough seas or swell from southwest with moderate to strong winds from some southerly to westerly direction; (iii) The strengthening and deepening of lower tropospheric west winds over extreme south peninsula and Sri Lanka and strengthening of upper tropospheric easterlies to 40 Kts for a few days at 14 to 16 km; at the time of onset, the easterlies reach a maximum speed of about 60 Kts. (iv) The tendency of the strong westerlies of the upper troposphere over north India to break up or shift northwards. (v) Persistent moderate to heavy clouding in the south Arabian Sea shown by satellite pictures and its tendency to shift northwards. In a note added to the criteria, the authors state that all these features may not always be present simultaneously. The organization of the circulation to the monsoon patterns extends over an interval ranging from a few days to 1 or 2 weeks. The above-mentioned study by Ananthakrishnan et al. (1968) is very significant in the sense that it marked a welcome re-thinking on the nature of monsoon and
100
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.7 Tracks of cyclonic storms that formed over the Bay of Bengal and the Arabian Sea in May during the 70-year period, 1891–1960 (After Rao, 1981)
its onset. It appeared to recognize for the first time that monsoon was not simply observed rainfall, nor a particular type of wind, but something beyond that, and that we must consider some aspects of the circulation that is associated with the rainfall or the wind to determine the onset. So, the problem of onset of summer monsoon over the Indian Ocean still remains to be addressed. In this regard, in addition to qualitative analysis of monsoon onset in Chap. 1 and the preceding sections, the author examined data and analyses of several meteorological variables over the Indian Ocean at surface and lower and upper-tropospheric levels at 12 GMT daily over a 5-year period (2004–2008) (January–August), available from NCEP Reanalysis. For tracing the movement of the cool, humid air of the monsoon current, stress was laid on the analysis of temperature, pressure, specific humidity, and wind fields of the different layers extending from the ocean surface to 500 hPa. The results of this examination are depicted in Fig. 4.8. For a closer look at the problem of advance, the total period of advance was divided into three stages, as follows: Stage 1. Advance over the Southwestern Indian Ocean, the Arabian Sea, and the Bay of Bengal to the Indian Subcontinent (April–June); Stage 2. Advance over the Indian Subcontinent (June–July); and Stage 3. Advance from the Indian Subcontinent to the Western Himalayas and the Tibetan Plateau (July–August)
4.4
Advance of Summer Monsoon to the Indian Subcontinent – General Remarks
101
Fig. 4.8 Schematic showing advance of summer monsoon from the Indian Ocean. The thick continuous lines with arrows show the routes of the major cross-equatorial flows. Thin continuous lines with wave structure mark the approximate monthly locations of the advancing monsoon current Troughs are indicated by thick double-dashed lines. The monsoon circulation at the beginning of advance showing locations of ITCZ and TCZ (short thick lines) over the Southwestern Indian Ocean is indicated by dotted lines with arrows. ITCZ and TCZ over India are also shown
The following aspects of the advance of monsoon are highlighted by Fig. 4.8: 1. Locations of the main heat lows (L) and their troughs (thick dashed lines) in all the land sectors bordering the Southwestern Indian Ocean, the Arabian Sea, and the Bay of Bengal; 2. The (J-F) location of the equatorial trough of low pressure with cyclonic circulation (dotted lines with arrows) around it over the Southern Indian Ocean with locations of ITCZ and TCZ, before it starts its seasonal movement; 3. Three major monsoon currents (thick bold continuous lines with arrow) transporting cool, humid air of the winter hemisphere to the summer hemisphere, each being a divergent current from a heat sink or a high pressure area. These currents converge into the circulations around low pressure areas on either side, one being the heat low over the neighboring land and the other over the oceanic equatorial heat source (Note that the Continents of Africa, Australia and the Maritime Continent play important roles in this regard). A major cross-equatorial flow of monsoon current is influenced by the trough of the heat low over Peninsular India. It appears to cross the equator in the longitudes of Sri Lanka and flow along the western boundary of the Bay of Bengal close to the east coast of the peninsula;
102
4 Monsoon over Southern Asia and Adjoining Indian Ocean
4. Approximate northern boundary of the advancing monsoon current (thin continuous lines with wave structure) in the northern hemisphere at certain epochs of time (months/dates); in the southern hemisphere, the lines refer to the southern boundary of the retreating monsoon; 5. Advance of the monsoon current along the west coast of Peninsular India as well as the Arakan coast of Myanmar appear to be facilitated by movement of a cold sector of the monsoon wave over the respective region; 6. Locations of the ITCZ and the TCZ (indicated by short thick lines) on either side of the equatorial heat source after its final arrival over the northwestern part of the Indian Subcontinent.
4.4.1 Advance over the Indian Ocean (April–June) – Stage 1 4.4.1.1 Retreat from the Southern Hemisphere The successive monthly locations of the southern boundary of the monsoon wave during the period, January–April, shown in Fig. 4.8, illustrate how summer monsoon withdraws from the southern hemisphere before it enters the northern hemisphere. Note how the heat low over the land sector jumps from its January location over the island of Madagascar to the African mainland and then travels along the western boundary of the Southwestern Indian Ocean to southern Somalia in April and disappears from the southern hemisphere. The equatorial troughs over the ocean also shift northward. By April they move into the northern hemisphere to start the process of advance of monsoon towards the Indian Subcontinent. Summer monsoon advances over the Northern Indian Ocean from different parts of the equator, as shown in Fig. 4.8. However, the first move appears to be made from the side of the Eastern Bay of Bengal. So, we start from that side first. 4.4.1.2 Advance over the Bay of Bengal During late March (M) or early April (A), the increased seasonal warming of the northern hemisphere and cooling of the southern hemisphere in the Southeast Asia sector and differential heating between land and ocean in general deepens a newly-formed heat low over the Malaysia-Sumatra region, which forces the North Equatorial Trough (N.E.T) of low pressure over the Bay of Bengal to move northward along with it. As the heat low crawls slowly northward along the narrow landstrips of Thailand and Myanmar, it carries the W/SW’ly monsoon airstream with it on its southeastern side. Just about this time, under the powerful influence of the heat low over the Indian Peninsula, there is a major cross-equatorial airflow from the western side of the South Equatorial Trough towards the east coast of the Indian Peninsula, a branch of which turns eastward to strengthen the circulation around the northward-moving heat low over the mountain ranges of the Tennaserim coast. Early May, there is another cross-equatorial current from the side of Australia to enter the extreme western part of the South China Sea and flow northward along
4.4
Advance of Summer Monsoon to the Indian Subcontinent – General Remarks
103
the eastern coast of the Malaysia-Thailand peninsula, a branch of which turns northward under the influence of a powerful heat low over Central Myanmar, while the other branch turns eastward to flow around a heat low over Thailand. From here, the S-ly current comes under the influences of several mountain ranges and finally turn towards the plains of northern India to flow as a ESE-ly current around the heat low over India along the foothills of the Himalayas. Needless to state, the cool, moist winds converging at the mountain slopes all along the routes produce heavy rainfall over the mountain ranges. Thus, on the Bay of Bengal side, summer monsoon advances to the Indian Subcontinent via two main routes: (i) A strong cross-equatorial airflow by the side of Sri Lanka and parallel to the east coast of the Indian Peninsula, and (ii) a cross-equatorial airflow from the side of Australia and Indonesia. 4.4.1.3 Advance over the Arabian Sea The Arabian Sea during early summer when the monsoon current crosses the equator is relatively cold compared to surrounding land areas where heat lows form, and remains under a strong temperature inversion with a strong anticyclonic circulation prevailing at low levels. The conditions inhibit further northward advance of the monsoon current. Further, part of the southern-hemispheric tradewinds that cross the equator near the Somali coast at this time appears to be diverted westward by the heat low circulation over the Congo region of Equatorial Africa, leaving only a feeble narrow coastal current to circulate around the heat low over Somalia. But, the situation changes, though gradually at first, when a warming ocean surface and deepening heat lows over land allows a further northward movement of the cool monsoon current towards the Indian Subcontinent. The land-sea thermal contrast across the Somali coast intensifies resulting in development of the so-called Somali jet and intense upwelling along the Somali coast. Similar developments also occur along the coast of the Arabian Peninsula, further north. These developments help the cold monsoon current which binds the heat low over the land to the oceanic heat low to move further northeastward and cover more than half of the Arabian Sea by end of May. So, on or around 1 June, the equatorial trough of low pressure over the northeastern Arabian Sea is so placed as to have its associated ITCZ oriented in a more or less NW-SE direction over mid-ocean and TCZ near the coast of Kerala, the southernmost Indian State, where a S/SW-ly flow is enforced partly by the movement of the cold sector of the monsoon wave and partly by the local orography of the Western Ghats Mountains. From this stage onward, monsoon current moves rapidly northward under the influence of the heat low over India-Pakistan to cover most of the northern Arabian Sea and the Indian Peninsula by June 15. The progress of the monsoon over the ocean and the land during this period is indicated in Fig. 4.9. 4.4.1.4 Weather over the Northern Indian Ocean During Advance of Monsoon Summing up the preceding paragraphs, it may be stated that the conceptual or idealized model of the advance of summer monsoon over the Indian Ocean suggested in
104
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.9 Dates of onset of summer monsoon over the Bay of Bengal and the Indian Subcontinent, as determined by India Met. Dept (Rao, 1981)
the Text appears to account qualitatively for several observed features of the weather over the Bay of Bengal and the Arabian Sea. These include cold SST anomaly, strong temperature inversion, low level jets with highly stormy seas and unusually strong ocean currents, little cloud and rain, and few cyclonic disturbances along and to the west of the cold monsoon current, as against warm SST anomaly, weak temperature inversion, light to moderate winds, heavy clouding and precipitation and a high frequency of cyclonic disturbances, such as depressions and cyclones, to the east. Both ITCZ and TCZ are characterized as zones of cloudy and rainy weather with relatively clear skies in between, confirming the wave structure of the monsoon circulation.
4.4.2 Onset over the Indian Subcontinent (June–July) – Stage 2 The India Meteorological Department (1943) worked out the normal dates of onset of summer monsoon over the Indian Subcontinent from climatological records of observed rainfall from coastal and inland stations. Figure 4.9 shows these dates by isolines over land and dashed lines over the Bay of Bengal. No isolines or dashed lines are drawn over the Arabian Sea. As stated in the preceding subsections, summer monsoon advances over the Indian Subcontinent in two broad airstreams; the Bay of Bengal stream from the southeast and the Arabian Sea stream from the SW, as shown in Fig. 4.8. In their
4.4
Advance of Summer Monsoon to the Indian Subcontinent – General Remarks
105
final position by end of June, these airstreams converge into the circulation around the heat low over northern India, forming the ITCZ on the equatorial side and the TCZ on the poleward side of the so-called monsoon trough. The progressive advance of the two streams to their final destinations over the Subcontinent is given in the map. From SE to NW, the Bay of Bengal branch of the monsoon advances to Bangladesh by 1 June, Bihar by 10 June, East Uttar Pradesh by 15 June, Punjab, Rajasthan, Pakistan and some parts of the State of Jammu and Kashmir by early July. The Arabian Sea branch approaching from the SW crosses the Indian Peninsula by 10 June, Madhya Pradesh by 15 June, and Rajasthan by 30 June. Monsoon remains established over the subcontinent till the end of August. During this period, moderate to heavy rain falls along the ITCZ to the S/SW of the monsoon trough and along the TCZ which runs along the foothills of the Himalayas to the N/NE. Rainfall appears to be deficient over the trough zone in between. However, as shown in the following subsection, the monsoon wave does not remain in this position for long, since it moves further north to Western Himalayas during late July or early August.
4.4.3 Advance to Western Himalayas (July–August) – Stage 3 After summer monsoon gets fully established over the Indian Subcontinent, an extraordinary development takes place further north, which shifts monsoon wave from the plains of northern India to the top of the Himalayan Mountain complex on account of the development of a series of heat lows to the north. It is well-known that during northern summer a heat low develops over the western part of the elevated Tibetan Plateau (e.g., Flohn, 1968; Yeh and Gao, 1979; Murakami and Ding, 1982; Luo and Yanai, 1984; Feng et al., 1984; Murakami, 1987a). Almost simultaneously, heat lows also develop to the north of the mountain complex over the extensive lowlands of Uzbekistan and Kazakhstan to the northwest, the extensive desert lands of the Sinkiang province of China to the north, and the Inner Mongolian region of Northeastern China to the northeast. All these heat lows have their warm anticyclonic circulations above them in the upper troposphere. But on a larger scale, they combine to form a powerful anticyclonic circulation centered over the Tibetan Plateau and extending from the Mediterranean Sea in the west to the Pacific Ocean in the east. It is the development of this giant anticyclonic circulation and its sudden poleward movement, which appears to draw the Monsoon and related Hadley circulations over the Indian Subcontinent within its fold. The poleward movement of the monsoon trough zone to the Himalayas in July–August causes a total re-organization of the associated Monsoon and Hadley circulation cells associated with it over the region. This movement implies a temporary bifurcation of the monsoon wave, one part remaining over the plains of Northern India with somewhat subdued activity, while the other part jumps over to the mountains to the north. The movement simply means a northward shift of the associated Monsoon and Hadley circulation cells, as shown schematically in Fig. 4.10.
106
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.10 Schematic showing mean meridional-vertical circulations and their associated rainbelts (dotted) over the Indian Subcontinent before (upper panel) and after (lower panel) development of the elevated heat source over Western Tibet
4.4.4 Source of Moisture for Monsoon Rainfall Over most parts of the globe, the main source of moisture for monsoon rainfall in the summer hemisphere is the cool, moist tradewinds of the winter hemisphere which after crossing the equator in narrow longitudinal segments converge into the ITCZ of the summer hemisphere (e.g., Simpson, 1921; Findlater, 1969a,b; Saha, 1970). While traveling over the ocean, the tradewinds pick up additional moisture from the underlying ocean surface. A limited amount of moisture is also injected into the TCZ. The cross-equatorial origin of moisture-bearing tradewinds ushering in moisture for monsoon rainfall in the summer hemisphere is evident over several other parts of the globe as well, such as Eastern Asia, Australia, Africa and South America. For details, see the respective chapter on regional monsoon in the present text.
4.5
Disturbances of the Summer Monsoon during the Onset Phase
107
4.5 Disturbances of the Summer Monsoon during the Onset Phase 4.5.1 Onset Vortex over the Arabian Sea and the Bay of Bengal During advance of the summer monsoon wave over the Arabian Sea as well as the Bay of Bengal, the atmosphere becomes dynamically unstable, often resulting in the formation of a cyclonic vortex, whenever it is disturbed by a traveling wave. Such a vortex has come to be known as an ‘Onset Vortex’. In the Arabian Sea, a few studies (e.g., Krishnamurti et al., 1981; Saha and Saha, 1993a) have shown that both barotropic and baroclinic instability may be involved in the genesis of these vortices. After formation, these disturbances usually move in a northerly direction and accelerate the advance of the monsoon current along the West Coast of India. However, a few of them may move westnorthwestward and develop into cyclonic storms over mid-ocean before they hit the coast of Oman and then fizzle out over the sandy Arabian Desert. In the Bay of Bengal, traveling E’ly as well as W’ly waves by their interaction with the quasi-stationary monsoon wave play an important role in the formation of an onset vortex. The formation of such a vortex may upset the normal schedule of advance of monsoon by either accelerating or delaying it by a few days.
4.5.2 Monsoon Depressions and Cyclonic Storms During the monsoon onset phase, June to August, a number of depressions and cyclonic storms form in the monsoon trough zone over the Bay of Bengal and the Arabian Sea. A few also form over the land area adjoining the Head Bay of Bengal. Table 4.1 gives the total number of depressions and storms that formed over these areas during an 80-year period, 1891–1970 (After Rao, 1976). The tracks of these disturbances are shown in Fig. 4.11. After formation, most of the disturbances move in a WNW direction and yield heavy precipitation over their SW quadrant. Several States in India, such as Orissa, Andhra Pradesh, Southern Bihar, Jharkhand, Chhatishgahr, and Madhya Pradesh, receive a significant proportion of their annual rainfall during the passage of these Table 4.1 Number of monsoon depressions (D) and cyclonic storms (S) during the 80-year period (1891–1970) in June, July and August over different areas Area
Bay of Bengal Arabian Sea Land area
June
July
August
D
S
D
S
D
S
71 18 12
35 15 1
107 9 39
38 3 1
132 2 42
26 2 0
108
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.11 Tracks of depressions and cyclonic storms that formed over the Bay of Bengal and the Arabian Sea in July during the 70-year period, 1891–1960 (after Rao, 1981)
disturbances. By contrast, little rain falls to the northeast of the trough axis. In the Arabian Sea sector, few lows and depressions form in this period. Those which form over the northeastern corner of the sea usually move in a northerly direction. Their contribution to annual rainfall is normally negligible. However, occasionally, midtropospheric cyclones form over the northeastern corner of the Arabian Sea and add significantly to coastal rainfall. More than 80% of all Bay disturbances during the onset phase (June–August) formed over the latitude belt, 17.5–22.5◦ N, and moved westnorthwestward.
4.5.3 Interaction of Monsoon with W’ly Waves During summer, large-amplitude W’ly waves moving across the Himalayan region, between about 30 and 50◦ N, interact with the circulation around the quasi-stationary monsoon trough over the Indian Subcontinent as well as with traveling monsoon disturbances, forming an extended trough between the two regions. On such occasions, a ridge of high pressure with anticyclonic circulations prevails over northwestern and Central India and weather remains dry over these areas (see Fig. 4.12). During the period of its eastward movement, the extended trough causes a relocation of the east-west oriented monsoon trough and its associated rainbelt which now lies across southern India. Thus, the interaction simply causes a redistribution of monsoon rainfall with two belts of heavy rain, one in the mountains in the north and other over Southern India in the south and a wide area of little or no rain in between over Central India. As the W’ly wave moves further eastward across the Himalayas taking the extended trough along with it, the belt of heavy rain also shifts eastward. A period of 3–5 days may be taken for the rainbelt to move across the mountains from west to east.
4.5
Disturbances of the Summer Monsoon during the Onset Phase
109
Fig. 4.12 Schematic showing interaction of a W’ly wave trough with: (a) the monsoon trough over the Indian Subcontinent, and (b) a Depression (D) over the Bay of Bengal. Heavy rainfall areas are hatched. Troughs are double-dashed
The W’ly wave troughs moving across the eastern Himalayas also interact with westward-propagating depressions over the Bay of Bengal forming an extended trough with it. In this case also, we have two areas of heavy rainfall, one in the north over the eastern Himalayas and the other over an extensive area of central and southern India. In fact, the ‘break monsoon’ situation over Central India almost disappears in this case. The windward slopes of the Western Ghats Mountain as well as the Arakan Yoma experience heavy rainfall on these occasions.
110
4 Monsoon over Southern Asia and Adjoining Indian Ocean
4.6 Rainfall over the Indian Subcontinent during the Onset Phase Several factors appear to contribute to the distribution of monsoon rain over the Indian Subcontinent during the onset phase. They include: (1) Orography, (2) Heat Lows, (3) Location of the monsoon trough, (4) Depressions and cyclones, and (5) Travelling wave disturbances. More than any other factor, orography appears to contribute the most to the distribution of monsoon rains over the Indian Subcontinent, as would be evident from Fig. 4.13, which records the highest concentrations of rainfall on the windward slopes of the mountains, and scanty rainfall on the leeside. According to Fig. 4.13, the Western Ghats Mountains of the Indian Peninsula, the Arakan Yoma and the Shan States of Myanmar, and the Mountains of Eastern Himalayas, especially in Bangladesh, Assam and Arunachal Pradesh experience heavy rainfall, while the leesides of these mountains have deficiency of rainfall. Along the foothills of the Himalaya Mountains, there appears to be a general decrease of rainfall from east to west. A rainfall maximum appears over Orissa and adjoining Central India to the southwest of the monsoon trough zone. While the eastern and southern parts of the Indian Subcontinent receive substantial rainfall during the onset phase of the summer monsoon, the northwestern part especially the Thar Desert area of Pakistan and India suffers from shortage of rain because of the presence of the heat low circulation over the region.
Fig. 4.13 Ten-year (1976–1985) mean July rainfall (unit: 10–5 kg m–2 s–1 ) over the Indian Subcontinent. The double-dashed line shows the location of the monsoon trough (after Saha and Saha, 1996)
4.7
Summer Monsoon – Withdrawal Phase (September–November)
111
The distribution of rainfall shown in Fig. 4.13 agrees substantially with that of normal summer monsoon rainfall during the period, June–September, as determined by the India Meteorological Department (not shown).
4.7 Summer Monsoon – Withdrawal Phase (September–November) 4.7.1 Dates of Withdrawal of Monsoon By and large, summer monsoon withdraws from the Indian Subcontinent and the Northern Indian Ocean by following the same route as it did during advance (Fig. 4.9), but in the reverse direction. After reaching its peak intensity during July–August, monsoon starts withdrawing from the Western Himalayas in early September. The process begins with the filling up of the heat low over the elevated Tibetan Plateau and the re-establishment of the heat low over the northwestern part of the Indian Subcontinent. However, the transition takes place very gradually and almost imperceptibly for a while. The withdrawal from northwestern India is marked by a weakening of the heat low and associated monsoon trough, reversal of the low-level wind from southerly to northerly and a decrease of rainfall. The change ushers in a regime of somewhat cooler and drier air from the north. Figure 4.14 gives isolines of the dates of withdrawal of the summer monsoon from India, as worked out by the India Meteorological Department on the basis of the climatological distribution of rainfall. According to Fig. 4.14, the SW monsoon pulls out of nearly half the subcontinent by 1 October when its northern boundary runs from the hills of Uttar Pradesh to the middle of the West coast of India. By 15th October, it moves further southeastward so as to have its western end over the middle of the Indian peninsula and the eastern end over central Myanmar or even further south. From this stage onward, its southward movement over land and ocean is very slow, and it is not until the end of November that monsoon totally withdraws from the Indian peninsula and reaches the latitude of Sri Lanka. The Myanmar branch of the monsoon continues to move south/southeastward to reach eventually its winter location as NET within about 5◦ of the equator in the eastern Indian Ocean.
4.7.2 Retreating Monsoon Rain over Tamil Nadu During the stage of withdrawal of summer monsoon from India, the area along the east coast of the Indian Peninsula, especially the state of Tamil Nadu, exposed to the moisture-laden ENE-ly winds at low levels, experiences retreating monsoon rains. In literature, this rain is often projected as winter monsoon rain. However, in reality, it is summer monsoon rain during its withdrawal phase. The mechanism of this rain is illustrated schematically in Fig. 4.15.
112
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.14 Normal dates of withdrawal of the SW monsoon from the Indian subcontinent (after Rao, 1981)
Fig. 4.15 Illustrating the mechanism for formation of a low level moisture convergence zone and rain in Tamil Nadu during monsoon withdrawal phase
4.7
Summer Monsoon – Withdrawal Phase (September–November)
113
Following the withdrawal of the monsoon trough from northern India, two high pressure cells with anticyclonic circulations develop over land, one over Central India and the other over Myanmar. Between their anticyclonic circulations, a trough of low pressure forms over the North Bay of Bengal (see Fig. 4.15). Now, the wind diverging from the high pressure cell over central India has a limited sea travel before it approaches the coast of Andhra Pradesh and Tamil Nadu from N/NE-ly direction and its moisture content is low. On the other hand, the wind diverging from the high pressure cell over Myanmar has first to flow northwestward over North Bay and then turn cyclonically around the oceanic trough southwestward towards the Indian peninsula. Thus, it has a long sea travel and is fully saturated with moisture by the time it arrives at the Tamil Nadu coast where it converges into the circulation around the high pressure cell over Central India. The resulting moisture convergence along the Andhra Pradesh-Tamil Nadu coast (indicated by a thick continuous line) along with the prevailing upper-air divergence over the area is responsible for producing the rainfall over Tamil Nadu. Heaviest rainfall occurs at the mountains of Tamil Nadu facing the moist winds.
4.7.3 Disturbances of the Withdrawal Phase 4.7.3.1 Western Disturbances Closely following the southward movement of the equatorial trough, the belt of the subtropical westerlies shifts southward and W’ly waves (WDs) follow a more southerly track. Their influence on weather over the northern part of the Indian Subcontinent, especially western Himalayas, increases. They also influence the track of cyclonic disturbances which move up from the Bay of Bengal and the Arabian Sea towards the mountain.
4.7.3.2 Depressions and Cyclonic Storms During the monsoon withdrawal phase, there is a spurt in cyclonic activity over the Bay of Bengal and the Arabian Sea. A larger percentage of the depressions develop into cyclonic storms and their places of origin shift continually equatorward from September to November, as shown by statistics over a 70-year period, 1891–1970 (Rao, 1981). It is estimated that in November, 90% of the depressions and cyclones formed over a wide area bounded by latitudes 7.5 and 15◦ N and longitudes 77.5 and 100◦ E. This continued southward movement and the widening of the area of cyclonic activity would be evident from Fig. 4.16 which shows the tracks of these disturbances, when it is compared with Fig. 4.11 for July. These changes appear to be characteristic features of the monsoon withdrawal phase. During the same 70-year period, the number of depressions and cyclonic storms that formed in the Arabian Sea was 21 and 15 respectively. About 50% of them formed over the area bounded by latitudes, 10–12.5◦ N and longitudes, 62.5–75◦ E.
114
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.16 Tracks of cyclonic storms that formed over the Bay of Bengal and the Arabian Sea in October during a 70-year period, 1891–1960 (after Rao, 1981)
In the Bay of Bengal, most of the depressions and cyclones after initial travel over the sea in a WNW direction recurved to a N/NE-ly direction after reaching the latitude belt, 15–20◦ N. Those amongst these which turned into severe cyclonic storms and entered land devastated many coastal belts. Low-lying river deltas which are particularly vulnerable in this regard suffered enormous losses due to high winds, storm surges and torrential precipitation. Table 4.2 lists a few of the deadliest cyclones on record which took heavy toll of lives through storm surge drowning during the monsoon withdrawal phase.
Table 4.2 Statistics of some killer tropical cyclones with record storm surge Cyclone
Date
Surge (m)
Death-toll
Bangladesh (Buckergunge) Calcutta (Kolkata) Bangladesh Andhra Pradesh
November 1, 1876 October 5, 1864 November 13, 1971 November 17, 1977
9.0–12.0 12.0 7.0 5.0
100,000 50,000 300,000 9000
4.8
Variability of the Indian Summer Monsoon Rainfall
115
4.8 Variability of the Indian Summer Monsoon Rainfall Observations show that the summer monsoon rainfall over India for the period, June–September, is not steady but varies on time scales ranging from a few days to a year or several years. Since large-scale variability often leads to disastrous floods and droughts which cause miseries to people and affect the economy of the country, the problem of variability has been studied extensively ever since the time of Blanford (1884, 1886) who first initiated the study. He was followed by Walker (1910a, b, 1914) and several scientists in India and abroad (e.g., Mooley, 1975, 1976; Hahn and Shukla, 1976; Kanamitsu and Krishnamurti, 1978; Bhalme and Mooley, 1980; Angell, 1981; Shukla and Paolino, 1983; Mooley and Parthasarathy, 1983, 1984; Rasmusseen and Carpenter, 1983; Parthasarathy, 1984). To date, there has been a large body of literature on the subject of variability of Indian summer monsoon rainfall. An excellent review of some of the recent studies has been provided by Mooley and Shukla (1987) as well as by Krishnamurti and Surgi (1987) and readers interested in details of these various studies may refer to the original papers mentioned in the references.
4.8.1 Interannual Variability For studying the rainfall variability, different workers used different sets of data. Mooley and Parthasarathy (loc. cit.) used data from a network of rain gauge stations which were fixed and evenly distributed over the country (one raingauge station per district) covering a period of 108 years from 1871 to 1978 (this was later extended to 114 years from 1871 to 1984) but by excluding the hilly stations (where rainfall depended upon height and was of a different pattern from rest of the stations) from the existing network. The season of rainfall considered in these studies is from June to September. On statistical tests, they found the series of rainfall data used by them to be homogenous. Mooley and Parthasarathy (loc. cit.) used the following statistical criteria for describing the various parameters of the variability.
4.8.1.1 Statistical Criteria
Mean or average (x) =
Standard deviation (SD) σx =
i=n i=1
i=n
xi /n
i=1
(4.8.1) 1/2
(xi − x) /(n − 1) 2
(4.8.2)
116
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Co-efficient of variation (CV) = Standard deviation × 100/Mean = (σx /x) × 100 (4.8.3)
Inerannual variability =
i=n−1
[xi+1 − xi ]/(n − 1)
(4.8.4)
i=1
where xi denotes the monsoon season rainfall for the ith year, x (x underlined) is the average rainfall of the total number of years n, and σ x is the standard deviation of the monsoon rainfall. Mooley and Parthasarathy describe the monsoon rainfall in terms of a standard unit, which is equal to deviation from mean divided by the standard deviation. Figure 4.17 shows the all-India summer monsoon rainfall in standard units for each year of the period, 1871–1984. According to Mooley and Parthasarathy, a small deviation up to 5% on either side of the long-term average value can be considered as normal or average rainfall. Some of the statistical properties of the all-India summer monsoon rainfall and its long-term variability as given by Mooley and Parthasarathy (loc. cit.) are given in Table 4.3.
Fig. 4.17 All-India summer monsoon rainfall in standard units (deviation from normal divided by standard deviation) for each year during the period, 1871–1984 (after Mooley and Parthasarathy, 1984)
4.8
Variability of the Indian Summer Monsoon Rainfall
117
Table 4.3 Statistical properties of the all-India summer monsoon rainfall, 1871–1984 Property
Value
Property
Value
Mean Mean/annual Median Lower quartile Upper quartile
852 mm 78.1 864 mm 800 mm 908 mm
Highest Rainfall Deviation from mean Range Coefficient of variation Mean interannual variation as % of mean
1017 mm 19% 413 mm 9.7% 11.9
Standard deviation Lowest rainfall (1877) Mean interannual variation in terms of standard deviation Deviation from mean
83 mm 604 mm 1.22
–29%
4.8.1.2 Floods and Droughts As should be expected, large excess (deficit) of rainfall in a year leads to large-scale floods (droughts) in India. Mooley and Parthasarathy (1983) from the results of their study lay down the following criteria for occurrence of these events: Flood if [(xi − x)/σx ] > 1.28, Drought if [(xi − x)/σx ] < 1.28 Using the above criteria, they worked out the years of large-scale droughts and floods in India during the period, 1871–1984, and identified the following years as those of worst flood and drought years in India: Floods Droughts
1892, 1917, 1956, and 1961 1877, 1899, 1918, and 1972
4.8.2 Factors Likely Responsible for Interannual Variability Since the time of Blanford (1886), there has been frantic search on a global scale for parameters which likely influence the interannual variability of the Indian summer monsoon rainfall. Studies have pointed at several factors around the globe and even at extra-terrestrial ones, with varying degrees of correlation, but the relationship with most cases have turned out to be fragile and undependable. Only a few have passed the rigorous tests of confidence. To date, the following factors appear to be promising in this regard: (a) (b) (c) (d)
Dates of onset and withdrawal of summer monsoon; Eurasian snow cover; Sea surface temperature, El Nino and Southern oscillation; Soil moisture, vegetation and albedo of the earth’s surface
118
4 Monsoon over Southern Asia and Adjoining Indian Ocean
Fig. 4.18 Dates of onset of summer monsoon over southern Kerala, 1901–1984, showing the mean date and the limits of one standard deviation on either side of the mean (Mooley and Shukla, loc. cit.)
(a) Variability in Dates of Onset and Withdrawal of Monsoon: During a period of 84 years, 1901–1984, there was a large variability in the dates of onset of summer monsoon in Kerala, the southernmost state of the Indian peninsula, as may seen from Fig. 4.18 (Mooley and Shukla, loc. cit.). According to Fig. 4.18, the long-term mean date of onset of summer monsoon over southern Kerala is 2 June with a standard deviation of 8 days. Most of the dates of onset lie within two standard deviations. The extreme dates are 11 May 1918 and 18 June 1972. However, as Mooley and Shukla remark, ‘it is strange that both these extremes occurred in drought years. While a late onset in 1972 may be consistent with a drought in that year, the same cannot be said about an early arrival in 1918. This inconsistency only highlights the nature of the problem and the fact that an early or late arrival or departure of monsoon alone is not uniquely related to the overall behaviour of the monsoon during a season. Other factors may be involved in determining the observed variability of the rainfall.’ (b) Eurasian Snow Cover: Blanford (1884) was, perhaps, one of the first to point out that excessive snowfall in the Himalayas during the winter and spring was prejudicial to the subsequent monsoon rainfall over India. His observation was substantiated by Walker in 1910 and the inverse relationship was made use of in monsoon rainfall forecasts. However, on account of uncertainty in observations of snow cover, the use of this parameter was discontinued after 1950. The advent of earth-orbiting satellites which started observing the earth’s snow cover changed the situation. Wiesnet and Matson (1976) on the basis of snow cover data furnished by satellites commented that ‘the December snow cover for the northern hemisphere was a very good predictor of the following January–March snow cover’. Subsequent studies (e.g., Hahn and Shukla, 1976; Dickson, 1983, 1984) found an apparent inverse relationship of the Himalayan snow cover with the Indian summer monsoon rainfall.
4.8
Variability of the Indian Summer Monsoon Rainfall
119
In regard to the suspected inverse relationship between the two parameters, Mooley and Shukla (1987) observes: ‘An inverse relationship between the Eurasian snow cover and summer monsoon rainfall is understandable, since large and persistent Eurasian snow cover would substantially reduce the rate of heating of the concerned land masses during spring and summer and thus delay the onset of the monsoon and prevent normal monsoon activity. It may be mentioned that the snow cover area as measured by the satellite is not quite a representative parameter to assess the amount of snow. The snow cover area in two cases may be the same but depths may be quite different, and the impact on monsoon would be quite different in the two cases’. It may be hoped that with rapid development of satellite technology which would enable both horizontal and vertical extent of snow cover to be measured, a reliable correlation will be found between Eurasian snow cover and the Indian summer monsoon rainfall. (c) Sea Surface Temperature, El Nino and Southern Oscillation: Need for an intensive study of the variation of sea surface temperature (SST) in the context of the Indian summer monsoon rainfall came to the fore after Bjerknes (1969) demonstrated the linkage between the ocean and the atmosphere and interpreted the Southern Oscillation and the Walker circulation in terms of air-sea interaction and El Nino and La Nina events in the equatorial Pacific Ocean. He found that in general a positive SST anomaly in the equatorial western Pacific was associated with heavy rainfall and a negative anomaly with deficient rainfall over India. In a La Nina year, the equatorial eastern Pacific ocean is cold (SST anomaly negative), but the equatorial western Pacific including a part of the eastern Indian ocean, lying between about 70 and 160◦ E, is warm with a positive SST anomaly. With such a distribution of SST anomaly, the Southern oscillation index is positive and the Walker circulation has its ascending branch over the equatorial western Pacific and the descending branch over the equatorial eastern Pacific. The result is normal or heavy rainfall over the western Pacific including the India-Australia sector and drought condition over the eastern Pacific. The situation reverses in an El Nino year when the warm anomaly in the SST shifts to the equatorial eastern Pacific Ocean with a cold anomaly over the western Pacific. In such a year, the rainbelt shifts to the eastern side of the ocean and the India-Australia sector experiences drought conditions with abnormally deficient rainfall. Following the investigations of Bjerknes (loc. cit.), there was a spurt in studies of Indian summer monsoon rainfall in relation to the zonal anomaly of SST in the Pacific Ocean. Several studies (e.g., Angell, 1981; Mooley and Parthasarathy, 1983, 1984; Sikka, 1980a; Rasmussen and Carpenter, 1983) were undertaken to find the degree of relationship of the Indian summer monsoon rainfall with the southern oscillation and the El Nino events by working out the correlation between them. Each of these studies used data for different years and often different parameters. Angell (1981) who used data for the period 1868–1977 obtained a correlation coefficient around –0.6 between all-India monsoon rainfall and SST anomaly over the equatorial eastern Pacific ocean (0–10◦ S, 90–180◦ W) one to two seasons later. The relationship is highly significant. Mooley and Parthasarathy (1983) who used
120
4 Monsoon over Southern Asia and Adjoining Indian Ocean
a dataset for the years, 1871–1978, found a significant relationship between the monsoon rainfall over India and the El Nino events in the equatorial Pacific. They considered 22 moderate and severe El Nino events and found that in all the severe El Nino years, the all-India monsoon rainfall in standard units was less than –0.60, with the exception of 1884 when it was +0.92. However, on careful examination, they found that in that particular year, 20.4% of the country had, indeed, experienced drought conditions as would be expected in an El Nino year, but the deficiency was offset by heavy rainfall in the remaining parts of the country caused by other more influential factors. In another investigation (Mooley and Parthasarathy, 1984) using the same dataset examined the relationship between the Indian monsoon rainfall and the SST anomaly in the Pacific in 3-monthly periods preceding or following the monsoon season and found inverse relationships that were significant at the 1% level for each of the JJA, SON and DJF seasons, and at the 5% level for the MAM season. They found the relationship to be consistent and stable. Sikka (1980a) used a different set of data. He used the Line Islands precipitation data as indicator of El Nino events in the Pacific and sought to relate them with the monsoon rainfall over India. He found a general association of El Nino events with deficient rainfall over India. Rasmusson and Carpenter (loc. cit.) found that in 25 El Nino years, the Indian monsoon rainfall was below the median rainfall in 21 years and below the mean in 19 years. They thought that the association had some predictive value. Shukla and Paolino (1983) examined the relationship between the Indian monsoon rainfall and composites of normalized Darwin pressure anomalies (3-month running mean) for heavy monsoon rainfall years and deficient monsoon rainfall years during the period, 1901–1981, and found that the tendency of Darwin pressure anomaly before the monsoon season was a good indicator of the subsequent monsoon rainfall over India. According to them, a negative tendency between the DJF (December–February) season and the MAM season was found to be associated with good monsoon rainfall years and a positive tendency with poor monsoon rainfall years. The results of their investigation are shown in Fig. 4.19. They expressed the view that whenever the tendency showed a large negative value, non-occurrence of drought in India could be predicted with a very high degree of confidence. (d) Soil Moisture, Vegetation and Albedo of the Earth’s Surface: There appears to be a symbiotic relationship between the albedo of the earth’s surface on one hand and soil moisture, vegetation and rainfall on the other. The atmosphere over a region with high albedo tends to make up for the loss of solar radiation by large-scale subsidence which inhibits precipitation and perpetuates conditions which lead to high albedo. This is likely to happen particularly over the dry desert regions of the subtropics and the marginal lands where an increase of albedo caused by overgrazing or deforestation can lead to long-term reduction of precipitation, and continuation of dry desert conditions. On the other hand, a lowering of albedo by afforestation and vegetation can help to reduce subsidence and thereby promote more convection and precipitation to occur over the region. Afforestation and vegetation thus allowing long-term increased precipitation helps to increase soil moisture which in turn
4.8
Variability of the Indian Summer Monsoon Rainfall
121
Fig. 4.19 Composite of the normalized Darwin pressure anomaly (3-month running mean) for years of heavy (good) monsoon rainfall and deficient (poor) monsoon rainfall (after Shukla and Paolino, 1983)
reduces albedo and increase rainfall. The validity of these cyclic processes has been well demonstrated by a series of numerical experiments using GCMs by Charney (1975); Charney et al., (1977); Shukla and Mintz (1982); Sud and Smith (1985) and several others. Besides the global and regional factors mentioned above, attempt has been made to link the variability of the Indian summer monsoon rainfall with such extraterrestrial factors as solar activity as judged by the sunspot numbers. Some of the original studies in this direction were carried out by Gilbert Walker (1915a,b,c). However, the results of his studies showed no significant or consistent relationships with the mean annual rainfall over India.
4.8.3 Intraseasonal Variability The variability of monsoon rainfall on time scales ranging from a few days to several weeks are caused by the familiar westerly and the easterly waves that move across the Indian longitudes and interact with the quasi-stationary monsoon wave, and also by low-frequency intraseasonal oscillations of the atmosphere during northern summer. 4.8.3.1 Variability on Scale of 3–7 Days – Active and Break Monsoons IActive and break monsoon cycles occur frequently during the monsoon season. They have drawn the attention of meteorologists over a long time. Defining a
122
4 Monsoon over Southern Asia and Adjoining Indian Ocean Table 4.4 Statistics of break in summer monsoons, 1888–1967
Month
No. of breaks
No. of break (days)
July August 1 July–31 August
53 55 5
306 356 47
Most frequent Average duration Longest break duration (days) (days) (days) 5.8 6.5
17 20 21
4 3
monsoon break as a disappearance of the monsoon trough over India from mean sea level and 850 hPa maps for at least a couple of days at a stretch, Ramamurty (1969) catalogued the breaks in July and August from 1888 to 1967. His statistics are presented in Table 4.4. Ramamurty found from his long-period statistics that August is slightly more susceptible to ‘break days’ and longer breaks, particularly around the middle of the month. He also found that in the long record, there was no break in 12 years. Another important effect of the W’ly wave trough on the monsoon is a north/northeastward recurvature of the track of a monsoon depression or cyclone from its usual west/northwestward track if it happens to come under the influence of the W’ly trough. 4.8.3.2 Variability on 30–50 Day Time Scale Atmospheric oscillations on this time-scale are known as Madden-Julian Oscillations (MJOs) after the name of their discoverers (Madden and Julian, 1971, 1972) and are believed to be excited by large-scale tropical convection. All subsequent studies (e.g., Murakami, 1976; Yasunari, 1979; Sikka and Gadgil, 1980; Murakami, 1984) have suggested a strong relationship of these oscillations with the slow meridional movement of a zonally-oriented low-level trough of low pressure associated with penetrative convection, cloudiness and heavy rainfall from the equatorial region to higher tropical latitudes in the Indian monsoon region Krishnamurti and Subrahmanyyam (1982) found that active and break monsoon cycles over the Indian longitudes were closely coupled to the meridional movement of these low-level troughs associated with rainfall. Oscillations on this time scale over the equatorial latitudes which moved slowly towards the pole have also been noted in the relative angular momentum of the earth’s atmosphere from the datasets of the zonal wind by Rosen and Salstein (1981). Interestingly, they also found a strong signal on this time scale in the variations of the length of the day computed from lunar laser ranging observations.
Chapter 5
Monsoon over Eastern Asia (Including China, Japan, and Korea) and Adjoining Western Pacific Ocean
5.1 Introduction The study of monsoon and related weather phenomena over Eastern Asia has a long history. Prior to the 3rd century B . C ., it was mostly the farmers who watched the weather seriously and maintained some kind of an ‘agricultural calendar’ of climatic events in connection with agricultural operations. In some central parts of China, these agricultural calendars are still in vogue, though other parts have opted for more modern methods. The modern instrumental period may be said to have begun about the close of the 19th century, but the observing network was very limited in the beginning and confined mostly to densely populated areas. Vast areas were uncharted. It is only recently from about the middle of the twentieth century that the observational network over the region as a whole has improved. Since 1959, a network of surface and upper-air observing stations was established on the highly elevated plateau of Tibet. It is mentioned that during the period, 1949–1963, the number of meteorological observing stations in China increased 30-fold (Cheng, 1963). A Chinese national project on the meteorology of the Tibetan Plateau was reported upon by Yeh and Gao (1979) who along with their many colleagues carried out excellent studies of the heat budget of the plateau and other related problems of the high-altitude region. The observational network on the plateau was further improved upon during a special experiment known as the Qinghai-Xizang Plateau Meteorology Experiment (QXPMEX) during the summer of 1979 which was conducted by Chinese scientists as part of the Global Weather Experiment, 1978–1979. Earlier, during the winters of 1974 and 1975, a GARP field project under the leadership mostly of Japanese scientists had conducted an ‘Air Mass Transformation Experiment’ (AMTEX) over the sea areas southwest of Japan to learn more about the energy and momentum exchanges between the sea and its overlying atmosphere and meso-scale cellular convection and cyclogenesis that occurs when there is a cold air outbreak over the East China Sea and the Kuroshio current. Like the Indian Ocean, the South China Sea and the Western Pacific Ocean play important roles in monsoon circulation over Eastern Asia, the Maritime Continent and the Australian region, especially during advance and retreat of monsoon current across these ocean areas. In order to learn more about these K. Saha, Tropical Circulation Systems and Monsoons, C Springer-Verlag Berlin Heidelberg 2010 DOI 10.1007/978-3-642-03373-5_5,
123
124
5
Monsoon over Eastern Asia and Adjoining Western Pacific Ocean
roles, the International Community led by Chinese scientists mounted an impressive array of field experiment called the South China Sea Monsoon Experiment (SCSMEX) in 1998, spanning the period from 1 May to 31 August, for carrying out intensive observations of surface and upper air parameters relating to monsoons. A wealth of new information was collected from this experiment, which became available for further studies (e.g., Lau et al., 1998, 2000). The recent studies of monsoon over Eastern Asia are, therefore, based on an excellent coverage of data, though in some areas long-period data are still lacking. An excellent review of some of the recent studies which were carried out on the seasonal march of the East Asian Summer Monsoon has been provided by Ding (2004). It is well-known that during the peak summer months of July-August, monsoon over the tropical belt of Eastern Asia suddenly jumps to extratropical latitudes to cover such areas as Northern China, Northern Japan, Korea and Eastern Siberia with its poleward boundary near about 60◦ N. We study monsoon over this extratropical belt of Eastern Asia in the latter part of this chapter.
5.2 Physical Features and Climate It is not easy to delineate the southern boundary of Eastern Asia which includes some of the most heterogeneous elements of the global terrain and features practically the whole range of global climate from tropical to arctic. Here, along the southwestern boundary of mainland China, the mighty Himalaya mountain complex with the world’s highest mountain peak, Mount Everest, rising to an altitude of about 8.85 km a.s.l. and associated Tibetan plateau with a mean elevation of well over 4.6 km a.s.l. stand guard over the extensive lowlands of Northern and Eastern China which have several smaller high ground or hill ranges scattered all over the region. The Tibetan Plateau descends steeply both northward and eastward to the plains of China and this is clearly indicated by the direction of flow of water of the two mighty rivers, the Yantzekiang and the Hwang-Ho which flow in a zig-zag course eastward to the China Sea. Also, along the northwestern boundary of China lie a series of high-rise mountains, the Tien Shan and the Altay mountains, and several other lesser mountain ranges which extend northeastward to as far north as 60◦ N or even beyond. A relief map of Eastern Asia showing the above-mentioned topographic features is at Fig. 5.1. Another important physical feature of China which exercises great influence upon the climate of the region is an extension of the vast Central Asian desert lowlands from Sinkiang in the west to the Gobi desert or even beyond to Manchuria and Eastern Siberia in the east. The Korean peninsula lies over the northeastern part of the region and juts out southward so as to have the Yellow Sea to its west and the Sea of Japan to the east. The Korean Strait separates the peninsula from the Islands of Japan which lie to the south and east. Besides the mainland, several large and small islands belonging to China and Japan lie scattered over the western North Pacific Ocean.
5.3
The Winter Season over Eastern Asia (November–March)
125
Fig. 5.1 Relief map of Eastern Asia
The topography and the geographical location of Eastern Asia are responsible for a wide variety of climatic conditions in terms of temperature, pressure, airflow and rainfall. The seasons also are somewhat different here from those over the Indian Subcontinent. On account of more northerly location and greater continental and oceanic influences, the winter season starts early in November and lasts till the end of March. Summer monsoon starts in May and lasts till the end of September. The transition periods are usually April and October.
5.3 The Winter Season over Eastern Asia (November–March) 5.3.1 Temperature, Pressure, and Wind Mean air temperatures over Eastern Asia start falling rapidly from October onward and by January extremely low temperatures often dipping to a minimum of < –30◦ C may prevail over the Gobi desert of Outer Mongolia and adjoining eastern Siberia as well as over the Korean peninsula (Fig. 5.2). In response to the temperature distribution, an extremely high pressure cell builds up over the region with maximum pressure exceeding 1032 hPa centered over the
126
5
Monsoon over Eastern Asia and Adjoining Western Pacific Ocean
Fig. 5.2 Mean air temperatures (◦ C) over Eastern Asia in January (after Watts, 1969)
Mongolian region and a steep pressure gradient to the south and east to cover practically the whole of Asia and a good part of northwestern Pacific Ocean close to the coast of Eastern Asia. Side by side, a deep low pressure cell develops in the vicinity of the Aleutian Islands area with the subtropical high pressure cell of the Pacific Ocean lying to its south with ridge along about 25◦ N. Consistent with the prevailing temperature and pressure distributions described in the preceding para, there is strong anticyclonic circulation over a vast region of Eastern Asia and adjoining Pacific Ocean where it appears to merge with the subtropical anticyclonic circulation with its axis along about 30◦ N. A strong cyclonic circulation prevails over the Aleutian Islands area. However, the circulations change rapidly with height, with westerlies dominating the flow at 500 and 200 hPa over the subtropical and midlatitude belts. These aspects of the airflow at low levels and at 500 and 200 hPa over Eastern Asia and adjoining Pacific Ocean during January are shown in Fig. 5.3 (Crutcher and Meserve, 1970). Several studies (e.g., Yeh and Gao, 1979; Murakami, 1981a,b; Boyle and Chen, 1987) have emphasized the great mechanical and thermodynamical influence of the Himalayan Massif and Tibetan Plateau on the upper air circulation over Asia, especially Central and Eastern Asia where the winds are predominantly westerly above
5.3
The Winter Season over Eastern Asia (November–March)
127
Fig. 5.3 Streamlines showing mean atmospheric circulation over Eastern Asia and adjoining western Pacific Ocean during winter: (a) Low-level (925 mb), (b) 500 mb, and (c) 200 mb. Thick continuous line in (a) shows the NE–SW oriented convergence line over the Western North Pacific Ocean
the low-level E/NE-ly tradewinds. During winter, upper-level midlatitude westerlies migrate southward and blow around the Himalayan mountain complex and the Tibetan plateau. On striking the western side of the mountain barrier, the flow appears to divide itself into two parts, one flowing northward around the northern boundary of the mountain block and the other flowing southeastward around the southern boundary. The divided aircurrents appear to merge on the leeside over China, some distance away from the eastern side of the mountains. An interesting aspect of the midtropospheric circulation over the region is that it is the weakest (