Global Environment: Water, Air and Geochemical Cycles [2 ed.] 9780691136783, 2011039852

This newly revised edition of Global Environment discusses the major elements of the geochemical cycles and global fluxe

171 30 17MB

English Pages 701 Year 2012

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Title
Copyright
Contents
Preface to the Second Edition
1. Introduction to the Global Environment: The Water and Energy Cycles and Atmospheric and Oceanic Circulation
Introduction
The Global Water Cycle
Major Water Masses
Fluxes between Reservoirs
Geographic Variations in Precipitation and Evaporation
The Energy Cycle
Introduction
Radiation and Energy Balance
Variations in Solar Radiation: The Atmospheric and Oceanic Heat Engine
Circulation of the Atmosphere
Oceanic Circulation
Introduction
Wind-Driven (Shallow) Circulation
Coastal Upwelling
Thermohaline (Deep) Circulation
2. Air Chemistry: The Greenhouse Effect and the Ozone Hole
Atmospheric Gases
Carbon Dioxide
Present and Future CO2 and the Surficial Carbon Cycle
Past CO2 Levels
Other Greenhouse Gases: Methane, Nitrous Oxide
Other Greenhouse Gases: Halogens and Tropospheric Ozone
Radiative Forcing by Anthropogenic Factors
Climatic Effects of Radiative Forcing: Climate Sensitivity, Global Warming, and Hydrologic Changes
Observed Changes in Temperature and Atmospheric Circulation
Observed Changes in the Water Cycle: Water Vapor, Precipitation, Streamflow, and Storms
Observed Changes in Ice, Sea Level and the Oceans
Predictions for Future Climate Change
Aerosols
Aerosol Cloud Effects
Types of Aerosols
Gaseous Emissions
Sulfate Aerosols
Black Carbon Aerosols
Organic Carbon Aerosols
Biomass Burning Aerosols
Nitrate Aerosols
Mineral Dust Aerosols
Sea-Salt Aerosols
Surface Dimming by Aerosols
Aerosols and the Hydrologic Cycle
Black Carbon Aerosols and Snow Cover
Ozone and the Ozone Hole
Stratospheric Ozone: The Ozone Hole
Tropospheric Ozone: Air Pollution
3. Air Chemistry: Rainwater, Acid Rain, and the Atmospheric Cycles of Sulfur and Nitrogen
Introduction
Formation of Rain (and Snow)
Water Vapor in the Atmosphere
Condensation
Sublimation
Rain (and Snow) Formation
Air Motion in Cloud Formation
Chemical Composition of Rainwater: General Characteristics
Cl−, Na+, Mg++, Ca++ and K+ in Rain
Gases and Rain
Sulfate in Rain: The Atmospheric Sulfur Cycle
Sea-Salt Sulfate
U.S. Sulfur Emissions
Conversion of Sulfur Dioxide to Sulfate in Rain
Biogenic Reduced Sulfur
Other Sulfur Sources: Biomass Burning, Volcanism, and Soil Dust
Sulfur Deposition on Land
Anthropogenic Sulfur Deposition in the United States
Atmospheric Sulfur Cycle: Human Perturbation
Radiative Forcing from Sulfate Aerosol
The Atmospheric Nitrogen Cycle and Nitrogen in Rain
N2, Nitrogen Fixation, Denitrification, and Total Nitrogen Fluxes
Nitrogen Cycle: Anthropogenic Changes and Climate
Atmospheric Nox and Nitrate in Rain
Nitrate in Rain: Anthropogenic Sources
Nitrate Deposition in Rain and the Nitrate-Nitrogen Cycle
Ammonium in Rain: Atmospheric Ammonium-Nitrogen Cycle
Ammonium in Rain
Reactive N Deposition
Acid Rain
The pH of Natural Rainwater
Acid Rain from Pollution
Acid Rain in Europe
Acid Rain in the United States from 1955 to 1985
Acid Deposition Changes in the United States from 1980 to 2007
Acid Rain in Other Parts of the World
Distinguishing Naturally Acid Rain from That Due to Pollution
Effects of Acid Rain
4. Chemical Weathering: Minerals, Plants, and Water Chemistry
Biogeochemical Cycling in Forests
Soil Water and Microorganisms: Acid Production
Chemical Weathering
Minerals Involved in Weathering
Silicate Weathering Reactions: Secondary Mineral Formation
Mechanism of Silicate Dissolution
Rate of Silicate Weathering
Silicate Weathering: Soil Formation
Carbonate Weathering
Sulfide Weathering
Groundwaters and Weathering
Garrels’s Model for the Composition of Groundwaters from Igneous rocks
5. Rivers
Introduction
Components of River Water
River Runoff
Major World Rivers
Suspended Matter in Rivers
Amount of Suspended Matter
Human Influence
Chemical Composition of Suspended Matter
Chemical Composition of Rivers
World Average River Water
Chemical Classification of Rivers
Relief and River-Water Composition
Major Dissolved Components of River Water
Chloride and Cyclic Salt
Sodium
Potassium
Calcium and Magnesium
Bicarbonate (HCO3-)
Silica
Sulfate
Sulfate Pollution and Acidic Rivers
Organic Matter in Rivers: Organic Acidity
Organic Acid Rivers
Chemical and Total Denudation of the Continents as Deduced from River-Water Composition
Nutrients in River Water
Nitrogen in Rivers: The Terrestrial Nitrogen Cycle
Reactive Nitrogen Deposition and River Transport in the United States
Phosphorus In Rivers: The Terrestrial Phosphorus Cycle
6.Lakes
Physical Processes in Lakes
Water Balance
Thermal Regimes and Lake Classification
Lake Models
Biological Processes in Lakes as They Affect Water Composition
Photosynthesis, Respiration, and Biological Cycling
Eutrophication
Limiting Nutrients
Sources of Phosphorus in Lakes
Pollutive Changes in Major Lakes: Potential Loading
Acid Lakes
Changes in Acid Lakes in the Northeastern and Upper Midwestern United States
Changes in Acid Lakes in Europe
Naturally Acid Lakes
Chemical Composition of Acid Lakes
Saline and Alkaline Lakes
7. Marginal Marine Environments: Estuaries
Introduction
Estuaries: Circulation and Classification
The Black Sea
Estuarine Chemistry: Conservative vs Nonconservative Mixing
Estuarine Chemical Processes
Inorganic (Nonbiogenic) Removal in Estuaries
Biogenic Nutrients in Estuaries
Limiting Nutrients: Nitrogen, Phosphorus, and Silica
Eutrophication from Nutrient Pollution of Estuaries
Coastal Hypoxia from Nutrient Loading and Eutrophication
Harmful Algal Blooms and Eutrophication
Suspended Sediment Deposition in Marginal Marine Environments
Antiestuaries and Evaporite Deposition
8. The Oceans
Introduction
Chemical Composition of Seawater
pH and the Human Acidification of the Oceans
Modeling Seawater Composition
Sillèn’s Equilibrium Model
Oceanic Box Models
Continuum Models
Energy Sources for Chemical Reactions
Major Processes of Seawater Modification
Biological Processes
Volcanic-Seawater Reaction
Interaction with Detrital Solids
Chemical Budgets for Individual Elements
Summary of Processes
Chloride
Sodium
Sulfur
Magnesium
Potassium
Calcium
Bicarbonate
Silica
Phosphorus
Nitrogen
References
Index
Recommend Papers

Global Environment: Water, Air and Geochemical Cycles [2 ed.]
 9780691136783, 2011039852

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

Global Environment

Global Environment Water, Air, and Geochemical Cycles Second Edition Elizabeth Kay Berner Robert A. Berner

Princeton University Press Princeton and Oxford

Copyright © 2012 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock, Oxfordshire OX20 1TW press.princeton.edu Cover photo: Scene from Makapuu Point. Oahu, Hawaii. Photo by the authors. All Rights Reserved Library of Congress Cataloging-in-Publication Data Berner, Elizabeth Kay, 1936Global environment : water, air, and geochemical cycles / Elizabeth Kay Berner, Robert A. Berner. — 2nd ed. p. cm. Includes bibliographical references and index. ISBN 978-0-691-13678-3 (cloth : alk. paper) 1. Atmospheric circulation. 2. Atmospheric chemistry. 3. Hydrologic cycle. 4. Water chemistry. 5. Geochemistry. I. Berner, Robert A., 1935– II. Title. QC880.4.A8B47 2012 551.5—dc23 2011039852 British Library Cataloging-in-Publication Data is available This book has been composed in Sabon LT Std and ITC Avant Garde Gothic Printed on acid-free paper. ∞ Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

To Abby, Katie, Sarah, Zach, Marshall, Charlotte, and Audrey

Contents

Preface to the Second Edition 1. Introduction to the Global Environment: The Water and Energy Cycles and Atmospheric and Oceanic Circulation Introduction The Global Water Cycle Major Water Masses Fluxes between Reservoirs Geographic Variations in Precipitation and Evaporation The Energy Cycle Introduction Radiation and Energy Balance Variations in Solar Radiation: The Atmospheric and Oceanic Heat Engine Circulation of the Atmosphere Oceanic Circulation Introduction Wind-Driven (Shallow) Circulation Coastal Upwelling Thermohaline (Deep) Circulation

2. Air Chemistry: The Greenhouse Effect and the Ozone Hole Atmospheric Gases Carbon Dioxide Present and Future CO2 and the Surficial Carbon Cycle Past CO2 Levels Other Greenhouse Gases: Methane, Nitrous Oxide Other Greenhouse Gases: Halogens and Tropospheric Ozone Radiative Forcing by Anthropogenic Factors Climatic Effects of Radiative Forcing: Climate Sensitivity, Global Warming, and Hydrologic Changes Observed Changes in Temperature and Atmospheric Circulation Observed Changes in the Water Cycle: Water Vapor, Precipitation, Streamflow, and Storms Observed Changes in Ice, Sea Level and the Oceans Predictions for Future Climate Change Aerosols Aerosol Cloud Effects Types of Aerosols Gaseous Emissions Sulfate Aerosols Black Carbon Aerosols Organic Carbon Aerosols Biomass Burning Aerosols Nitrate Aerosols Mineral Dust Aerosols Sea-Salt Aerosols Surface Dimming by Aerosols Aerosols and the Hydrologic Cycle Black Carbon Aerosols and Snow Cover Ozone and the Ozone Hole Stratospheric Ozone: The Ozone Hole

Tropospheric Ozone: Air Pollution

3. Air Chemistry: Rainwater, Acid Rain, and the Atmospheric Cycles of Sulfur and Nitrogen Introduction Formation of Rain (and Snow) Water Vapor in the Atmosphere Condensation Sublimation Rain (and Snow) Formation Air Motion in Cloud Formation Chemical Composition of Rainwater: General Characteristics Cl–, Na+, Mg++, Ca++ and K+ in Rain Gases and Rain Sulfate in Rain: The Atmospheric Sulfur Cycle Sea-Salt Sulfate U.S. Sulfur Emissions Conversion of Sulfur Dioxide to Sulfate in Rain Biogenic Reduced Sulfur Other Sulfur Sources: Biomass Burning, Volcanism, and Soil Dust Sulfur Deposition on Land Anthropogenic Sulfur Deposition in the United States Atmospheric Sulfur Cycle: Human Perturbation Radiative Forcing from Sulfate Aerosol The Atmospheric Nitrogen Cycle and Nitrogen in Rain N2, Nitrogen Fixation, Denitrification, and Total Nitrogen Fluxes Nitrogen Cycle: Anthropogenic Changes and Climate Atmospheric Nox and Nitrate in Rain Nitrate in Rain: Anthropogenic Sources Nitrate Deposition in Rain and the Nitrate-Nitrogen Cycle

Ammonium in Rain: Atmospheric Ammonium-Nitrogen Cycle Ammonium in Rain Reactive N Deposition Acid Rain The pH of Natural Rainwater Acid Rain from Pollution Acid Rain in Europe Acid Rain in the United States from 1955 to 1985 Acid Deposition Changes in the United States from 1980 to 2007 Acid Rain in Other Parts of the World Distinguishing Naturally Acid Rain from That Due to Pollution Effects of Acid Rain

4. Chemical Weathering: Minerals, Plants, and Water Chemistry Introduction Biogeochemical Cycling in Forests Soil Water and Microorganisms: Acid Production Chemical Weathering Minerals Involved in Weathering Silicate Weathering Reactions: Secondary Mineral Formation Mechanism of Silicate Dissolution Rate of Silicate Weathering Silicate Weathering: Soil Formation Carbonate Weathering Sulfide Weathering Groundwaters and Weathering Garrels’s Model for the Composition of Groundwaters from Igneous rocks

5. Rivers Introduction Components of River Water River Runoff Major World Rivers Suspended Matter in Rivers Amount of Suspended Matter Human Influence Chemical Composition of Suspended Matter Chemical Composition of Rivers World Average River Water Chemical Classification of Rivers Relief and River-Water Composition Major Dissolved Components of River Water Chloride and Cyclic Salt Sodium Potassium Calcium and Magnesium Bicarbonate (HCO3-) Silica Sulfate Sulfate Pollution and Acidic Rivers Organic Matter in Rivers: Organic Acidity Organic Acid Rivers Chemical and Total Denudation of the Continents as Deduced from River-Water Composition Nutrients in River Water Nitrogen in Rivers: The Terrestrial Nitrogen Cycle Reactive Nitrogen Deposition and River Transport in the United States Phosphorus In Rivers: The Terrestrial Phosphorus Cycle

6.Lakes Physical Processes in Lakes Water Balance Thermal Regimes and Lake Classification Lake Models Biological Processes in Lakes as They Affect Water Composition Photosynthesis, Respiration, and Biological Cycling Eutrophication Limiting Nutrients Sources of Phosphorus in Lakes Pollutive Changes in Major Lakes: Potential Loading Acid Lakes Changes in Acid Lakes in the Northeastern and Upper Midwestern United States Changes in Acid Lakes in Europe Naturally Acid Lakes Chemical Composition of Acid Lakes Saline and Alkaline Lakes

7. Marginal Marine Environments: Estuaries Introduction Estuaries: Circulation and Classification The Black Sea Estuarine Chemistry: Conservative vs Nonconservative Mixing Estuarine Chemical Processes Inorganic (Nonbiogenic) Removal in Estuaries Biogenic Nutrients in Estuaries Limiting Nutrients: Nitrogen, Phosphorus, and Silica Eutrophication from Nutrient Pollution of Estuaries Coastal Hypoxia from Nutrient Loading and Eutrophication Harmful Algal Blooms and Eutrophication

Suspended Sediment Deposition in Marginal Marine Environments Antiestuaries and Evaporite Deposition

8. The Oceans Introduction Chemical Composition of Seawater pH and the Human Acidification of the Oceans Modeling Seawater Composition Sillèn’s Equilibrium Model Oceanic Box Models Continuum Models Energy Sources for Chemical Reactions Major Processes of Seawater Modification Biological Processes Volcanic-Seawater Reaction Interaction with Detrital Solids Chemical Budgets for Individual Elements Summary of Processes Chloride Sodium Sulfur Magnesium Potassium Calcium Bicarbonate Silica Phosphorus Nitrogen References Index

Preface to the Second Edition During the past three decades, there has been a veritable explosion of books on environmental problems. However, this book is different in that it approaches the environment almost entirely in terms of global geochemical cycles both as they occur naturally and as they are affected by human activities. Here we emphasize such important problems as global warming, acid rain, rock weathering, erosion, eutrophication of both lakes and estuaries, and ocean acidification. This book is intended for those who have a fundamental understanding of elementary chemistry, but it requires no other background in science, whether it be in biology, geology, meteorology, oceanography, hydrology, soil science, or environmental science. Our approach is multidisciplinary and covers all of these fields, but we do it from an elementary standpoint. Mathematical complexity is held to an absolute minimum, with the only requirement being some previous training in chemistry at the college freshman, or even the advanced high school, level. The book is appropriate as a primary or secondary text in junior- or senior-level undergraduate courses, or beginning graduate courses, in environmental geochemistry, environmental geology, global change, biogeochemistry, water pollution, geochemical cycles, chemical oceanography, and geohydrology. Because we provide extensive data on natural fluxes of chemicals, the book is also of reference value to researchers on global geochemical and environmental problems. Much of this book is devoted to the natural behavior of the Earth’s surface. We attempt to quantify the rates by which the major constituents of rocks, water, air, and life are transferred from one

reservoir to another and to track down the sources of each constituent. We feel that a knowledge of geochemical cycles in the prehuman state is necessary before one can discuss how humans have perturbed these cycles. The present book is a second edition of our 1996 book by the same title. Here we have put in extra effort to update the increasing amount of information being published on important global environmental problems. This includes, among other things, an exhaustive study of changing climate and atmospheric chemistry by the Intergovernmental Panel on Climate Change (ICPP) in 2001 and 2007; new findings on how the problem of acid rain is being ameliorated somewhat; further information on the euthrophication of lakes, rivers, and estuaries; major advances in the study of chemical weathering; and the new global environmental problem of ocean acidification. The material in this book has been used in courses over the past three decades by E. K. Berner at Wesleyan University and the University of Connecticut and by R. A. Berner at Yale University. We are grateful for the various suggestions made to us by both students and teaching assistants during the preparation of this book and its predecessor. Special thanks are due to Danny Rye for help in photograph preparation. Elizabeth Kay Berner Robert A. Berner North Haven, Connecticut April 12, 2011

Global Environment

1 Introduction to the Global Environment: The Water and Energy Cycles and Atmospheric and Oceanic Circulation Introduction In this book we shall be concerned with the principal constituents of rocks, water, and life as they circulate through the land, the sea, and the air. In other words, our concern will be with the geochemistry of the Earth’s surface and how it operates naturally and how it has been perturbed by human activities. The approach is global in scope, and because of their special importance in Earth surface geochemical cycles, water and air will receive major attention. Water moves from the atmosphere to the land surface as rain containing pollutants added to the air by humans. Humans also add other gases and solids to the atmosphere that result in changes in atmospheric composition and climate. The air flows from one place to another, distributing pollutants over wide areas. Water once on the ground reacts with minerals contained in rocks, via a process known as chemical weathering, with a change in chemical composition of the water and with soil formation. Plants circulate elements between the atmosphere and soils. Water eventually makes its way into rivers, which obtain dissolved pollutants as well as suspended material

derived from human-accelerated erosion. The rivers then deliver their load to the oceans, where a variety of chemical and biological processes occur. Overall, it is these fluxes of water and air that ultimately act to maintain the overall chemical and physical conditions at the Earth’s surface. Throughout most of this book we will concentrate on the principal constituents transported by water and air. These are sodium, potassium, calcium, magnesium, silicon, carbon, nitrogen, sulfur, phosphorus, chlorine, and, of course, hydrogen and oxygen. We shall point out how the global cycles of some of these elements have been perturbed by humans, resulting in such things as greenhouse gases, acid rain, eutrophic lakes, and oceanic acidification. Although they are of geochemical and environmental interest, we shall not be concerned with minor and trace elements (e.g., lead and mercury) or exotic synthetic chemicals (e.g., pesticides), since our goal is not allinclusive environmental coverage. (The interested reader is referred to books such as those by Laws 2000 and Mackenzie 2011 for detailed discussion of environmental problems.) Rather, the approach is that of geochemists trying to understand how global chemical cycles operate and how they affect the major constituents of rocks, water, air, and life. Because of their importance to climate and chemical changes in the global environment, emphasis in this chapter will be placed on the atmosphere and the oceans, and how their circulation operates; a discussion of the water and energy cycles of the Earth and their role in meteorology and oceanography will be included. The chapter, then, helps to set the stage for discussions of such subjects as the atmospheric greenhouse effect (chapter 2), acid rain (chapter 3), and chemistry of the oceans (chapter 8).

The Global Water Cycle Major Water Masses

Earth is the only planet in the solar system having an abundance of liquid water on its surface; about 70% of the Earth is covered by liquid water. Because of the particular combinations of temperature and pressure on the planet’s surface, water can exist here in three states: as liquid water, as ice, and as water vapor. This is in sharp contrast to the surface of the planet Mars, for example, which is so cold and dry that water can exist there only as ice or as water vapor. Water is by far the most abundant substance at the Earth’s surface. There are 1444 × 106 km3 of it in its three phases: liquid water, ice, and water vapor. As shown in table 1.1, most of the Earth’s water (97%) is stored as seawater in the oceans. The remaining 3% is either on the continents or in the atmosphere. The amount of water in the atmosphere, in the form of water vapor, is very small in comparison with the other reservoirs, only around 0.001% of the total. However, it plays a very important role in the water cycle, as we shall see. Of the fresh water stored on the continents, around two-thirds is in the form of ice in ice sheets, polar ice caps, and glaciers. Most of the rest of the continental water is present either as subsurface groundwater or in lakes and rivers. It is this small part of the Earth’s total water (1%) that humans draw on for their water supplies. Here we shall focus on water near the Earth’s surface and how it moves within and between the various reservoirs.

Fluxes between Reservoirs Water does not remain in any one reservoir but is continually moving from one place to another in the hydrologic cycle. This is illustrated in figure 1.1. (For a more detailed discussion, see chapter 5 for runoff; also Penman 1970, Baumgartner and Reichel 1975, NRC 1986, and Chahine 1992). Water is evaporated from the oceans and the land into the atmosphere, where it remains for only a short time, on the average about eleven days, before failing back to the surface as snow or rain. Part of the water falling onto the continents runs off in rivers and, in some places, accumulates temporarily in lakes.

Some also passes underground only to emerge later in rivers, lakes, and the ocean. The remaining portion of the precipitation on the continents is returned directly to the atmosphere via evaporation. Over the oceans, evaporation exceeds precipitation, with the difference being made up by input via runoff from the continents. An idea of the sizes of these various fluxes of water (mass transported per unit time) is shown in figure 1.1. Table 1.1 Inventory of Water at the Earth’s Surface

Reservoir

Volume 106 km3 (1018 kg)

Oceans Mixed layer Thermocline Abyssal

1400 50 460 890 0.7 Ice shelves (floating)a Ice caps and 0.09 glaciersbb Ice sheetsa Greenland Antarctica Groundwater Lakes Rivers Soil moisture Atmosphere totalb Terrestrial Oceanic Biosphere Approximate Total

Percent of Total 96.95

0.048 0.006

27.6

1.9

2.9 24.7 15.3 0.125 0.0017 0.065 0.0155

1.06 0.009 0.0001 0.0045 0.001

0.0045 0.0110 0.002 1444

0.0001

Source: NRC 1986; Berner and Berner 1987; Lemke et al. 2007. a Lemke et al. 2007. b As liquid volume equivalent of water vapor.

In order to conserve total water, evaporation must balance precipitation for the Earth as a whole, since the total mass of water at the Earth’s surface is believed to be constant over time. The average global precipitation rate, which is equal to the evaporation rate, is 0.506 × 106 km3/yr. For any one portion of the Earth, by contrast, evaporation and precipitation generally do not balance. On the land, or continental part of the Earth, the precipitation rate (0.108 × 106km3/yr) exceeds the evaporation rate (0.071 × 106 km3/yr), whereas over the oceans evaporation (0.435 × 106 km3/yr) dominates over precipitation (0.398 × 106 km3/yr). The difference in each case (0.037 × 106 km3/yr) comprises water transported from the oceans to the continents as atmospheric water vapor, or that returned to the oceans as river runoff (see fig. 1.1). Although inaccurately known, there is a considerably smaller amount of direct groundwater discharge to the oceans (0.0022 × 106 km3/yr)(Korzun et al. 1977).

Figure 1.1 The hydrologic cycle. Numbers in parentheses represent inventories (in 106 km3 = 1018 kg) for each reservoir. Fluxes are in 106 km3/yr per year (1018 kg/yr) Data from

table 1.1, NRC 1986, and chapter 5.

The values given in figure 1.1 have built-in errors. Precipitation is difficult to measure, rainfall measurements over the oceans are few, and oceanic evaporation is necessarily estimated from models. Also, because the evaporation rates over many land areas of the Earth have not been measured, the values given in figure 1.1 are based, by necessity, on the difference between measured worldwide precipitation on land and river runoff values. (Recently, attempts have been made to obtain evaporation rates by other means, including atmospheric water vapor balance and heat balance; see the section Radiation and Energy Balance below.) Assumption of constant volume of water in a given reservoir (water mass) enables the use of the concept of residence time. The residence time is defined as the volume of water in a reservoir divided by the rate of addition (or loss) of water to (from) it. It can be thought of as the average time a water molecule spends in a given reservoir. For the oceans, the volume of water present (1,400 × 106 km3; see fig. 1.1) divided by the rate of river runoff to the oceans (0.037 × 106 km3/yr) gives a residence time of 38,000 years. This long residence time, which can also be thought of as a filling or replacement time, reflects the very large volume of water in the oceans. By contrast, the residence time of water in the atmosphere, relative to evaporation from both the oceans and the continents, is only about eleven days. Lakes, rivers, glaciers, and shallow groundwater have residence times lying between these two extremes, but because of extreme variability, no simple average residence time can be given for each of these reservoirs.

Geographic Variations in Precipitation and Evaporation The values shown in figure 1.1 are only average values for precipitation and evaporation over the continents and oceans. From one region to another there is considerable variation, as can be seen

in figure 1.2, which shows the mean annual precipitation for different areas of the continents.

Figure 1.2 Global average annual precipitation. After McKnight 1996.

In order to have rain or snow, there must be both sufficient water vapor in the atmosphere and rising air that can carry the water vapor up to a height where it is cold enough for condensation and precipitation to occur (see chapter 3). Net precipitation (precipitation minus evaporation), as shown in figure 1.3, is highest near the equator (10°N to 10°S) and at 35° to 60° north and south latitudes, where there is frequent storm activity with its accompanying air motion. Net precipitation is lowest in the subtropics (15° to 30° N and S), where the air is stable, and near the poles, which have both stable air and a very low moisture content due to low temperatures. (However, since there is also very low evaporation near the poles, precipitation can exceed evaporation in certain places, resulting in the formation of the ice caps of Greenland and Antarctica.) In continental areas of high rainfall, runoff is also high. Examples are the large rivers of equatorial regions (Amazon, Zaire, Orinoco) and those of middle latitudes such as the Mississippi and Hwanghe (Yellow).

Evaporation also varies considerably over the Earth’s surface. For net evaporation to occur, there must be a heat source (i.e., radiation from the sun), a low moisture content in the air, and the presence of water available for evaporation. In arid regions the evaporation rate is high but is limited by the availability of water. Overall, as shown in figure 1.3, evaporation exceeds precipitation at subtropical latitudes (15° to 30° N and S). Over the continents this leads to the formation of large deserts at these latitudes (for example, the Sahara Desert in Africa and the Great Desert of Australia). The greatest evaporation rates on the Earth (more than 200 cm/yr), however, occur over the subtropical oceans, such as over the Gulf Stream, in winter where warm water, which is carried northward, encounters cooler, drier air and evaporates. High evaporation rates can lead to locally higher ocean salinities due to the removal of almost pure water during evaporation, leaving dissolved sea salt behind. An outstanding example of this is the Mediterranean Sea, which has a salinity notably higher than that of the open ocean (e, g., see chapter 7).

Figure 1.3 Net precipitation (precipitation minus evaporation) as a function of latitude. Positive values represent net precipitation while negative values represent net evaporation. After Peixoto and Kettani 1973.

The Energy Cycle Introduction In this section we shall take a look at the energy cycle of the Earth. This is what drives the water cycle and the circulation of the atmosphere and oceans. Atmospheric water vapor remains in the atmosphere for only about eleven days on the average. However, during this time it has travelled a mean distance of 1000 km (Peixoto and Kettani 1973), and this transport is controlled by the energy cycle of the Earth. The energy cycle, in turn, is greatly influenced by the presence of water vapor in the atmosphere. Thus, the Earth’s energy and water cycles are intimately interconnected, and they exert strong influences on one another. (For more information on the Earth’s energy cycle, see Ingersoll 1983; Ramanathan 1987; and Trenberth et al. 2009).

Radiation and Energy Balance The primary energy sources for the Earth’s surface are summarized in table 1.2. As can be seen, radiation from the sun is by far the most important source of energy (99.98%), and consequently it is the dominant influence on the circulation of the atmosphere and oceans. It is the only energy source that will be discussed here. The incoming solar radiation impinges on the top of the atmosphere, below which it is reflected, absorbed, and converted into other forms of energy. Apportionment of this energy into different forms is described in terms of the radiation or energy balance of the Earth (fig. 1.4). As shown by the flux values near the top of figure 1.4, 341 Wm–2 of incoming, mainly short-wave (4 μm) energy flux from the Earth (239 Wm–2). This must be true to avoid rapid heating up or cooling of the earth and has been verified by satellite measurements

(Ramanathan 1987). This does not mean, however, that very slight imbalances over extended periods cannot bring about cooling or heating. An example of this is the large climate changes that occurred at northern midlatitudes during the past 20,000 years, as evidenced by extensive glaciation and deglaciation that occurred during this time. Table 1.2 Primary Energy Sources for the Earth

Source

Energy Flux cal/cm2/min)

Percent of Total Energy Flux

Solar radiation

0.5a

99.98

Heat flow from interior 0.9 × 10–4 of earth Tidal energy 0.9 × 10–5

0.018 0.002

Source: Hubbert 1971; Flohn 1977. a 0.5 cal/cm2/min is approximately equal to 341 Watts/m2, the solar flux used in Fig.1.4, after Trenberth et al. 2009; (1 Watt = 0.2389 cal/sec).

Figure 1.4 The mean annual radiation and heat balance of the atmosphere and Earth for the period March 2000 to May 2004. Units are in watts per square meter. Short-wave solar

radiation is that with 4 µm and is shown on the right side of the diagram. After K. E. Trenberth et al. 2009; © Copyright 2009 American Meteorological Society.

The 102 Wm–2 of reflected short-wave solar energy (fig. 1.4) represents about 30% of the incoming flux and is known as the Earth’s albedo. It is a measure of how bright the Earth would appear if viewed from outer space. Note that 79 Wm–2 of this reflected radiation is by clouds and the atmosphere and does not reach the Earth’s surface. At the surface just 23 Wm–2 is reflected back to space. Thus, changes in cloud cover are very important to radiative heating of the surface. The remaining 70% of the incoming solar radiation represents absorption and reradiation as long-wave (>4 μm) infrared radiation by the Earth’s surface and various components of the atmosphere. Because the sun is very hot (6000°C), most of the solar radiation is in shorter wavelengths (less than 4 μm), with the peak radiation in the visible wavelengths (0.4–0.7 μm). Much of this visible radiation, or light, penetrates the atmosphere and ultimately reaches the ground. This is important because life is dependent on the absorption of light by photosynthetic organisms. By contrast, almost all of the ultraviolet solar radiation (4 μm) with a maximum in the infrared at about 10 μm (see fig. 1.5). Atmospheric water vapor, carbon dioxide, and rarer gases such as methane are good absorbers of infrared energy in this range of wavelengths. Because of this, most of the infrared radiation originating from the Earth’s surface is absorbed by these gases

(mostly by water vapor), and very little of it escapes directly to space. As a result the spectrum of outgoing radiation leaving the Earth differs considerably from that expected for a black body at 15°C (see fig. 1.5). Of the 396 Wm–2 of long-wave (infrared) radiation emitted by the Earth’s surface, 333 Wm–2 are both absorbed by atmospheric water vapor and other gases and reradiated back to the ground. Here it is reabsorbed, keeping the Earth warm. Clouds also contribute to warming of the Earth by reducing long-wave emissions to space because at their bases they absorb radiation emitted by the warmer Earth surface and at their tops they emit to space at colder temperatures. However, as discussed above, clouds also reflect solar radiation, and because this process outweighs their warming effect, the net effect of clouds is global cooling (Meehl et al. 2007; see also chapter 2). The role of atmospheric water vapor and carbon dioxide in allowing the incoming short-wave solar radiation to pass through to the ground, while absorbing and reradiating to the Earth’s surface most of the Earth’s outgoing long-wave radiation, is referred to as the atmospheric greenhouse effect by comparison to the glass in a greenhouse. A greenhouse lets solar radiation in but keeps the greenhouse warm by preventing long-wave radiation from leaving because of absorption of the long-wave radiation by the glass. The greenhouse effect makes the Earth’s surface much warmer (around 30°C warmer, according to IPCC 2007) than it would be otherwise. Lately, there has been much concern about the atmospheric buildup of carbon dioxide, released from fossil fuel burning, and other greenhouse gases in that they may absorb more than the normal amount of the Earth’s outgoing radiation, resulting in an enhanced greenhouse effect and global warming (see chapter 2). In order to maintain a constant Earth surface temperature, the amount of incoming solar radiation received at the surface (161 Wm−2) should be balanced by the net loss of long-wave radiation from the surface to the atmosphere and space (396 – 333 = 63

Wm−2) plus the fluxes of sensible heat via thermal updrafts (17 Wm–2) and latent heat from evapotranspiration (80 Wm–2). However, figure 1.4 shows that the sum of the three upward fluxes is less, by about 1 Wm–2, than that of the incoming solar radiation. This represents the estimated warming of the Earth, mainly by human activities. When liquid water is evaporated to form atmospheric water vapor, heat (energy) is absorbed. This is what is called latent heat, since upon subsequent condensation of the water vapor into rain and snow, the previously added energy is released as heat to the atmosphere. Since condensation can occur at great distances from the original site of evaporation, the transport of water vapor in the atmosphere also involves the transport of heat. Of the 161 Wm–2 of solar energy absorbed at the Earth’s surface, 80 Wm–2 are used to evaporate water, giving rise to an equivalent latent heat flux from the land and oceans to the atmosphere (see fig. 1.4). It is this latent heat flux that drives the water cycle. Knowledge of the latent heat flux to the atmosphere can be used to calculate the global rate of evaporation. Since it takes 2.46 × 109 joules to convert 1 m3 of liquid water to water vapor at the average Earth surface temperature of 15°C, the rate of evaporation corresponding to the dissipation of 80 Wm–2(fig. 1.4) as latent heat over the Earth’s surface (510 × 106 km2, as follows (after Miller et al. 1983 and Budyko and Kondratiev 1964, updated to include accepted units and the energy data of Trenburth et al. 2009):

This is equivalent to 522,000 km3 of water per year, which agrees fairly well with estimates of total annual evaporation from the Earth (506,000 km3 from fig. 1.1) based on setting the total evaporation equal to measured total precipitation in the global water balance.

Variations in Solar Radiation: The Atmospheric and Oceanic Heat Engine The amount of solar radiation absorbed by the Earth decreases with latitude from the equator to the poles. It is this variation in the Earth’s heating that drives the circulation of the ocean and atmosphere and, thus, much of the hydrologic cycle. Latitudinal variations in the input of solar energy are due to two factors. First, the Earth is a sphere and the angle at which the sun’s rays hit its surface varies from 90° (or vertical) near the equator to 0° (or horizontal) near the poles. This is shown in figure 1.6. Less energy is received at the poles because the same amount of radiation is spread out over a much larger area at high latitudes (compare situation C with situations A and B in fig. 1.6) and because at high latitudes the sun’s rays must travel through a much greater thickness of atmosphere where more absorption and reflection occur. The second factor affecting latitudinal variations in heating is the duration of daylight. Because the polar axis of the Earth is tilted at an angle of 23.5° with respect to the ecliptic (the plane of the Earth’s orbit about the sun), we have a progression of seasons where the angle of the sun’s rays striking any given point varies over the year. This is shown in figure 1.7. In the Northern Hemisphere winter no sunlight strikes the area around the North Pole, during a full day’s rotation of the Earth, because it is in the Earth’s shadow. Thus, little or no solar heating occurs in this area at this time. Conversely, at the South Pole there is continual daylight, but at a very low sun angle, during the Northern Hemisphere winter. As the seasons shift, the South Polar region eventually becomes plunged into twenty-four-hour darkness (during the Southern Hemisphere winter and Northern Hemisphere summer), just as the North Pole had been earlier. Low-latitude regions near the equator, by contrast, undergo little seasonal change in the duration of daylight, whereas intermediate latitudes are subjected to changes intermediate between those of the poles and

the equator. Thus, because of seasonality, more annual radiation is received per unit area at lower, as compared to higher, latitudes. The effects of variation in angle of the sun’s rays with latitude and seasonal changes in the amounts of daylight result in strong variation with latitude in the total solar radiation received during a year, and this result applies equally to both the Northern and Southern Hemispheres. However, because the Northern Hemisphere is dominated by land and the Southern Hemisphere by water, one might expect hemispheric differences in received radiation due to differences in the albedo of land vs water. Nevertheless, according to satellite measurements, the annual mean albedo (relection of solar radiation) of both hemispheres is nearly the same. This is because of the dominant influence of clouds (over surface effects) in determining the mean albedo of each hemisphere (Ramanathan 1987).

Figure 1.6 Schematic diagrams showing the variations of solar intensity (energy per unit area) with angle of incidence to the Earth’s surface. Lower angles (higher latitudes) result in the same energy spread out over a larger area and, thus, in a lower intensity of radiation. Scene depicted is for Northern Hemisphere winter. Adapted from Miller et al. 1983.

The long-wave radiation leaving the Earth also varies with latitude but less strongly. The differences between the amount of solar radiation received and long-wave radiation emitted results in radiation imbalances over the surface of the Earth (fig. 1.8). From 35° north and south latitudes to the poles (mid-and polar latitudes) there is a net deficit of radiation (more leaves than enters), whereas from 35° to the equator (tropical and subtropical latitudes) there is a net surplus (more solar radiation enters than the Earth radiates back). To keep the poles from getting colder and the tropics from getting warmer, heat must be transported from lower to higher latitudes. This is accomplished by the circulation of the atmosphere and oceans. Thus, the atmosphere and oceans act like a “heat engine” driven by latitudinal variations in solar radiation. By contrast there is no transfer of heat across the equator because both hemispheres have similar zoned energy balances (Ramanathan 1987).

Figure 1.7 Revolution of the earth in its orbit around the sun, showing the changing seasons (also length of daylight). The seasons given are for the Northern Hemisphere; they are reversed in the Southern Hemisphere. Modified from Lutgens and Tarbuck 1992, fig 2–3, p. 29.

Heat is transported from the equator to the poles in three ways: (1) by ocean currents carrying warm water, (2) by atmospheric circulation (wind) carrying warm air, and (3) by atmospheric circulation carrying latent heat in the form of water vapor. The maximum meridional heat transport occurs around 30° latitude (Ramanathan 1987). In the northern hemisphere, the heat transport by the oceans and the atmosphere are roughly comparable in size but the exact fluxes are not well known (Chahine 1992). Warm, wind-driven ocean currents transport heat poleward from the zone between 20°N and 20°S; they are well known, examples being the Gulf Stream in the North Atlantic and the Kuroshio Current in the North Pacific. The poleward transport of warm water by the oceans tends to warm the overlying atmosphere at higher latitudes, especially during the winter. This provides a moderating influence on climate and helps to explain the relatively mild winters experienced,

for example, by western Europe (Gulf Stream) and by Japan (Kuroshio Current). The atmosphere carries heat poleward as warm air and latent heat. A major source of warm air is the tropical region between 10°N and 10°S, which is the zone of maximum surplus of solar radiation. The tropical air is heated both by sensible heat and by the release of latent heat upon condensation of moisture. (The hot tropics are a zone of both high evaporation and high rainfall, with the latter predominating). Much additional latent heat in the form of water vapor is injected into the atmosphere in the subtropic zones (15°– 30° N and S) where evaporation exceeds precipitation. From there, poleward transport of the latent heat takes place. The latent heat is subsequently released by condensation, which warms the atmosphere in the midlatitude zones of intense storm activity at 30°– 50° north and south. Latent heat from condensation in clouds provides 30% of the heat energy that is carried by the atmospheric circulation (Chahine 1992).

Figure 1.8 Annual zoned mean estimates for both hemispheres, which are nearly the same, of absorbed solar radiation and outgoing long-wave radiation emission obtained by satellites. Shaded regions denote net heating and dashed region denotes net cooling. After Ramanathan 1987, fig. 3, p. 4076, based on data from Ellis and Vonder Haar 1976.

A small part (about 0.7%) of the incoming solar radiation is converted into the energy of motion (kinetic energy) of ocean currents, winds, and waves. Although this is a small number, it represents an energy of major interest to the hydrologic cycle, that associated with the circulation of the atmosphere and oceans. This circulation will be discussed next.

Circulation of the Atmosphere The atmosphere circulates as a consequence of the latitudinal heat imbalance discussed above. If the circulation were due solely to heating, hot air would rise at the equator and flow poleward at high levels. As the air was cooled in transit and piled up at the poles, it would tend to sink at the poles. To complete the cycle, there would be a return flow near the Earth’s surface of cool air toward the equator, where heating would produce two symmetrical closed circuits, or cells, one in the Northern and one in the Southern Hemisphere. Such a circulation (incorporating the Earth’s rotation) was originally proposed in 1735 by George Hadley, but because of a number of factors, the actual circulation has turned out to be considerably more complicated.

Figure 1.9 Schematic representation of the general circulation of the atmosphere. Modified from Lutgens and Tarbuck 1992, fig. 8–3, p. 170.

The general circulation of the atmosphere, showing the mean annual winds, is depicted in figure 1.9. Note that the winds do not simply blow along north-south, or meridional, lines. This is because they are deflected by the rotation of the Earth. The force that deflects moving objects—in this case, moving air masses to the right in the Northern Hemisphere and to the left in the Southern Hemisphere—is referred to as the Coriolis force. The general circulation differs from the simple Hadley circulation by being broken up into several latitudinal zones; however, there is still symmetry more or less between the Northern and Southern Hemispheres. For the sake of brevity, we will discuss only the Northern Hemisphere, but what is said applies equally to the Southern Hemisphere (with a leftwardinstead of rightward-directed Coriolis force). (For details of circulation not discussed here, the interested reader is referred to books on

meteorology and climatology such as Lutgens et al. 2009 and Barry and Chorley 1998.) Hot moist air rises at the equator, and as it rises the moisture condenses and intense precipitation results, releasing latent heat. Here there are only very weak surface winds, giving rise to the equatorial doldrums. After rising, the air flows northward at high levels, cools, and eventually sinks around 30°N. The descending air is very dry (having lost most of its moisture in the tropics), and when it is warmed, its capacity to take up moisture is further increased. The resulting hot dry air causes intense evaporation at the Earth’s surface, and this gives rise to the subtropical belt of deserts centered between 15° and 30°N. After reaching the surface, the air flows southward, picking up moisture as it flows over the ocean and being deflected to the right by the Coriolis force. This surface flow is known as the northeast trade winds. Upon reaching the equator the northeast trades converge with the southeast trades from the Southern Hemisphere, in the Intertropical Convergence Zone (ITC), and the air rises at the equator to complete the low-latitude cycle known as the Hadley cell (which behaves rather as Hadley expected the whole atmospheric circulation to behave). At around 30°N, additional air descends and then flows north at the surface rather than south. This is the beginning of the Ferrel cell. The northward flowing air is deflected to the right, forming the prevailing westerlies, which flow from southwest to northeast in the Northern Hemisphere. The westerlies continue until they encounter a cold mass of air moving south from the North Pole at about 50°N. This zone where the air masses meet is known as the polar front, and it is a region of unstable air, storm activity, and abundant precipitation. The polar jet stream (a very fast air stream) occurs at this boundary. The warmer air from the south rises over the polar air and then turns south at high altitude to complete the Ferrel cell. Meanwhile the southward flowing polar air (polar easterlies) becomes warmed by condensation at the polar front and by contact with the southern air. As a result, it too rises and then flows northward at high altitude to the pole, where it sinks, thus completing the polar cell.

In the midlatitudes, the west-to-east flow (Northern Hemisphere westerlies) is subject to considerable turbulence because of the Earth’s rotation. At higher levels of the atmosphere, the flow forms waves called planetary waves) that transport warm air from the surface to the top of the atmosphere (Ingersoll 1983). These waves are expressed, in the lower atmosphere, in a series of storms that travel west to east around the globe, transporting warm air poleward and cool air equatorward and releasing heat by precipitation.

Oceanic Circulation Introduction The oceans can be divided into two portions for the purpose of discussing circulation. The top 50–300 m, or surface layer (fig. 1.10), is stirred by the wind and is well mixed from top to bottom. Below the surface layer (also referred to as the mixed layer), the remaining deeper water is colder, less well mixed, and divided into a number of roughly horizontal layers of increasing density. The deep water is separated from the surface water by a region of steeply decreasing temperature gradient, known as the thermocline, about 1 km thick (fig. 1.10), across which there is limited communication between the surface and deep water. The deep part of the ocean below the thermocline is referred to as the abyssal zone. The volume of the surface (mixed) layer is small, comprising only 3.5% of the total ocean volume, while about one-third of the remaining volume is in the thermocline and two-thirds in the abyssal zone (see table 1.1).

Figure 1.10 Generalized temperature-vs-depth profile for the oceans (at low to-midlatitudes) showing vertical stratification into surface and deep water masses.

In the surface ocean, lateral circulation is predominantly driven by the wind; in the deep ocean, circulation is driven by density variations due to differences in temperature and salinity, giving rise to the term thermohaline circulation. In this section the wind-driven and thermohaline circulations are discussed separately.

Wind-Driven (Shallow) Circulation The circulation of the shallow ocean is driven by prevailing winds, which are in turn caused by uneven heating of the Earth’s surface. The circulation pattern can be summarized as a number of current gyres that flow clockwise in the Northern Hemisphere and counterclockwise in the Southern Hemisphere due to the stresses imparted by the prevailing winds. These gyres extend poleward from about 10°N and S of the equator to about 45°N and S. This is shown in figure 1.11. Each gyre has a strong narrow poleward current on the western side with much weaker currents on the east. The most

pronounced of these “westward intensification” currents are found in the Northern Hemisphere as the Gulf Stream in the Atlantic Ocean and the Kuroshio Current in the Pacific. The circulation pattern shown in figure 1.11 is not what is expected if water were simply carried downwind. Transport of surface water, on the scale of the oceans, involves factors such as the Earth’s rotation and friction against continents, as well as wind stresses on the surface. The interaction of wind and water is complicated, and only a brief summary will be given here. (For details, the interested reader should consult references on physical oceanography such as Pickard and Emery 1982, and Pedlosky 1990, or general oceanography texts such as Knauss 1997.)

Figure 1.11 Surface currents of the oceans. After Drake et al. 1978, fig. 6.1, p. 88.

The origin of the prominent gyres can best be understood by reference to the North Atlantic. Here the prevailing winds are westerlies (from the west) at 40°–50°N and trade winds (from the

east) at 15°–30°N. Because of the Coriolis force (see previous section), water in the top layer does not simply move downwind, but instead is moved to the right of the wind direction (Ekman flow). This brings about a convergence or piling up of water from both the north and south into the central portion of the North Atlantic in the Subtropical Convergence. (This area is also known as the Sargasso Sea.) The piled-up water then sinks, and just below the surface, begins to return to the north and south. As it does, it is turned to the right by the Coriolis force, which results in a strong east-flowing current on the north and a strong west-flowing current on the south, just below the surface. These so-called geostrophic currents flow until they encounter the European continent at the northeast portion of the gyre and the North American continent along the southwest portion. Here they must turn. Since friction is strong along the continents and reduces the effect of the Coriolis force, the currents will flow from high pressure (where water is accumulating due to the current) to low pressure (where it is being removed), resulting in a north-to-south-flowing current on the European side and a south-tonorth-flowing current (Gulf Stream) on the North American side. In this way a clockwise-flowing gyre results. This type of explanation for the North Atlantic circulation can be applied to the other oceans, except that in the Southern Hemisphere the Coriolis force is to the left of the direction of motion, and as a result, the gyres are counterclockwise (see fig. 1.11). In the northern hemisphere the current on the west side of the Atlantic becomes intensified because of the superimposition of an additional pressure gradient (decreasing pressure from south to north due to the variation in Coriolis force) that reinforces the western current and opposes the eastern current. Another prominent surface current is the Antarctic Circumpolar Current, which flows from west to east entirely around Antarctica and is driven by strong westerly winds. This is the primary current connecting the three ocean basins.

Coastal Upwelling

A special case of the wind-driven circulation is coastal upwelling, which has an important effect on biological productivity in the ocean (see chapter 8). Major upwelling occurs along the western boundaries of the continents, where the surface currents flowing toward the equator are broad and relatively weak. Winds blowing toward the equator along the coasts bring about a transport of surface water offshore. This is because the net transport of water (Ekman drift) is to the right of the wind in the Northern Hemisphere and to the left of the wind in the Southern Hemisphere. Transport of surface water away from shore leaves a nearshore deficit that is replaced by deeper water flowing up from below (see fig. 1.12). This deeper water is rich in nutrients, which results in high planktonic productivity and teeming life, including abundant fish. Some classic examples of upwelling areas are those off Peru and Chile, the bulge of West Africa, Namibia in southwest Africa, and the California coast. Upwelling also results when surface waters are blown or transported away from an open-ocean area, bringing about the phenomenon of divergence. In such cases, subsurface waters move in to replace the missing surface water. Some noncoastal upwelling areas include the eastern equatorial Pacific, the sea around Antarctica, and the oceans at high northern latitudes (Kennett 1982).

Thermohaline (Deep) Circulation Below the top few hundred meters, the oceans are not directly affected by the winds. Here circulation is brought about by density differences arising from differences in temperature and salinity. (The following discussion of deep ocean circulation is from Pickard and Emery 1982, Warren 1981, and Gordon 1986; see also Drake et al. 1978 and Knauss 1997.) In seawater, density increases continuously with decrease of temperature, and no density maximum, as is found in fresh water at 4°C (see chapter 6), is encountered. Density also increases with increasing salt content, or salinity. Overall the deep ocean is vertically stratified into various water masses, with the densest water at the bottom and the lightest at the top, and the density stratification is due primarily to the decrease of temperature

with depth. The stratification severely inhibits vertical motion; in other words, it is much easier to move water along surfaces of constant density (isopycnals) than it is to move it across them. Thus, the deep-water circulation can be viewed as primarily horizontal.

Figure 1.12 Upwelling, or the result of Ekman drift, in response to a north-blowing wind in the Southern Hemisphere. After Turekian1976.

The density stratification and density differences between water masses of the deep sea owe their origin to surface processes. Here changes in density are brought about by heating and cooling, evaporation, addition of fresh water, and freezing out of sea ice. Surface water migrating to high latitudes becomes more dense because of evaporation (which causes both cooling and increased salinity) and because of loss of sensible heat and sea-ice formation. (When sea ice forms, dissolved salts are excluded from the ice and the remaining water becomes more saline and, thus, more dense.) In certain locations in the far North and far South Atlantic, this surface density increase becomes so great that the cold, salty surface water

occasionally, during winter, becomes denser than the underlying water and sinks downward to replace it. In this way deep ocean water originates and is replenished from above. Once it reaches great depths, the water tends to conserve its temperature and salinity as it flows laterally throughout the oceans, bringing about the deep water circulation. An example of stratification and deep circulation for the Atlantic Ocean is shown in figure 1.13. Here each water mass is identified by its characteristic temperature and salinity. The deep water is dominated by two cold water masses, the North Atlantic Deep Water and the Antarctic Bottom Water. The North Atlantic Deep Water (NADW) originates in the Norwegian Sea off Greenland from the cooling of Gulf Stream surface water by evaporation and by surface cooling to form sea ice, leaving denser salty water behind. This water sinks and flows at depth southward, where it is joined by water sinking in the Labrador Sea off Canada. Ultimately this water crosses the equator. In the Antarctic an even denser water, the Antarctic Bottom Water, is formed in the Weddell Sea by the cooling and freezing-out of sea ice in the winter. This sinks to the bottom and flows north. In the Antarctic region further north, at the Antarctic Convergence, is formed the Antarctic Intermediate Water, which also sinks, but not to the bottom, because it cannot displace the underlying denser North Atlantic Deep or Antarctic Bottom waters. This water thus occupies intermediate depths, as implied by its name.

Figure 1.13 South-north vertical section of water properties of the Atlantic Ocean along the western trough as delineated by lines of constant temperature and salinity. N. Atl. Deep = North Atlantic Deep Water; Ant. Bott. = Antarctic Bottom Water; Ant. Int. = Antarctic Intermediate Water; Medit. = Mediterranean Water. Adapted from Pickard and Emery 1982, based on data from Bainbridge 1976.

Another intermediate-type water is formed in the Mediterranean Sea. Here intense evaporation causes the water to be sufficiently saline and dense that, upon passing out into the Atlantic through the Straits of Gibraltar, it sinks and fills intermediate depths. It does not sink to the bottom because, although it is more saline than North Atlantic Deep Water, it is also warmer and the temperature difference counteracts the salinity effect making it less dense than NADW. For the same reason, surface water at lower latitudes remains at the surface, even though it is more saline (see fig. 1.13, bottom); that is, the density-lowering effect of higher temperature overpowers the density-raising effect of higher salinity. Ultimately the Mediterranean water moves west and south across the Atlantic and joins the NADW in the western boundary current (Gordon 1986).

Figure 1.14 Global cycle of thermohaline circulation. Deep water circulation (darker lines) originates in the North Atlantic by sinking of North Atlantic Deep Water (NADW). NADW starts with water from the Norwegian Sea that is joined by water that sinks in the Labrador Bay. Both areas are denoted as sources of heat to the atmosphere as a result of cooling of the downwelling water. This water then flows south at depth as intensified currents along the western side of the ocean basin to the South Atlantic, where it is joined by Antarctic Bottom Water that sinks in the Weddell Sea (another “heat source” due to cooling). The deep water then flows into the deep Indian Ocean and Pacific Ocean. There is upwelling to the surface in all oceans. In addition, a warm water return circulation in the thermocline layer (shown by the lighter lines) serves as a way to return water to the North Atlantic. There is also return flow of surface water to the Atlantic through the Drake Passage south of South America. After Intergovernmental Panel on Climate Change (IPCC) 2001.

The lateral deep-water circulation over the entire ocean, which was originally modeled by Stommel (1958), is shown in figure 1.14. (Deepwater circulation is not well documented and maps, such as that shown in figure 1.14, are based largely on theoretical models). North Atlantic Deep Water flows away southward from its source as an intense bottom current on the western side of the North Atlantic. This meets a strong northward flowing current of Antarctic Bottom Water in the South Atlantic, and as a result they merge and flow east

through the Antarctic into the deep Indian Ocean and ultimately into the deep Pacific Ocean (Gordon 1986, Warren 1981). Thus, the deep water originates only in two areas, and both are in the Atlantic Ocean. During the thermohaline circulation, deep waters remain out of contact with the atmosphere for long periods of time, which can result in appreciable change in their chemical composition (see chapter 8). The residence time of deep water, or average time it spends out of contact with the atmosphere, is about 200–500 years for the Atlantic and 1000–2000 years for the Pacific. As the deep water traverses the ocean floor, there is very slow, diffuse upwelling (at the rate of about 1 m/yr) of deep water into surface layers. This slow upwelling is supplemented by the more intense but localized coastal and open-ocean upwelling discussed in the previous section. In addition, there are two direct routes proposed to return water to feed North Atlantic Deep Water formation. One is the flow of cold surface water through the Drake Passage (south of South America) into the South Atlantic. The other is a proposed return flow of warm water toward the North Atlantic in the ocean’s thermocline layer. (Both routes are shown by lighter lines in fig. 1.14) There is considerable interest in the thermohaline circulation of the North Atlantic and rate of formation of NADW because of its climatic implications. Possible changes in NADW formation have been cited as a cause of rapid climate change over the last 18,000 years (e.g., Street-Perrott and Perrott 1990), and increased global temperatures from greenhouse warming (see chapter 2) may affect the rate of formation of NADW in the future (e.g., Gates et al. 1992 and Broecker 1987). Manabe et al. (1991) find that enhanced greenhouse warming may reduce the rate of NADW formation, thus reducing the surface water flow from the south and delaying the future greenhouse response in the northern North Atlantic. NADW formation is sensitive to changes in surface water temperature and salinity (the latter due to the addition of fresh water from ice melting [Street-Perrott and Perrott 1990]) and changes in ocean currents (Shaffer and Bendtsen 1994). Lozier et al. (2010) have recently

emphasized that the thermohaline circulation changes in response to natural variations in such things as winds and oceanic eddy circulation.

2 Air Chemistry: The Greenhouse Effect and the Ozone Hole In order to discuss some major global environmental problems and to more properly understand the chemistry of rainwater, it is necessary to delve into several aspects of air chemistry. Air consists of a mixture of gases and suspended particles, and the composition of this mixture has been perturbed in recent decades by human activities. This leads us to a discussion of such subjects as the atmospheric greenhouse effect, the ozone hole, and cooling and warming by the scattering of sunlight by global particle layers. In keeping with the general approach of this book, emphasis is on global or large-scale regional problems.

Atmospheric Gases Air consists of three gases—nitrogen, oxygen, and argon—that make up over 99.9% of the total volume. These major gases, along with a number of minor inert gases (helium, neon, krypton) because of their long residence times, occur in constant ratios to one another throughout the atmosphere, and these ratios stay constant over the human timescale. This is not true for the other gases that are impacted by changes in rates of input by natural processes, but more importantly by human processes that have brought about major perturbations in their concentrations, both locally and globally.

A major example is the rise of CO2 over the past century due to the burning of fossil fuels, a subject dealt with in detail later in this chapter. A listing of gas concentrations in air is shown in table 2.1. Because of their long residence times, the two major atmospheric gases N2 and O2 cannot be affected by anthropogenic activities on the human timescale. For instance, to remove all O2 from the atmosphere, it would take over 80,000 years of fossil fuel burning at the present rate. On a larger scale the natural processes of global photosynthesis (O2 production) and respiration (O2 consumption) are almost perfectly balanced (to within 0.4%). If human activities caused photosynthesis to cease and respiration continued at its present rate (a drastic and incredibly unrealistic scenario), it would still take over 8,000 years to consume all atmospheric O2. Similarly, if the production of N2 by global denitrification (see chapter 3) ceased tomorrow and consumption of N2 by nitrogen fixation (see also chapter 3) continued at its present rate, it would take more than 9,000,000 years to strip all N2 from the atmosphere. Gases of lesser abundance, but which are impacted anthropogenically, are carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), ammonia (NH3), ozone (O3), carbon monoxide (CO), sulfur dioxide (SO2), and nitrogen oxides (a combination of NO2 + NO represented as NOx). All of these gases have increased in atmospheric concentration as the result of human activities. Carbon dioxide, methane, and N2O are the most long lived (residence times on the scale of several years) so that their concentrations at any one time are relatively uniform over the globe. Carbon dioxide, methane, and nitrous oxide are contributors to the atmospheric greenhouse effect (see chapter 1), and because of this, emphasis on them is placed in this chapter. Because of their influence on the composition of rain, especially acid rain, SO2, NO2, and NH3 are discussed in chapter 3 in terms of the atmospheric cycles of sulfur and nitrogen. An important atmospheric gas, water vapor, is omitted from table 2.1. Its concentration is highly variable, both spatially and temporally,

and varies from a low of less than 0.01% to as much as 3.0%. Water vapor is omitted only because it is a topic of special consideration that is discussed extensively in chapter 1 as it affects the global energy cycle and in chapter 3 as it affects rain formation. Because it is a greenhouse gas it is also briefly discussed in the present chapter. Table 2.1 Concentration of gases in air at sea level

Gas

Volume percent in air

Nitrogen (N2)

78.084

Oxygen (O2)

20.942

Argon (Ar) Carbon dioxide (CO2)

0.934 0.039

Neon (Ne) Helium (He) Methane (CH4)

0.0018 0.0005 ≈0.0002

Sulfur dioxide (SO2)

0-0.0001

Krypton (Kr) Hydrogen (H2)

0.0001 ≈0.00005

Nitrous oxide (N2O)

≈0.00003

Carbon monoxide (CO) Nitrogen dioxide (NO2)

≈0.00001 0-0.000002

Ammonia (NH3)

≈0.000001

Ozone (O3)

0-0.000001

Source: modified from Turekian (1972) and Walker (1977), with CO2 updated to 2010 (Tans, P. 2010, www.esrl.noaa.gov/gmd/ccgg/trends); O2 updated to 2010; N2O 319 ppb (2005); CH4 1774 ppb (2005).

Carbon Dioxide Present and Future CO2 and the surficial Carbon Cycle Although carbon dioxide is the fourth most abundant gas in the atmosphere, it constitutes only some 0.039% (by volume) (see table 2.1). Atmospheric CO2 is important for two reasons: (1) it strongly absorbs infrared (long-wave) radiation given off by the Earth and reradiates energy back to the Earth, thus helping to maintain the Earth’s surface temperature (the so-called greenhouse effect—see chapter 1); and (2) it is a source of carbon, which is the dominant element in life and in the biogeochemical cycles of the Earth. Beginning in 1958, the atmospheric concentration of CO2 has been measured at the Mauna Loa Observatory in Hawaii (fig. 2.1). There is an obvious annual oscillation in atmospheric CO2 concentration of around 6 ppm which is a result of terrestrial biological cycling, with uptake of CO2 by plants during the spring and summer due to excess photosynthesis, and its release during the fall and winter due to excess respiration. (Because there is more land and thus more terrestrial biosphere in the Northern Hemisphere, seasonal cycles are larger there than in the Southern Hemisphere.) Superimposed on the annual Northern Hemisphere oscillation one can see from figure 2.1 that the yearly average value of CO2 has clearly increased from 315 ppm in 1958 to 389 ppm in 2010. The CO2 concentration has been rising faster in recent decades (1995– 2005, at the average rate of 1.9 ppm per year) than it did earlier (1.4 ppm per year average for 1960–2005) (Forster et al. 2007). The average rise on Mauna Loa from 2000 to 2009 was about 2.0 ppm per year. The CO2 concentration measured at many marine locations is in close agreement with the data from Mauna Loa. Figure 2.2 (Forster et al. 2007) emphasizes the links between the increase in atmospheric CO2, fossil fuel combustion, the drop in atmospheric O2, and the 13C/12C ratio of the atmosphere. Figure

2.2a shows the increase in the concentration of atmospheric CO2 in Hawaii and New Zealand. The rise in atmospheric CO2 has been attributed mainly to the burning of fossil fuels (coal and oil), which release CO2 to the atmosphere. A much smaller amount of CO2 comes from the production of cement. Estimates of contributions from fossil fuel combustion plus cement production from 1970 to 2005 is shown in figure 2.2b. The global emissions of CO2 accelerated sharply after 2000. The growth rate for the 1990s was around 1.1% y−1 compared to a 3.3% growth rate over the period 2000–2006. By 2008 fossil fuel emissions were 8.7±0.5 Gt Cy−1 (Le Quéré et al. 2009), an increase of 2% from 2007 and 29% from 2000 to 2008. The increase is due to growth in the world economy and more carbon use per unit energy use (what is called carbon intensity (Canadell et al. 2007a; Raupach et al. 2007). Because most of the fossil fuel emissions are in the Northern Hemisphere, there is a north-south atmospheric CO2 gradient that correlates with fossil fuel emissions. Along with the increases in fossil fuel CO2 emissions, there has been a shift in the largest fuel emission source from oil to coal. Coal contributed 40% of the CO2 emissions in 2008 versus 37% from 1990 to 2000. Oil dropped from 41% (from 1990 to 2000) to 36% in 2000 (Le Quéré et al. 2009). Almost all coal (93%) is burned for electric power generation, whereas most oil (72%) is used to make fuel for transportation (U.S. Energy Information Administration, 2009). Coal produces more CO2 per unit energy than oil.

Figure 2.1 Mean monthly concentration of atmospheric CO2 at Mauna Loa, Hawaii, 1958– 2010. The yearly oscillation is explained mainly by the annual cycle of photosynthesis and respiration of plants in the Northern Hemisphere. (Note: 1 ppm CO2 = 2 Gt C, where 1 Gt C = 109 tons C.) These measurements were started by C. D. Keeling of Scripps Institution of Oceanography in March 1958 at a NOAA facility. NOAA started its own measurements in May 1974, and they have run parallel with those made by Scripps since then. Pieter Tans 2010, NOAA/ESRL. www.esrl.noaa.gov/gmd/ccgg/trends/co2_data_mlo.html.

Anthropogenic carbon dioxide emissions also result from changing land use by humans, from deforestation and the burning or oxidation of organic carbon stored in plant material. From 2000 to 2008 net CO2 emissions from land use change were 1.4±0.7 Gt Cy−1 dominated by tropical forests. By 2008 emissions from land use change were nearly constant at 1.2 Gt Cy−1 (Le Quéré et al. 2009). Figure 2.2(a) also has a graph of the atmospheric O2 concentrations. O2 and CO2 are closely coupled due to photosynthesis, which removes atmospheric CO2 and generates O2.

In addition, the combustion of fossil fuel removes O2 from the atmosphere and adds CO2 to the atmosphere. Thus, the concentration of atmospheric O2 has been dropping as the concentration of atmospheric CO2 rises from fossil fuel combustion. However, Manning and Keeling (2006) showed that atmospheric O2 is decreasing at a faster rate than CO2 is increasing, which demonstrates the existence of an oceanic sink for fossil fuel CO2. Figure 2.2b has a graph of the 13C/12C ratio of atmospheric CO2, which has become more negative as fossil fuel emissions rise. This is because the 13C/12C isotope ratio in fossil fuel emissions is considerably lower than that in natural atmospheric CO2.

Figure 2.2 CO2 concentrations and emissions from 1970 to 2005. (a) Left: CO2 concentrations (monthly averages) from Mauna Loa, Hawaii (19° N; Keeling and Whorf 2005) and Baring Head, New Zealand (41° S; following techniques by Manning et al. 1997). Right: atmospheric O2 concentrations per meg (derived from deviation of O2/N2 from a standard on a parts-per-million basis) from Alert, Canada (82° N) and Cape Grim, Australia (Manning and Keeling 2006). (b) Left (dark line): Global annual emissions of CO2 from fossil fuel combustion and cement production in gigatons of carbon per year (Gt = 1015 tons), from 1970 through 2005, using data from the CDIAC website (Marland et al. 2006) to 2003. Emissions data for 2004 and 2005 are extrapolated from CDIAC, using data from the BP Statistical Review of World Energy (BP 2006). Right (light line): Annual averages of the 13C/12C ratio measured in atmospheric CO2 from Mauna Loa from 1981 to 2002 are shown (Keeling et al. 2005). The isotope data are expressed in del 13C(CO2) (per mil) deviation from a calibration standard.

After Forster et al. 2007, fig. 2.3a and b, p. 138. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Figure 2.3 summarizes the surficial carbon cycle as perturbed by humans mainly by the burning of fossil fuels. The natural surficial cycle consists of photosynthesis, respiration, gain and loss of carbon by soils, transport of carbon to the sea by rivers, and exchange of CO2 between the oceans and atmosphere. Using the mean 2000– 2008 values, fossil fuel combustion emissions were 7.7±0.5 Gt Cy−1 (Le Quéré 2009). Combined with land use emissions of 1.4 Gt Cy−1, the total CO2 emissions in 2008 were 9.1 Gt Cy−1. The fraction of human fossil fuel and cement production–derived CO2 that remain in the atmosphere is referred to as the airborne fraction of CO2. Between 1959 and 2008, 43% of each year’s total CO2 emissions (fossil fuel plus land use change) remained in the atmosphere on average (Le Quéré et al. 2009), but this fraction varies a great deal from year to year. In the last fifty years the airborne fraction has increased from 40 to 45% due to less CO2 uptake by carbon sinks.

Figure 2.3 The surficial Carbon Cycle. Reservoirs in 1015g C = 109t C = Gt C. Fluxes in Gt C/yr (dashed fluxes are yearly fluxes due to human activity; solid fluxes [arrows] natural). (Note: 1 Gt C = 1 Pg C.) For reservoirs, the first number is prehuman reservoir size, and the second number is the change in the reservoir size by humans. The value of +4.1 represents annual storage in the atmosphere. Anthropogenic yearly fluxes to and from reservoirs from Le Quéré (2009) are the mean for 2000–2008 (see also table 2.2.) Reservoir sizes, changes in reservoir, and natural fluxes are from Denman et al. (2007), fig.7.3, p. 515; for more details see the reference. Atmospheric reservoir is for 385 ppm CO2 (1 ppm CO2 = 2 Gt C). Marine C cycle gas exchange (natural and anthropogenic) and NPP (net primary productivity) from Bishop (2009). GPP = gross primary terrestrial productivity.

In the 1990s, for which the numbers are quite well known, the airborne fraction of fossil fuel CO2 is about 45% (see table 2.2). As noted by Takahashi (2004), from the beginning of the industrial age (about 1750 or 1800), when CO2 was 280 ppm, until 2004, when CO2 was 380 ppm, the increase in atmospheric CO2 was about 50% of that expected from estimates of fossil fuel CO2 production. (The rise of 1 ppm of atmospheric CO2 = 2.12 Gt C/yr). This means that about 50% of fossil fuel CO2 has been stored elsewhere—what has been called the “missing CO2.” (For the moment we will just consider

fossil fuel CO2 and consider the less well determined anthropogenic CO2 from land use changes later). Most of the recent growth of atmospheric CO2 is due to greater fossil fuel emissions (see table 2.2). There are two major carbon reservoirs that could take up the missing CO2 and which are both much larger than the atmosphere and exchange rapidly with it on a timescale of years to hundreds of years. These reservoirs are the oceans and the terrestrial biosphere plus soils. Together with the atmosphere, these reservoirs are involved in the surficial carbon cycle, which operates over a historical or human timescale (Berner 2004). This is illustrated in figure 2.3. Carbonate rocks and buried organic matter represent much larger carbon reservoirs that are important over geologic time, but because fluxes to and from them are so slow, they are not important on a human timescale. However, they are of interest to demonstrate how CO2 has varied over geologic time and are discussed in the next section on the long-term geological carbon cycle. The oceans represent the largest of the rapidly exchanging reservoirs, and most of the missing atmospheric fossil fuel CO2-C has probably been stored in them. Carbon is present in the oceans primarily as inorganic carbon in the form of dissolved bicarbonate ion (HCO3−), and carbonate ion (CO3−−). When atmospheric CO2 is added to ocean surface water, the following reaction occurs:

Thus, atmospheric CO2 is converted to HCO3− ion and stored in the oceans, and the ability of the oceans to take up CO2 depends on the availability of carbonate ion. The surface oceans (top 75 m) are well mixed with the atmosphere. Below this lies the thermocline (1000 m thick), where exchange with atmospheric CO2 is on a timescale of several decades, and below the thermocline lies the deep sea, which is

isolated from the atmosphere and mixes very slowly with it on a timescale of several hundred to a thousand years (Broecker et al. 1979; see also ocean structure in chapter 1). Thus, on a timescale of decades, only the surface ocean, and certain other regions extending down to 1000 m, takes up CO2. CO2 also dissolves in surface waters in the polar regions and is advected into abyssal waters during deep water formation. There is another means whereby excess carbon dioxide can be stored in the oceans. This is known as the biological pump (Sarmiento 1993a; Bishop 2009). CO2 is taken up in surface waters by photosynthesis to produce living planktonic organic matter. When the plankton die, a fraction of their remains is transported into deep water, where it decays back to CO2. This CO2 dissolves in the deep water, which is out of rapid exchange with the atmosphere. Thus, the biological pump results in a net transfer of CO2 from the atmosphere to deeper layers of the ocean. Some of this transfer may take place by intense biological production in coastal waters, with the organic carbon transported off the continental shelves into deeper waters. Since the production of organic matter in the oceans is generally limited by the availability of nutrients other than CO2 (see chapter 8), an increase in atmospheric CO2 does not necessarily mean an increase in the rate of downward biological transfer. To increase this rate and to try to accomodate excess CO2 produced by fossil fuel burning, it has been suggested that extra nutrients, specifically iron, be added to the Antarctic Ocean (Martin et al. 1990), as a corrective. However, this proposal has proven to be controversial. The effectiveness of the biological pump may be decreased by the greater ocean acidity that results from increased anthropogenic CO2 uptake. Particulate CaCO3 (particularly coccoliths) acts as a ballast to increase the sinking rate of particulate organic carbon. With less CaCO3 ballast, the conversion of more slowly sinking particulate organic carbon to dissolved inorganic C occurs in shallower water, which slows the biological pump and increases

surface CO2 concentrations. If CO2 is returned to the surface faster, the capacity of the oceans to take up more CO2 from the atmosphere is decreased. Greater acidity may or may not slow calcium-carbonate coccolith formation, so the overall effect of increased CO2 on the biological pump is not well established (Bishop 2009). Some atmospheric CO2 removed by marine organisms during photosynthesis is eventually deposited in bottom sediments. However, most of this CO2 is released fairly rapidly by organic matter decay, and only a very small amount is eventually buried and thereby permanently removed from the ocean-atmosphere system. Calculations indicate that sediment burial probably is not a major mechanism for the removal of anthropogenic CO2 (Berner 1982). The mean estimate of the global ocean sink for anthropogenic CO2 is 2.3±0.5 Gt Cy−1 for 2000 to 2008 (Le Quéré 2009). (See table 2.2.) Estimates of the ocean uptake are quite consistent based on a number of different methods with the uncertainty varying with the method. The methods used include oceanic uptake of CO2 based on the O2/N2 ratio (Manning and Keeling 2006); chlorofluorocarbon observations of water age and atmospheric CO2 history (McNeil et al. 2003); estimates of anthropogenic CO2 in the ocean and information about ocean circulation and mixing from ten ocean general circulation models (Mikaloff Fletcher et al. 2006); joint ocean-atmosphere inversions; surface ocean pCO2 and gas exchange (Takahashi et al. 2002); and changes in the 13C/12C ratio of seawater (Quay et al. 2003). Canadell et al. (2007a) estimated ocean uptake from 1959 to 2006 with an ocean general circulation model, coupled to a biogeochemical model forced by observed climate and CO2 concentrations, and got an average yearly oceanic CO2 uptake (2.2±0.4 Gt Cy−1) There is some concern that positive feedbacks may make the ocean a less efficient sink for anthropogenic CO2. As the surface

ocean dissolved CO2 increases with increasing atmospheric CO2, a drop in pH (acidification of the oceans) occurs as well as a drop in carbonate concentration. The pH of the ocean has been lowered since 1750 by about 0.1 (Denman et al. 2007); see also chapter 8. As the carbonate decreases, the buffering decreases and the ocean can absorb less atmospheric CO2. There are also effects of the acidification of the oceans on marine organisms. Acidification causes aragonite and calcite, which form shells for many marine organisms, to precipitate with greater difficulty. For example, reef-building corals are less able to produce shells when the ocean is more acid (and has a lower carbonate concentration), and there is increased dissolution of calcium carbonate on the ocean floor (Feely et al. 2004). Table 2.2 Mean annual anthropogenic carbon cycle fluxes (in Gt Cy-1) (Pg Cy-1)

Source: Le Quéré et al. (2009) for 2008, Le Quéré (2009) for 2000–2008, and Denman et al. (2007) for 1990–2000. Note: Emissions from fossil fuel and land use change are based on economic and deforestation statistics. Atmospheric CO2 is measured directly. Land and ocean sinks are estimated from observations from 1990–2000, and for 2000–2008, the ocean CO2 sink is from the average of several models and the land CO2 sink is to balance. Fluxes for 2008 were determined independently and don’t necessarily balance. aAn increase of 1.8 ppm. The total atmospheric CO concentration in 2008 was 385 2 ppm.

With a warmer climate the ocean circulation is likely to slow down which also slows vertical transport of surface water containing anthropogenic CO2 and its replacement by ocean water which has not yet contacted anthropogenic CO2. This also reduces oceanic uptake of anthropogenic CO2. How much anthropogenic CO2-C could eventually be absorbed by the ocean? Humans have already (1850–2006) released about 330 Gt C from fossil fuel and cement production (Canadell et al. 2007a). The IPCC business-as-usual scenario (Nakicenovic et al. 2000) projects a release of about 1600 Gt C from fossil fuels and deforestation by 2100. The maximum amount of fossil fuel carbon available for energy is thought to be about 5000 Gt C, dominantly coal but also oil (250 Gt C) and natural gas (200 Gt C), and this is expected to be released over several centuries (Rogner 1997). Some models predict that the terrestrial biosphere will no longer be a sink for CO2 after 2100 and it will release 1000 Gt C (Cox et al. 2000). Denman et al. (2007) estimate about 50% of the atmospheric CO2 increase will be removed by the oceans within 30 years, and a further 30% will be removed within a few centuries. Archer (2005), using a model of the ocean and seafloor carbon cycle, estimates the majority of the carbon will be taken up by the ocean in 300 years, when the atmosphere reaches equilibrium with the ocean, with the uptake dropping after that. He also estimates that in 1000 years, depending upon the size of the fossil fuel release, 67–83% of the fossil fuel will be in the oceans, with the rest in the atmosphere. According to Sabine et al. (2004), about 118 Gt C of the fossil fuel emissions from 1800–1994 is stored in the ocean, based on inorganic carbon measurements and a tracer-based separation technique. (This is about half of the estimated 244 Gt C released.) They estimate that 118 Gt C is about a third of the long-term potential ocean storage (which is then around 350 Gt C). In addition to the ocean, the other large carbon reservoir and possible CO2 sink is forests and the terrestrial biosphere (550 Gt C), plus soils and detritus (1500 Gt C). There is rapid annual exchange

between the terrestrial biosphere and soils and the atmosphere of approximately 100 Gt C/yr, and this causes the annual oscillation in atmospheric CO2 concentration mentioned previously. However, the terrestrial biosphere is both a source of CO2-C from land-use change or deforestation and a sink from the uptake of forest fuel emissions CO2. The size of the carbon source to the atmosphere, from land-use changes. including deforestation and burning of carbon, is poorly known. At present land-use change comes about mainly from the clearing of land in the tropics for agriculture, and it is this information which is not very reliable. These CO2 emissions are partly compensated for by CO2 uptake, by regrowth of secondary vegetation and rebuilding of soil CO2. Le Quéré et al. (2009) made a revised estimate of the net CO2 emissions from land-use change. The long-term average for net CO2 emissions from 1990 to 2005 is 1.5±0.7 Gt Cy−1 and is dominated by tropical deforestation. This number would amount to 20% of total anthropogenic CO2. In 2008 the estimate for land use change CO2 is 1.2 Gt Cy−1 and the fossil fuel burning CO2 flux is 8.7 Gt Cy−1 for a total anthropogenic flux of 9.9 Gt Cy−1 of which the land use change flux is 12%. Land-use change emissions are more nearly constant from year to year than fossil fuel emissions, which are constantly increasing. Most other estimates of land-use change (e.g., Canadell et al. 2007a are similar to these estimates. It is predicted that the land-use change source of CO2 will decrease in the future. Then, looking at table 2.2, for 2008 the total sources of anthropogenic C are fossil fuel emissions, 8.7 Gt C/yr, and land use change, 1.2 Gt C/yr, for a total of 9.9 Gt C/yr. The combined sinks of carbon are the atmosphere (3.9 Gt C/yr), the oceans (2.3 Gt C/yr), and a land sink of 4.7 Gt Cy−1 for a total sink of 10.9 Gt C/yr. Because the sources and sinks are determined independently, they don’t necessarily add up to zero because of the errors in the various

estimates. The “residual” refers to the difference between the sum of the various sources and sinks and zero. Le Quéré et al. (2009) believe that overall the residual is due to land-use change anomalies caused by “regional responses of terrestrial vegetation to climate variability,” which are not well captured by models. The tropical biomass is both a deforestation (land-use change) source and apparently in areas of undisturbed tropical forest also a considerable sink. According to Denman et al. (2007), the tropical live biomass sink could be 1.2±0.4 Gt C/yr, based on measurements in tropical forest plots and extrapolated worldwide. Lewis et al. (2009) estimate the carbon storage in tropical forests is 1.3 Gt C/yr from forest inventories and suggest that carbon storage is increasing in tropical forests due to increasing atmospheric CO2. The estimates of the Northern Hemisphere forest sink for C vary from 1.5 Gt C/yr (Stephens et al. 2007) for 1992–1996 to 1.7 Gt Cy−1 (Denman et al. 2007). There has been an increase in Northern Hemisphere forested land on former agricultural land and these forests are less than seventy years old and accumulating biomass. Adding the tropical biomass sink to the Northern Hemisphere sink would give a total terrestrial sink of 2.7 to 3.0 Gt C/yr. This is somewhat more than the estimated mean land sink for 2000–2008 of Le Quéré (2009) of 2.7 ± 1.0 Gt Cy−1. Stallard (1998) also estimated a terrestrial sink in sediment of artificial reservoirs of 0.6–1.5 Gt C/yr. It has also been suggested that the northern boreal forest is expanding northward due to global warming, which adds a further sink for anthropogenic CO2. The land biomass sink can vary from year to year with several factors: (1) Climatic conditions such as El Niño, a worldwide climate change driven by a change in an equatorial ocean current, cause a smaller biomass sink than average; this resulted in a rise in atmospheric CO2 of about 1.0 ppm (Sarmiento 1993b).

(2) Volcanic eruptions such as Mt. Pinatubo in 1991 caused cooler temperatures in the Northern Hemisphere, which resulted in less respiration and a greater net biomass sink and a 1.5 ppm drop in atmospheric CO2 in the two years following the eruption (Sarmiento 1993b). (3) Drought such as that in North America in 1998–2003 reduced the biomass CO2 uptake. These variations, added to the fact that observations of tropical biomass changes are sparse, makes estimating the land biomass sink more difficult. The amount of terrestrial biomass carbon storage can in general be affected by a number of processes (Canadell et al. 2007b): (1) CO2 fertilization refers to increased plant growth at higher concentrations of atmospheric CO2. For doubling of CO2, plant growth increases 10–25%, but only when there is an adequate supply of nutrients (especially nitrogen) and water. The present increase in CO2 should cause a few tenths of a percent change in plant growth, which is difficult to detect. CO2 fertilization is likely to be most important in the tropics (see Lewis et al. 2009, who suspect this effect) and in young forests. (However, there does seem to be some evidence of this effect in North American forests; see below.) (2) Atmospheric nitrogen deposition from air pollution is important in increasing plant growth in N-limited forests such as in Europe and the eastern United States. The source of the nitrogen is fossil fuel emissions from cars, biomass burning, and volatilization of nitrogen fertilizers. (3) Longer growing seasons at high latitudes cause more plant growth, which is partly cancelled out by restricted growth during hot, dry summers. (4) Soil respiration by microbes also increases with warming if there is enough water. Based on a carbon-climate model, Cox et al. (2000) predict that the terrestrial biosphere acts as a carbon sink

until about 2050 but turns into a source thereafter due to greater soil respiration from warming. (5) Natural savannah and forest fires, which release CO2, can be triggered by heat and drought. (6) Permafrost contains 900 Gt C, which could potentially be released by global warming–induced melting at high latitudes. Peatland soils contain 400 Gt C, which can be released by drying. The amplitude of the annual seasonal oscillation of atmospheric CO2 at Mauna Loa Observatory in Hawaii seems to be increasing (see fig. 2.1), which leads some to suspect that the size of the terrestrial biosphere may be growing in response to increasing atmospheric CO2; this would be an example of the “CO2 fertilization effect.” A growing biosphere would constitute increased CO2 storage. An increased oscillation amplitude occurred from the early 1970s until the 1990s, apparently due to North American carbon uptake by forests during the warm season. After an interruption from 1998 to 2003, when the size of the oscillation decreased due to a North American drought, it increased again in 2004 with returning rain in the United States (Buermann et al. 2007) However, droughts may worsen in general with global warming. In order to make projections of future rises in atmospheric CO2, and in order to estimate the size of the biosphere and fossil fuel contributions of CO2 over the past two centuries, it is necessary to know what the concentration of CO2 in the atmosphere was before human intervention. Measurements of CO2 concentrations of air bubbles trapped in buried Antarctic ice have shown that over the millennia preceding the pre–Industrial Revolution era, the concentration of CO2 ranged from about 275 to 285 ppm (Forster et al. 2007). The mean concentration of atmospheric CO2 was 280 ppm before the anthropogenic rise in CO2 started, and 280 ppm is considered to be the “pre-industrial value.” Since 1750 the concentration of CO2 has risen about 100 ppm, with most of the rise

taking place during the past century. Figure 2.4 shows that natural variations in CO2, at least over the past 1700 years, did not approach the human-impacted changes of the past century.

Fig. 2.4 Increases in concentrations of greenhouse gases since AD 0. Concentrations of CO2 and methane, which were relatively constant up until the 1700s, have increased steeply since then due to human activities. Nitrous oxide concentrations have risen since about 1750, with the steepest increases after 1950. After Forster et al. 2007, FAQ 2.1, fig. 1, p. 135. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Past CO2 levels Over the past million years, natural CO2 variations have been much larger than over the past 1700 years. For example, a study of air bubbles in deeply buried Antarctic ice found that from 650,000 to 430,000 yr B.P. the range of CO2 was between 180 and 260 ppm (Siegenthaler et al. 2005b). Petit et al. (1999) found that over the past 420,000 yr B.P. the range of CO2 concentration was between 180 and 280 ppm (fig. 2.5), a slightly larger range. Fischer et al. (1999) found that rapid atmospheric CO2 increase and temperature change occurred at the end of the last ice age, between around 18,000 years ago, when CO2 was about 200 ppm, and 11,000 years

ago, when CO2 had risen to 280 ppm. In addition, Siegenthaler et al. (2005b) found from deuterium (a proxy for temperature) records that between 650 and 390 ky B.P., CO2 change lagged temperature change by 1900 years. Fischer et al. (1999) concluded that CO2 concentration lagged Antarctic warmth by 600±400yr during the most recent glacial-interglacial transition. The lag of CO2 behind temperature has been suggested to have several possible causes (Fischer et al. 1999), which include:

Figure 2.5 Glacial-interglacial ice core data for the past 650,000 years derived from air bubbles trapped within buried Antarctic ice. Curves show variations of deuterium (delD) which is a proxy for local temperature, and the atmospheric concentrations of the greenhouse gases carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). The shaded bands indicate current and previous interglacial warm periods. After Solomon et al. 2007, Technical Summary, fig. TS1. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. (Fig. TS.1 is adapted from Fig. 6.3 using data from Petit et al. 1999, Indermuhle et al. 2000, EPICA community members 2004, Spahni et al. 2005, and Sigenthaler et al. 2005a, b).

(1) the ocean uptake of CO2 is greater during cold temperatures and lesser during warm temperatures (a positive feedback); (2) ocean mixing between the deep ocean and the atmosphere takes about 1000 years;

(3) the terrestrial biosphere, which stores carbon and controls its release to the atmosphere, waxes and wanes during temperature changes involved in glacial advances and retreats; (4) as glaciation increases, sea level drops, exposing organic carbon stored on the continental shelf to respiration which releases CO2. In addition, exposed continental shelf at lower latitudes would provide a place for terrestrial biomass expansion. The lag between changes in CO2 and temperature should not be interpreted as meaning that the CO2-induced atmospheric greenhouse effect is not important. Siegenthaler et al. (2005b) state that the estimated lags “are small compared to glacial-interglacial time scales and do not cast doubt on the strong coupling of CO2 and temperature or on the importance of CO2 as a key amplification factor of the large observed temperatures of glacial cycles.” The changes in CO2 amplify the warming that occurs due to changes in the distance between the Earth and sun because of changes in the Earth’s orbital parameters. In addition, the atmospheric concentration of CO2 over the past 650,000 years before the preindustrial era did not exceed 300 ppm. And coupling of the rate of change of CO2 and temperature (as measured by deuterium) did not change significantly during the same time interval. Thus, there appears to have been a constant correspondence between CO2 levels and Antarctic climate. Over even longer timescales of many millions of years, carbon dioxide concentrations have varied to a much greater degree. This variation is the result of natural perturbations of the long-term or geological carbon cycle (Berner 2004). This cycle involves the removal of carbon from the ocean-atmosphere-biosphere surficial cycle to rocks and the return from the rocks (fig. 2.6). One major process is the uptake of CO2 by the weathering of Ca and Mg silicates (see chapter 4) followed by the delivery of the resulting dissolved Ca, Mg, and bicarbonate to the sea where Mg exchanges

with basalt for Ca and where Ca and bicarbonate are precipitated on the seafloor as calcium carbonate (see chapter 8). The lost CO2 is returned to the atmosphere and oceans via the thermal breakdown of carbonates at depth and by degassing from volcanoes and smaller thermal springs and vents. The other major process of the geological carbon cycle is the burial in sediments of organic matter (representing the remains of dead organisms) and the oxidative weathering of old carbonaceous rocks exposed to the atmosphere by tectonic uplift and the thermal decomposition of deeply buried organic matter. Models for calculating past CO2 levels according to the long-term carbon cycle have been done by a variety of workers. Many factors affecting CO2 over time are considered in the modeling (see Berner 2004 for a summary). They include the effect on CO2 uptake by rock weathering resulting from past changes in land temperature, rainfall, continental size and position; exposure of rocks via erosion accompanying mountain uplift; and the evolution of land plants. Changes in the global rate of degassing of CO2 via volcanic and metamorphic activity are accounted for by following changes in plate tectonic activity. An important aspect of the modeling is the presence of negative feedbacks that stabilize CO2 against excessive variations that would otherwise result in frozen oceans or broiling lands due to the greenhouse effect. Two principal atmospheric greenhouse-related feedbacks are illustrated in figure 2.7. Increases in CO2 (e.g., by increased volcanism) are compensated for by uptake of CO2 during increased weathering due to rises in temperature, which in turn induce increased rainfall due to an accelerated global water cycle (see chapter 1). Both higher temperatures and increased rainfall accelerate the rate of rock weathering.

Figure 2.6 The geological or long-term (million-year) carbon cycle.

The results of the latest (2008) revision of the GEOCARBSULF model (Berner 2006) are shown in figure 2.8. The large drop in CO2 between 400 and 270 million years ago (the Devonian, Carboniferous, and Permian periods) can be explained by the rise of large land plants, mainly trees, that accelerated the uptake of CO2 due to both increased weathering of Ca and Mg silicate rocks and to increased global burial of organic matter. By secreting soil acids and increasing water recirculation, plants greatly accelerate rock weathering (see chapter 4). The burial of plant-derived organic matter, because much of it is relatively nonbiodegradable lignin, leads to the formation of coal and dispersed coaly carbonaceous matter in sedimentary rocks. The global burial of organic matter during the Carboniferous and Permian periods was the largest of all time, as evidenced by the great abundance of coal of this age.

Figure 2.7 System diagram showing negative feedback on atmospheric CO2 due to weathering. Arrows without concentric circles represent positive responses (e.g., if CO2 goes up, temperature goes up). Arrows with concentric circles represent negative responses (e.g., if weathering goes up, CO2 goes down). Complete loops with an odd number of concentric circles represent negative feedback. No circles or an even number represent positive feedback or reinforcement (latter not shown here).

Other Greenhouse Gases: Methane, Nitrous Oxide Measurements during the past several decades have shown that the concentrations of certain trace gases, specifically methane, nitrous oxide, and chlorofluorocarbons (CFCs), in the atmosphere have been increasing along with CO2 (see fig. 2.9), and these increases are also likely due to human activities. (CFCs, which are discussed later in this chapter with regard to the ozone hole, are entirely artificial in origin.) These gases are excellent absorbers of infrared radiation, in fact much stronger absorbers per unit of concentration than CO2, and thus serve to enhance the atmospheric greenhouse effect. Because of their high efficiency in absorbing long-wave radiation, the combined effect of methane and the other trace gases in raising the Earth’s surface temperature known as radiative forcing

is nearly equal to the effect of CO2. (This is shown in table 2.3.) Note that the ozone loss in the lower stratosphere, which is caused largely by CFCs, has resulted in net negative radiative forcing or cooling, which counterbalances part of the role of CFCs in causing global warming in the same period. This is because ozone is also a greenhouse gas.

Figure 2.8 Atmospheric CO2 vs time for the past 550 million years (Phanerozoic eon) (after Berner 2008). For method of calculation see Berner 2004.

Figure 2.9 Global average versus time of the concentrations of the major long-lived greenhouse gases, carbon dioxide, methane, nitrous oxide, and CFC-12 and CFC-11, from NOAA global flask sampling network from 1978 to 2009. After NOAA 2007. The NOAA Greenhouse Gas Index (AGGI). 2009. Fig.2. http://www.esrl.noaa.gov/gmd/aggi. Table 2.3 Extra trapping of infra-red radiation DQ (positive radiative forcing) by excesses of trace atmospheric gases above their pre-industrial concentration

Source: After Solomon et al. 2007, chap. 2 and table 2.1, p. 141.

a 2007 CO Raupach et al. 2007; rest 2005. 2 b From IPCC 2007; removal of CO involves a range of processes that can span long 2 time scales. c Ice core data; ice core range (up to 650,000 years ago) 0.32–0.79. d Stratospheric water vapor from methane oxidation. e In addition to CFCs include HCFCs (see Solomon et al. 2007 Table 2.1, p. 141). f Tropospheric O range of ΔQ = 0.25 to 0.65. 3 g Net negative forcing from stratospheric O destruction by CFCs 1750–2005. 3

The most important greenhouse gas after CO2 is methane, CH4, whose concentration has also been increasing over past decades The contribution of methane to the enhanced atmospheric greenhouse effect is about 16% of the total trapped-energy flux (table 2.3). Methane had a concentration in the atmosphere of about 1.77 parts per million in 2007, and studies of methane in air bubbles trapped in polar ice have shown that concentrations have more than doubled during the past 150 years (table 2.3). The sources of methane, summarized in table 2.4, are diverse. The largest individual source is from natural wetlands with a much smaller natural source from termites. Total anthropogenic sources are greater than total natural sources, with ruminant (cattle, etc.) digestive tracts being the largest, followed by a number of similarsized sources, including rice paddies, biomass burning, waste dumps, fossil fuel production (coal mining and natural gas), and industry. The total amount of CH4 emissions is fairly well known from measurements of atmospheric concentrations and estimates of removal rates. However, the exact size of each of the various methane sources is not well known (see table 2.4 for recent estimates). All processes, except fossil fuel production, involve the anoxic fermentation of carbohydrates, such as cellulose, by microorganisms living either in flooded soils or within the digestive tracts of animals (e.g., cattle, termites). Note that the estimates for 1990 and 1992 methane sources are smaller than those for 2006.

Figure 2.9 also shows an increase from 1990 to 2009 of 1000 ppm methane. Table 2.4 Sources and sinks for atmospheric methane TgCH4 y−1

Source: Model data: Wang et al. 2004; Bosquet et al. 2006; Mikaloff Fletcher et al. 2004 (table 1: process based; table 3: inverse source del13 C model); Watson IPCC 1990, 1992, process based. a Houweling et al. 2006. b Includes biofuel. c Includes industry. d IPCC 2001, revised by IPCC 2007. e coal mining plus natural gas.

There are also geologic sources of methane; it seeps out from such bacterial and geothermal sources in the Earth’s crust as organic matter-rich sediments, mud volcanoes, and geothermal emissions. Estimates of the size of the geologic source range from 40–60 Tg CH4 y−1(Etiope 2004), 45 Tg CH4 y−1 (Kvenvolden and Rogers 2005) to 18 Tg CH4 y−1 (Houweling et al. 2006). Recently, the emission of CH4 from living plants under aerobic conditions has been suggested based on lab measurements (Keppler et al. 2006; Keppler et al. 2008). The original larger estimate of the emissions was reduced to 10–60 Tg CH4 y−1 (Kirschbaum et al. 2006). Other researchers (Dueck et al. 2007; Beerling et al. 2008) have not been able to confirm Keppler’s results so far (Hopkin 2007), and more research is needed (Dueck and van der Wherf 2008). Thus, we have not included green plants as a methane source in the methane budget. Green plant methane production would replace some of the wetland output in the methane budget (Houweling et al. 2006). The best documented and largest sink for methane is atmospheric oxidation by the OH radical, mostly in the troposphere. Another sink is the oxidation of CH4 by microbiota living in soils (Whalen and Reeburgh 1992). A small amount of methane is also lost to the stratosphere. The atmospheric residence time of methane is about 8.9 years (Dlugokencky et al. 2003). Starting around 1982, the annual rate of growth of atmospheric methane decreased from 15 ppb (1% per year) to 0.2 ppb per year, the average for 1999 to 2006. (0.2 ppb equals 0.6 Tg CH4 y−1, using a conversion rate of 2.78 Tg CH4 per ppb.) This decreased growth is reflected in the curve of atmospheric CH4 concentrations, which flattens out from around 1999 to 2006 (See fig. 2.9). Why did the rate of increase in atmospheric methane level out? The implication is that the sources and sinks of atmospheric CH4 (see

table 2.4) were nearly balanced because the excess of emissions over sinks is the amount of methane that stays in the atmosphere. Since studies suggest that there was no significant change in the amount of OH in the atmosphere, and oxidation by OH is the main mechanism for methane removal, the total emissions of methane must have dropped (Solomon et al. 2007). This was likely caused by several factors, such as a drop in fossil fuel emissions, particularly a drop in natural gas emissions from the former Soviet Union combined with the short lifetime of atmospheric methane (8.9 years) (Dlugokencky et al. 2003). However, there is high interannual variability in the CH4 growth rate. For example, El Niño– ENSO produces warm and dry conditions and thus more biomass burning and increased wetland and rice growing emissions, resulting in an increase in atmospheric methane (particularly in the El Niño of 1998). Bosquet et al. (2006) found that from 1999 to 2006 atmospheric methane concentrations were quite stable. A rise in the anthropogenic methane emissions was balanced by a decrease in wetland emissions, which are very sensitive to temperature and water. Most models predict an increase in wetland emissions with global warming. According to Rigby et al. (2008), methane concentrations in the atmosphere rose again, beginning in 2007, for the first time in a decade in all monitoring locations. They evaluated the relative importance of an increase in the methane emission rate and a decrease in the concentration of hydroxyl radical, the biggest methane sink. Their measurements and modeling suggest that a small drop in hydroxyl concentration was accompanied by a methane rise in the Northern Hemisphere. The drop in hydroxyl concentration may be due to increased concentrations of carbon monoxide (CO), the main OH sink, because of more wildfires and changes in temperature, water vapor, and cloud cover. The 2007 methane emissions rise is most likely due to economic development in Asia and emissions from Arctic wetlands (NOAA 2007). The 2007 temperature was unusually high in Siberia, the site of many wetlands.

Methane hydrate is a highly concentrated solid form of CH4 but is stable only at high pressures and low temperatures. Methane hydrates are also climate sensitive, with methane being released by warming. There is probably a large mass of methane presently trapped as crystals of methane hydrate at several hundred meters’ depth in high latitude permafrost (Kvenvolden 1988; MacDonald 1990). If global warming continues, the methane hydrate should eventually be liberated, but because of its depth and the time it takes for a thermal signal to be conducted downward, the release should be delayed for several centuries. There are also methane hydrates in ocean sediments that occur at continental margins from 300 to 600 meters in depth and are a huge methane reservoir (500–2500 Gt C) (Reeburgh 2007). Oceanic methane hydrates could be released by an oceanic temperature rise. Ice core data (fig. 2.5) show that concentrations of CO2 and CH4 tend to rise during interglacial warm periods and fall in unison during glacial periods. However, present concentrations of methane are much greater than those found in the ice cores, which go back 650,000 years. The ice core concentrations are in the range of 400 to 600 ppb CH4 and never exceed 773 ppb (Spahni et al. 2005) versus a present concentration of about 1770 ppb, tending to confirm that the post-industrial rise in CH4 is not due to natural mechanisms (Solomon et al. 2007). The present atmospheric concentration is certainly unprecedented in the last 10,000 years. During the geologic past, over multimillionyear timescales, due to changes in the global abundance of wetlands, there have been large changes in the emission of methane to the atmosphere. This has led to changes in the concentration of atmospheric methane as large as several thousand ppb CH4 (Beerling et al. 2009). Nitrous oxide (N2O) is not only an important greenhouse gas but is also involved in the destruction of stratospheric ozone (see below in the section Ozone and the Ozone Hole). N2O is produced from various microbial N cycle transformations in the soil and ocean where N2O leaks off (Nevison et al. 2007). Natural sources include

denitrification in aerobic soils, dominantly in tropical regions but also in temperate forests. The atmospheric concentration of N2O has risen from about 270 ppb in 1750 during the preindustrial era to 319 ppb in 2005 (Denman et al. 2007) (table 2.3 and fig. 2.10a), with the steepest increases since 1950. It has continued to rise linearly by about 0.8 ppby−1 or 0.26%y−1 (Forster et al. 2007) for the last 30 years. Measurements of N2O from ice cores show that the atmospheric concentration of N2O varied by less than 10 ppb in the 11,500 years before the beginning of the industrial period (Solomon et al. 2007) and never was as high as it is now, only reaching 228 ppb (Spahni et al. 2005). Anthropogenic sources contribute nearly 40% of N2O emissions (table 2.5). Agricultural N cycle perturbations including N fixation to make fertilizer, cultivation and fertilization of farmland, land-use change, and raising livestock are the largest anthropogenic N2O sources. Several measures of agricultural N production are shown in figure 2.10b, including manure production and fertilizer production. From 1850 to 1950 the rise in manure production parallels the rise in N2O. After 1950 there is a sharp rise in fertilizer production, which is parallel to the rise in atmospheric N2O. In the twentieth century agricultural land has been expanded in addition to more intense land use (Denman et al. 2007). Davidson (2009) estimated from modeling that 2% of the manure nitrogen and 2.5% of the fertilizer N was converted to N2O between 1860 and 2005. The second largest anthropogenic N2O source is from rivers and estuaries (Kroeze et al. 2005) and coastal oceans (Naqvi et al. 2000; Nevison et al. 2004). An oversupply of N to coastal areas results in eutrophication with high surface water phytoplankton productivity and oxygen depletion in bottom waters from dead organic matter breakdown by bacteria. Anthropogenic nitrate runs off into coastal waters, where low bottom-water oxygen favors denitrification with N2O release. Increased coastal N2O production has been observed over a large area on the Indian continental shelf (Naqvi et al. 2000).

Figure 2.10(a) Changes in the emissions of fuel combustion NOx and atmospheric N2O mixing ratios from 1750 to 2005. (b) Several measures of agricultural N production since 1850: manure production, fertilizer production, and crop N fixation. After Denman et al. 2007, fig. 7.16, p. 545. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

There are a number of other smaller anthropogenic N2O sources (biomass burning, fossil fuel combustion, atmospheric deposition, and human excretia) (see table 2.5). N2O is removed from the

stratosphere by photochemical breakdown (photolysis). This process oxidizes N2O to NO, which then reacts with ozone to form NO2, destroying the ozone (see below in the section Ozone and the Ozone Hole). The lifetime of N2O is about 120 years (Nevison et a.l 2007), and it is the only long-lived tracer of human changes in the nitrogen cycle. It is assumed that 2% of anthropogenic N leaks off as N2O (Nevison et al. 2007). Other N gases such as NOx and NH3 are very short lived with atmospheric lifetimes of hours to days, not years. The atmospheric increase of N2O-N is about 3.5 Tg Ny−1, or an increase of 0.7 ppb y−1 in 1999–2001 (Hirsch et al. 2006). (Note that 1 ppb atmospheric N2O = 4.8 Tg N2O-N [Kroetze et al. 1999]). Table 2.5 Sources and sinks for nitrous oxide (N2O) for 1990s

Source: Sources from Denman et al. 2007; Davidson 2009 for 1860–2005. Sinks from *Denman et al. 2007; ** Hirsch et al. 2006, based on 0.7.ppb N2O increase; 1 ppb = 4.8 Tg N (Kroeze et al. 1999). a Davidson (2009) includes indirect emissions from downwind and downstream ecosystems including human sewage.

Other Greenhouse Gases: Halogens and Tropospheric Ozone Halocarbons, in addition to their ozone-destroying potential, are strong greenhouse gases, and their combined warming effect is 0.34Wm−2 (table 2.3). These gases, dominantly chlorofluorocarbons (or CFCs) and to a much lesser extent hydrofluorocarbons or (HCFCs), are entirely anthropogenic in origin and are being phased out (as ozone-destroying gases) under the Montreal Protocol. The phase-out began in 1991, and by 1996 the consumption and production of the three major CFCs ceased. As a result the atmospheric concentration of CFC 12, the most important gas, has leveled out, and the concentrations of CFC 11 and CFC 113 are now decreasing gradually (see fig. 2.9 for graphs of the concentrations of CFC 11 and CFC 12 from 1978 to 2009). These gases have very long lifetimes (100 years for CFC 12) and therefore decline very slowly (see also table 2.3). Other halogens (such as HCFCs) are also strong greenhouse gases and will not be phased out until 2030. Ozone is an important greenhouse gas. Changes in ozone concentration, both in the lower stratosphere and upper troposphere can change the radiative forcing of the troposphere-surface system (Solomon et al. 2007). Increases in ozone concentrations in the upper troposphere cause positive radiative forcing (warming) of the troposphere-surface system of 0.35 Wm−2 (table 2.3). On the other hand, decreases in lower stratospheric ozone by destruction by CFCs from 1750 to 2005 results in cooling of - 0.05 Wm−2 (see table 2.3). The main discussion of stratospheric ozone destruction and the ozone hole and tropospheric ozone as a pollutant is later in this

chapter. The combined effect of the lower stratospheric and upper tropospheric ozone changes is net warming of 0.30 Wm−2.

Radiative Forcing by Anthropogenic Factors The climate of the Earth is determined by incoming solar energy, which is modified by the reflection, absorption, and emission of energy by the Earth’s atmosphere and surface (see chapter 1). The properties of the atmosphere and surface have been changed by humans in such a way that the Earth’s energy budget has been altered, which can also change global climate. These anthropogenic changes include increases in greenhouse gases that absorb outgoing Earth radiation, changes in surface reflectivity, and increases in aerosols (very small atmospheric particles and droplets), which reflect and absorb incoming solar radiation and change cloud properties. (Aerosols will be further discussed later in this chapter.) These changes cause radiative forcing of the climate system. Radiative forcing measures the influence of a factor in changing the balance of incoming and outgoing energy in the Earthatmosphere system and is also an indication of the potential of a factor in causing climate change (Solomon et al. 2007). Radiative forcing is expressed as Wm−2 per °C, where W = Watts. Positive radiative forcing increases the temperature and negative radiative forcing cools the Earth. The combined radiative forcing of all the long-lived greenhouse gases amounts to 2.63±0.26 Wm−2 and is the dominant and bestknown effect in producing global warming along with ozone (+0.3 Wm−2) as is summarized in figure 2.11. Details of forcing by each gas are shown earlier in table 2.3. Direct aerosol effects come from reflection or absorption of incoming solar radiation by anthropogenic aerosols, including sulfate aerosols, fossil-fuel organic carbon aerosols, fossil-fuel black carbon aerosols, nitrate aerosols, and mineral dust (see table 2.8 in section on aerosols). Anthropogenic aerosols also cause an indirect cloud

albedo effect (reflection) by nucleating water clouds; larger concentrations of aerosols lead to increased albedo of clouds. In addition, there is other (mainly negative) anthropogenic radiative forcing (or cooling) that is less well determined. Changes in the land cover, such as cutting down forest and replacing it with agricultural land, increases the surface albedo (or reflectivity of the surface). This is accentuated by the presence of snow cover on agricultural land, a very reflective surface, particularly as compared to dark trees (Forster et al. 2007). Other changes in land use can affect the local climate through shifts in cloudiness, surface roughness, surface temperature, and water balance. The deposition of black carbon aerosol from incomplete combustion of fossil fuel or biomass burning on snow cover reduces the high surface albedo of the snow. A positive radiative forcing of 0.1±0.1 Wm−2 results from this effect (Forster et al. 2007). Surface albedo effects (land use changes and black carbon on snow) and aerosol effects collectively amount to about –1.1 Wm−2(see fig. 2.11). The total net anthropogenic forcing is 1.6 (0.6–2.4) Wm−2. This value is determined by combining the positive (warming) from longlived greenhouse gases plus ozone with the negative forcing (cooling) from aerosol plus land albedo and cloud albedo (see below) effects. The best estimates and uncertainty ranges cannot be obtained by adding the various terms directly due to assymetrical uncertainty ranges for certain factors; the best estimate and uncertainty ranges are, therefore, obtained from a Monte Carlo technique (Solomon et al. 2007). From the net positive mean anthropogenic global radiation forcing of 1.6 Wm−2 one might expect an approximately linear increase in global mean surface temperature (Solomon et al. 2007). However, the spatial warming on the Earth’s surface will vary because of various climate effects. For example, the high latitude response is greater due to sea-ice albedo feedbacks. Also, the oceans have a greater thermal inertia and thus respond more slowly than do land

areas. Aerosol forcings vary considerably due to the greater concentrations of aerosols in the Northern Hemisphere and are harder to model. In addition to anthropogenic radiative forcing, there has been a natural change in solar irradiance of 0.12 (0.06–0.30) Wm−2 since 1750 (Solomon et al. 2007; see fig. 2.11). This is due to reevaluation of the long term-change in solar irradiance since 1610 (Mauder Minimum). There are also episodic natural changes in solar irradiance due to volcanic eruptions, such as Mount Pinatubo in 1991 and Krakatau in 1883, which inject reflective sulfate particles into the lower stratosphere where they remain for several years, causing a decrease in solar irradiance of - 3 Wm−2 and cooling the Earth by 1–2° C (Forster et al. 2007).

Climatic Effects of Radiative Forcing: Climate Sensitivity, Global Warming, and Hydrologic Changes The term equilibrium climate sensitivity, which equals “the equilibrium global average warming expected if CO2 concentrations were to be sustained at double their preindustrial values (about 550 ppm)” is used by Solomon et al. (2007). The CO2 concentrations are CO2-equivalent concentrations, which means concentrations of greenhouse gases, etc., equivalent in radiative forcing to a certain amount of CO2. According to Solomon et al. (2007), the best estimate of climate sensitivity is about 3°C (2°–4.5°C). And, the climate sensitivity is very unlikely to be below 1.5°C. This estimate is based on atmosphere-ocean general circulation models (AOGCMs) constrained by observations.

Figure 2.11 Global Mean Radiative Forcing. (a) Global mean radiative forcing from 1750 to 2005. The term radiative forcing (RF) refers to the absorption of long-wave radiation (in Watts per m2) by the various greenhouse gases, aerosols, and other mechanisms. Positive radiative forcing causes global warming while negative radiative forcing causes cooling. Best estimates and uncertainty ranges cannot be obtained by direct addition of individual terms due to asymmetric uncertainty ranges for some factors; the values are obtained by Monte Carlo techniques. LOSU means level of scientific understanding. In addition, the forcing for black carbon aerosols (top of the atmosphere) is +0.9 W m-2 (Ramanathan and Feng 2009). Black carbon added here is not included in total net anthropogenic forcing. After Solomon et al. 2007, Technical Summary, fig. TS5, p.32. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. Note: The IPCC (2007) climate sensitivity is about 1.25 Wm-2 per °C (Ramanathan and Feng 2009). (b) Probability distribution of global mean combined forcing from all anthropogenic agents in (a). Three cases are shown: total of all anthropogenic RF terms (filled dark-gray curve); long-lived greenhouse gases and ozone RF’s only (dashed light-gray curve); direct aerosol and cloud albedo RF’s only (dashed black curve). Surface albedo, contrails and stratospheric water vapor RF’s are included in the total curve but not the others. After Forster et al. 2007, fig. 2.20B, p. 203. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Water vapor is the most important greenhouse gas. Although the concentration of upper tropospheric water vapor is relatively low, it makes a very important contribution to the “natural” greenhouse effect. The concentration of atmospheric water vapor increases as the Earth warms, resulting in a positive radiative feedback (the water vapor feedback) and enhanced warming. In fact, water vapor provides the largest positive radiative feedback; alone, it doubles the

warming due to increases in greenhouse gases. However, in the tropics the change in the lapse rate partially offsets the temperature increase, so the combined water vapor/lapse rate feedback increases the estimated climate sensitivity from greenhouse forcing by 50%. (The lapse rate is the rate at which temperature decreases with altitude.) This means that the effective temperature rise for doubling CO2 would be about 4.0°C (Randall et al. 2007). There are several other feedbacks that amplify the warming: (1) Increased surface temperatures cause melting and decreased areal extent of surface ice and snow at high latitudes, which reduces the Earth’s albedo—the amount of solar radiation reflected from the Earth (see chapter 1). Ice and snow are much stronger reflectors of solar radiation than vegetation or bare ground, so that increased absorption of solar radiation accompanying a lower albedo further increases surface temperature. (2) The warming of soils, especially at high latitudes (Schlesinger 1991; Kvenvolden 1988) should cause greater microbiological activity in the soils with the release of extra CO2 by decay.

Other feedbacks may be important and need to be taken into consideration. The effects of clouds is complex. The net effect of high-level clouds is to reduce the terrestrial radiation leaving the top of the atmosphere (warming effect), while low-level clouds reflect more entering solar radiation (cooling effect). For 1980 to 1999 the global annual mean radiative forcing due to clouds was -22.3 Wm−2 (i.e., cooling) (Meehl et al. 2007). The cloud feedback from global warming—i.e., changes in cloud coverage, altitude, and water content (and therefore reflectivity)—is a source of uncertainty in climate models. One study (Clement et al. 2009) suggests a reduction in low-level cloud cover over much of the Pacific Ocean when greenhouse gases were increased, i.e., a positive low-level cloud feedback. Differences between models in climate sensitivity— i.e., the amount of temperature rise for doubling of CO2—is primarily due to the differences in representing cloud feedbacks, particularly those due to low clouds.

Observed Changes in Temperature and Atmospheric Circulation The global average surface temperature of the Earth has increased especially since 1950. From 1850–99 to 2000–2005 the temperature has warmed 0.76°C±0.19°C (see fig. 2.12). Note that the linear trends of temperature are shown for different time intervals, with the last 25-year interval having a much steeper slope (an increase in temperature of 0.177°C/decade) as compared to the 50-year interval where the temperature increase is only 0.128°C/decade. The rate of temperature increase is accelerating over the approximately 150 years shown on the graph. The temperature increase varies with location. The surface temperature over the land has warmed at a greater rate (0.27°C/decade) than that over the ocean (0.13°C/decade). In addition, the warming is greatest at high northern latitudes (Solomon et al. 2007). For example, in the area north of 65°N temperatures have increased by about twice the average global temperature increase from 1965 to 2005 (Lemke et al. 2007). The two warmest years of temperature record are 1998 and 2005.

Figure 2.12 Annual global mean surface temperature (black dots) with linear fits to the data. The left-hand axis shows temperature anomalies relative to the 1961–90 average, and the right-hand axis shows estimated actual temperatures, both in °C. Linear trends are shown for the last 25, 50, 100, and 150 years (see figure legend for period symbols). The smooth pale-gray curve shows decadal variations, with the decadal 90% error range shown as a

pale gray band surrounding that line. The total temperature increase from the period 1850– 99 to the period 2001–2005 is 0.76°C ± 0.19°C. (FAQ 3.1, fig.1. Climate Change 2007: The Physical Science Basis). After Solomon et al. 2007, Technical Summary, fig. TS.6, p. 37. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Changes in temperature extremes might be expected from global warming. The most obvious change is the decrease in the number of cold nights for 1951–2003 for all areas. In addition, heat waves have lasted longer in the last half of the twentieth century. In the summer of 2003, western and central Europe had a heat wave that was the warmest since 1780. The increase in surface temperature has been accompanied by a drop in stratospheric temperature of 0.3°C to 0.6°C per decade since 1979. This drop has been interrupted by periodic rises in temperature following volcanic eruptions. Long-term changes in the large-scale atmospheric circulation have occurred, such as stronger midlatitude westerly winds in both hemispheres with a poleward shift of the corresponding Atlantic and south polar front jet streams. In the Northern Hemisphere stronger westerlies change the flow from the oceans to the continents and are an important factor in changes in winter storm tracks and resulting patterns of precipitation and temperature trends at mid-and high latitudes. The increased westerlies reflect a change in the preferred patterns of atmospheric circulation variability in the North Atlantic, either a change in the North Atlantic Oscillation (NAO) or in the North Annular Mode (NAM). The North Atlantic Oscillation is a measure of the strength of the Icelandic Low and Azores High and of the westerly winds between them. The North Annular Mode is a winter fluctuation in the amplitude of a pattern of low surface pressure in the Arctic and strong midlatitude westerlies. In a positive phase of the NAO and NAM when the atmospheric pressure is stronger over the Central Atlantic, the strong westerlies push warmth and precipitation toward northern Europe. It has been suggested that the trend and increased variance in NAO/NAM from 1968 to 1997 was greater than might be

expected from internal variability and might be related to global warming (Solomon et al. 2007). In the Pacific Ocean, ENSO (El Niño Southern Oscillation) dominates ocean-atmosphere change. The 1976–77 climate shift related to the phase change in the Pacific Decade Oscillation (PDO), which is a measure of sea surface temperatures in the North Pacific, caused an increasing number of El Niño events. El Niño results in a pool of warm water in the western Pacific off Peru. Over North America ENSO changes have led to the western part of North America warming more than the eastern part, which has become cloudier and wetter (Solomon et al. 2007). In addition, the 1998 mean temperature was the highest on record until 2005. (1997–98 was an El Niño year). El Niño also causes extremes of hydrologic cycle—floods and drought. There may be a link between the observed changes and global warming.

Observed Changes in the Water Cycle: Water Vapor, Precipitation, Streamflow, and Storms Water vapor in the troposphere has been increasing at 1.2±0.3%/decade from 1988 to 2004 (Solomon et al. 2007). This is what would be expected from rising temperature. Because water vapor contributes to the greenhouse effect, an increase in water vapor provides an important feedback to global warming. Willett et al. (2007) found that specific humidity, which is the ratio of water vapor to air in a particular volume of air, has increased significantly during the late twentieth century, mainly due to human influence. They used observations of specific humidity along with a coupled climate model to determine this. This is important because greater humidity tends to cause more precipitation, more intense precipitation, and more intense tropical cyclones. Overall, increased temperature is predicted to result in a small increase in global mean precipitation according to models (Emori and Brown 2005; Zhang et al. 2007). Models also predict greater precipitation at high latitudes and less in the subtropics with an areal redistribution in the tropics. However, global anthropogenic-caused

precipitation changes are difficult to detect because changes in precipitation in different areas tend to cancel each other out. Precipitation changes tend to vary more with the season and location than do warming changes. Zhang et al. (2007) used models to predict anthropogenic forcing of land precipitation from 1925 to 1999, then compared model results with actual observations. They believe that human forcing contributed to observed increases in land precipitation in midlatitudes of the Northern Hemisphere and drought in the subtropics and tropics of the Northern Hemisphere, plus wetter Southern Hemisphere subtropics and tropics. Global annual mean land precipitation anomalies that have been observed are shown in figure 2.13. These changes are not unidirectional with time, showing an overall increase from 1900 until the mid 1950s, a decrease until the early 1990s, and then a rise thereafter. Changes in global precipitation are difficult to interpret because different areas have large anomalies of different signs. Models have been used to predict the global pattern of streamflow changes in the twentieth century from global changes in temperature and precipitation. The results have been compared with observations (Milly et al. 2005). Areas observed and predicted to have less river runoff are sub-Saharan Africa, southern Europe, southern South America, southern Australia and western midlatitude North America. Runoff has increased in the LaPlata basin of southern South America and in southeastern, central, and far northeastern North America; southeastern Africa; northern Australia; and southern Eurasia. Climate models tend to predict a drier climate in the southwestern United States in the twenty-first century that should already be under way (Seager et al. 2007). Barnett et al. (2008) show that the mountain precipitation and hydrology in the arid western United States changed from 1950 to 1999. Using hydrologic and global climate models, they show that 60% of the climate-related trends in the hydrology, such as more winter rain rather than snow, earlier snow melt, increased spring river flow, lower summer river flow, and a warming in the area, are due to human-induced changes from

greenhouse gases and aerosols. Together these changes will result in future water shortages in this very dry area.

Figure 2.13 Time series for 1900 to 2005 of global annual mean land precipitation anomalies from GHCN data set for a number of other different data sets (see original for particulars). After Trenberth et al. 2007, fig. 3.12, p. 254. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

Barnett and Pierce (2008) also point out that when the water from the Colorado River was divided among the western states in 1922, the river was having relatively high annual flows. The river flow has been declining for a number of years, and Lake Mead and Lake Powell are at 50% of capacity. If the predicted drop of 10–30% in the river flow in the next thirty to fifty years due to global warming actually occurs, and particularly if a natural drought occurs, there is a 50% chance the Colorado River water stored in Lakes Mead and Powell could be so low that it could not flow downstream by gravity within the next ten years or so without changes in water allocations. There has been an increase in the number of stronger tropical cyclones (hurricanes) in the North Atlantic since 1970, which correlates with increased tropical sea surface temperatures. These hurricanes have been more destructive, lasting longer and having greater intensity. Variations in the total number of tropical cyclones

and in their tracks come from ENSO and decadal variability. However, the number of North Atlantic hurricanes has also increased from 1995 to 2005. Records are much better with the use of satellites since 1970. There are more anthropogenic aerosols. More aerosols, by producing more cloud condensation nuclei and more cloud droplets, increase the cloud lifetime. This is because smaller droplets do not rain out as easily causing drizzle suppression, increased cloud liquid water content, and increased cloud height. This cloud lifetime effect alters the hydrologic cycle. Also, aerosols absorb solar radiation, heating the troposphere and causing cloud burn-off.

Observed Changes in Ice, Sea Level, and the Oceans Ice, snow, and frozen ground on the Earth’s surface have melted due to global warming. Table 2.6 summarizes the major contributions of ice loss to sea level changes. In addition, changes in snow and ice have a strong influence on the Earth’s surface because of (1) their high reflectivity (albedo) compared to bare ground or trees, which cool the surface; (2) changes from ice to water involve the uptake of latent heat; (3) glacier ice and snow are an important source of freshwater in many areas (see previous discussion of the Colorado River). Since the temperature in high latitudes has increased at about twice the rate of the average temperature rise from 1965 to 2005 (Lemke et al. 2007), the Greenland and Antarctic ice sheets have been subjected to greater than average global warming. These ice sheets are the main reservoir capable of changing sea level drastically. If they melted completely, sea level would rise 64 m, of which Greenland contributes 7 and Antarctica 57 m (Lemke et al. 2007). Ice formed from snow falling on the interior moves under gravity toward the coast, where it either melts or calves off into the ocean as icebergs. In the past it had been assumed that spreading velocity of the ice does not change rapidly, so the main effects of global warming would be on the snowfall rate, which is accelerated by the warming of the oceans, and on the surface melting rate of the

ice, which is also accelerated. However, rapid increase in ice-flow velocities have been observed in Greenland in summer. Additionally, it has been observed that floating ice shelves act as a brake on the flow of glaciers that feed them, which speeds up upon the breakup of the ice shelves. Both these factors have caused concerns about more rapid loss of ice from the ice sheets (Alley et al. 2005; Lemke et al. 2007). Table 2.6 Contributions of ice loss and thermal expansion to observed sea level change

Recent satellite observations of the western ablation zone of the Greenland ice sheet (van de Wal et al. 2008) suggest that meltwater lubricates the bedrock and speeds up flow at the ice sheet margins in the summer, but the yearly ice movement is actually fairly constant or even dropping in some places and not rapid as had been feared. Thus, because the positive feedback between melting and ice velocity is seasonal, it may not greatly speed up the ice melting from global warming in the next decades at least. At present the authors cannot conclude that it is important.

Venke Sundal et al. (2011) found that melt-induced speedup of the Greenland ice sheet flow was offset by efficient subglacial drainage, and in warm years summer ice flow is slower. The ice sheet was not flowing from basal lubrication alone, but rather, like mountain glaciers, subglacial drainage is necessary. An improved understanding of subglacial drainage is critical to understanding the effect of melting changes on ice sheet velocity. Thus, the response of the ice sheet to climate change and its contribution to sea level change remain uncertain. There is a second impact from the melting of the Greenland and Antarctic ice sheets. The locations of ocean deepwater formation are located near Greenland, where sinking of cold, salty surface water to the bottom leaves a void that draws warm surface currents north, setting up the North Atlantic conveyor belt which warms northern Europe. (See chapter 1 for a discussion of ocean circulation.) The sinking is dependent partly on the salinity (and thus density) of the surface water, so that an influx of freshwater might make the water less dense and stop the sinking, which would turn off the Atlantic conveyor belt. This would greatly impact ocean heat transport to northern Europe and its climate (Vellinga and Wood 2002; Alley 2004). Deepwater formation also occurs near Antarctica. Global warming is causing a gradual rise in sea level through several processes (see table 2.6): (1) transfer of water to the ocean by slow partial melting of ice and snow in the Greenland and Antarctic ice sheets (sea level rise of 0.4 mm/yr), along with mountain glaciers and ice caps such as those in Alaska, continental northwestern United States, southwestern Canada, and Patagonia (rise of 0.8 mm/yr); and (2) expansion of ocean water due to heating (sea level rise of 1.6 mm/yr). The total sea level rise from known global warming causes is 2.8 mm/yr for 1993–2003 (Solomon et al. 2007). Satellite measurements indicate an actual rise of 3.1 mm/yr. Sea level rose 1– 2 mm/yr over the past century, with 0.5±0.2 mm/yr from seawater expansion and the rest mainly from land glacier melting (Alley et al. 2005). The rise in sea level from 1870 to 2000 is shown in figure 2.14. This rise should have inundated some very low coastal lowlands. Real problems occur when sea level rise combines

with river floods and/or tropical storms, making the combined affect devastating (e.g., Bangaladesh or Hurricane Katrina in New Orleans). In the geologic past ice sheet shrinkage occurred much more rapidly than ice sheet growth. For example, 21,000 years ago at the end of the last glacial maximum, when glacial melting occurred due to global warming from changes in the Earth’s orbital parameters and natural changes in greenhouse gases, sea level rose on average 10 mm/yr. There were even two episodes at 19,000 years ago and 14,500 years ago when rates may have been 50 mm/yr (Alley et al. 2005). In addition, global sea level was 4–6 m higher during the last interglacial 125,000 years ago (Jansen and Overpeck 2007). A 2.7%/decade reduction in the average annual areal extent of sea ice in the Arctic Ocean has occurred since 1978, although no trend in Antarctic sea ice has been observed (Solomon et al. 2007). This change greatly affects albedo but does not affect sea level, however.

Fig 2.14 Global Mean Sea Level rise (mm) from 1870 to 2000. Error bars show 90% confidence intervals. After Bindoff et al. 2007, fig 5.13, p. 410. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

There has been a decrease in Northern Hemisphere land snow cover observed in satellite observations since 1966, particularly in spring. In addition the mountain snowpack has declined in the western United States with implications for less freshwater supply. Permafrost and seasonally frozen ground have declined since the 1980s in the Canadian Arctic, Alaska, Siberia, Tibet, and Europe. The oceans are warming. The global ocean average temperature rose from 1961 to 2003 by 0.10°C from the surface to a depth of 700 m. The subpolar ocean has become less saline, and the shallower parts of the tropical and subtropical oceans have become saltier. The Pacific Ocean is fresher and the Atlantic and Indian Oceans are more saline (Bindoff et al. 2007).

Predictions for Future Climate Change Warming of average surface temperature of 0.1°C per decade is predicted for the next two decades. This assumes that there will not be any volcanic eruptions, which cool the Earth, or solar changes (Solomon et al. 2007). This is accompanied by a sea level rise over the next several decades of 1.3±0.7 mm/yr due to thermal expansion alone (but augmented by melting of glaciers and the Greenland ice sheet). Projections are made of the equilibrium climate sensitivity, which is the equilibrium global average warming expected for CO2 concentrations of double their pre-industrial values (or CO2 = 550 ppm). The equilibrium climate sensitivity is about 3°C with a range of 2.0°–4.5°C but is very unlikely to be less than 1.5°C (Solomon et al. 2007). Warming reduces the amount of CO2 taken up by the land and ocean, increasing the amount of anthropogenic emissions in the atmosphere. The greatest uncertainty in model predictions is the effect of clouds (particularly low clouds). The IPCC (2007) climate sensitivity is about 1.25 Wm−2 per °C (Ramanathan and Feng 2009). IPCC estimates of temperature change for a doubling of CO2 are independently confirmed by the analysis of the effect on climate of

varying CO2 over geologic time. Comparison of CO2 values obtained from modeling the temperature dependence of the longterm carbon cycle with hundreds of independent proxy estimates over the past 420 million years produces a best fit between the two sets of data when using a temperature change of 2.8°C for a doubling of CO2 (Royer et al. 2007). This study further found that temperature responses of less than 1.5°C for CO2 doubling were highly unlikely in the geologic past. The global mean temperature increase to 2090–99, relative to 1980–99, made by a number of models, is 1.8°–4.0°C, with an average of 2.8°C and most models giving around 2.4°–2.8°C. The twenty-first-century temperature change is greatest over the land and at high latitudes in the Northern Hemisphere during the winter, increases from the coast to continental interior, and is larger in arid regions because evaporation of water vapor cools the air. The modeled average sea level rise that would accompany this warming is an average of 0.34 m with a range of 0.2–0.5 m. This is 70–75% due to thermal expansion of the ocean, with the rest due to ice melting. Antarctica is too cold for much melting, so it comes primarily from glaciers and ice caps and the Greenland ice sheet. As noted earlier, there is evidence for greater land ice flow when ice shelves are removed and disagreement over how much enhanced production of summer ice-melt water contributes to annual ice-flow velocity. If there is greater ice melt and sea level increased linearly with increasing surface temperature, this could add 0.1–0.2 m to the upper limit of sea level rise (Solomon et al. 2007). Models also predict that the North Atlantic oceanic conveyor belt which was described previously (and is sometimes referred to as the meridional overturning circulation, or MOC) will slow down over the twenty-first century due to less dense high-latitude surface water (which will be warmer and fresher due to more precipitation and ice melting). However, it is very unlikely that the conveyor belt will abruptly stop in the twenty-first century (Solomon et al. 2007). Predictions suggest more extreme weather events. Heat waves are expected to be more intense and more frequent, and to last

longer, particularly in the tropics. Higher sea surface temperatures should produce more frequent and more intense tropical storms such as hurricanes and typhoons. Models suggest reduced snow cover and thawing of permafrost to greater depths. Precipitation patterns should change to more precipitation at high latitudes (>50°) due to more atmospheric water vapor and a resulting increase in water vapor transport from low latitudes poleward, particularly in the Northern Hemisphere. This pattern is accompanied by decreased precipitation in the subtropics. Southern and Southeast Asia are expected to have more rainfall in the summer monsoon (Solomon et al. 2007). Increased atmospheric CO2 leads to more acid surface oceans, and over the last two hundred years 40% of the anthropogenic CO2 has ended up in the oceans, decreasing their pH and making the precipitation of carbonate minerals such as calcite and aragonite more difficult (see chapter 8). Since these minerals are the main constituent of many marine shells and skeletons and make up coral reefs, ocean acidification is harmful to many marine organisms (Zeebe et al. 2008). A decrease in 0.2–0.3 pH units inhibits calcification in many marine organisms, including corals, foraminifera, and some calcareous plankton. The key variable determining future seawater chemistry changes are the size and timescale of anthropogenic CO2 release. Climate change from already emitted anthropogenic greenhouse gases depends largely on their lifetime in the atmosphere. For example, methane has a short lifetime (decades or less), whereas N2O will be in the atmosphere for around a century. The removal time for CO2 varies considerably, but the transfer of CO2 from the atmosphere and terrestrial reservoirs to the oceans is slow and can take thousands of years for its removal from the atmosphere. And although methane is short lived, its increased release from melted permafrost could raise its future concentration. Even if the emission of greenhouse gases ceased in 2100, it is expected that the temperature would rise about 0.5°–0.6°C within the following century

and thermal expansion of the oceans would continue for two hundred years, causing another 0.3–0.8 m of sea level rise by 2300 (with respect to 1980–99) (Solomon et al. 2007). In addition, the Greenland ice sheet would continue to melt with continued elevated temperatures.

Aerosols Aerosols are small particles of solid or liquid ranging in size from a few molecules to about 20 µm in radius. Here the role of aerosols in reflecting solar radiation is emphasized; their role in the formation of acid rain and in the cycles of sulfur and nitrogen is discussed in chapter 3. Direct radiative forcing by aerosols occurs because they absorb or reflect short-wave (solar) or long-wave (Earth) radiation, thus altering the radiative balance of the Earth-atmosphere system. Depending on their properties, some aerosols cool the Earth and others warm it. The main sources of particles or aerosols in the atmosphere are given in table 2.7. Although the particle fluxes are not well known and represent approximate estimates, it is still apparent that the main particle sources are natural: mostly soil dust, and sea salt. Anthropogenic aerosols include industrial aerosols, made up of sulfate, organic carbon, and black carbon from fossil fuel burning, as well as nitrate and mineral dust; they also include aerosols from biomass burning. Total anthropogenic radiative forcing from all aerosols amounts to - 1.2 Wm−2 (Solomon et al. 2007; see table 2.8). Ramanathan and Feng (2009) get total radiative forcing from anthropogenic aerosols of - 1.4 Wm−2. Because the total radiative forcing from anthropogenic aerosols is negative (i.e., cooling) and is similar in magnitude to greenhouse gas warming, it tends to counterbalance it (see fig. 2.11 and fig. 2.16, which compares the radiative forcing from various sources). For an excellent discussion of the development of our knowledge about greenhouse warming and the effect of atmospheric aerosols, see Ramanathan and Feng (2009).

Table 2.7 Fluxes of Particles smaller than 20 µm radius emitted into or formed in the atmosphere (in Tg/y; Tg = 1012g)

Aerosol Cloud Effects There are a number of effects of aerosols on clouds (these are summarized in fig. 2.15; see also Lohmann and Feitcher 2005). The indirect “cloud albedo effect” with water clouds formed by anthropogenic aerosols is also referred to in IPCC (2001) as the first indirect effect, or Twomy effect (Forster et al. 2007; see fig. 2.15). Larger concentrations of aerosols tend to lead to increased cloud albedo. This is because more aerosols produce more cloud condensation nuclei (CCN) and a larger number of cloud droplets, which increases cloud reflection of solar radiation. The indirect cloud albedo effect is estimated to have a radiative forcing of - 0.7 Wm−2 (- 0.3 to - 1.8).

Figure 2.15 Radiative mechanisms associated with cloud effects from aerosols. The small black dots represent aerosol particles; the larger open circles cloud droplets. Straight lines represent the incident and reflected solar radiation, and wavy lines represent terrestrial radiation. CDNC means cloud droplet number concentration. The unperturbed cloud contains larger cloud drops as only natural aerosols are available as cloud condensation nuclei, while the perturbed cloud contains a greater number of smaller cloud drops, as both natural and anthropogenic aerosols are available as cloud condensation nuclei (CCN). The vertical gray dashes represent rainfall, and LWC refers to the liquid water content. After Forster et al. 2007, fig. 2.10, p. 154. In Climate Change 2007. The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press.

There is a second indirect effect (IPCC 2001) of aerosols on clouds referred to in Forster et al. (2007) as the cloud lifetime effect, or Albrecht effect (Albrecht 1989). This induced effect by aerosols on cloud lifetimes is not considered to be radiative forcing because by suppressing drizzle, increasing cloud height, or cloud lifetime, the hydrological cycle and particularly precipitation is changed. There is also a “semidirect effect of aerosols on clouds,” which is due to the absorption of short-wave radiation by tropospheric aerosols and which heats the troposphere, causing “cloud burn-off” and shrinking of the cloud (Ackerman et al. 2000b).

Types of Aerosols There are two main types of particles or aerosols in the atmosphere: primary particles emitted directly into the atmosphere (such as anthropogenic particles from burning, windblown dust, sea salt, plant

fragments, etc.) and secondary particles formed from gaseous emissions that subsequently condense in the atmosphere. Gas-toparticle conversion results in the formation of fine particles (1 µm). In chemical composition, aerosols may consist of one or more fractions: (1) water-soluble ions (such as sulfate, nitrate, ammonium, and several seasalt-derived ions); (2) a mostly insoluble inorganic part (silicates, oxides, etc.); and (3) a carbonaceous part (soluble and insoluble organic matter). In form, aerosols range from dry dust particles to sea-salt particles that are sometimes drops of salty water (at a high relative humidity). Most continental aerosols are a mixture of soluble and insoluble components (mixed particles), whereas most marine aerosols are soluble, consisting of sulfate from oceanically produced reduced sulfur gases and sea salt (Junge 1963; Fitzgerald 1991).

Gaseous Emissions Fine secondary aerosols formed from gases emitted to the atmosphere include sulfuric acid formed from fossil fuel burning SO2, volcanic sulfur gases, and biogenic dimethyl sulfide (DMS); (NH4)2 SO4 and NH4 HSO4 formed by the reaction of NH3 with sulfuric acid aerosols; nitric acid formed from emissions of nitrogenous gases; and organic aerosols formed from the oxidation of biogenic gases. These secondary aerosols are important both to the thermal budget of the Earth and to the atmospheric cycles of sulfur and nitrogen.

Sulfate Aerosols Sulfate aerosols dominate the anthropogenic aerosol radiative forcing (see table 2.8), producing 80% of the net cooling that occurs. These are fine secondary aerosols formed from anthropogenic SO2 gas emitted into the atmosphere, where it is converted into sulfate acid aerosols. Because sulfate aerosols are water soluble

(hygroscopic), they convert to sulfuric acid at low relative humidities and are in the size range of 0.1–1 µm, thus making good CCN (cloud condensation nuclei). The sulfuric acid aerosols are partly or totally neutralized by ammonia and can either be liquid or partly crystallized. The major source of anthropogenic sulfate aerosols is SO2 from fossil fuel burning, with biomass burning contributing a much smaller amount (Forster et al. 2007). According to Stern (2005), there has been a decrease in the total global SO2 emissions from 73 to 55 Tg Sy−1 in the interval 1980–2000. There has also been a shift in SO2 emissions from the United States, Europe, Russia, and the North Atlantic Ocean to Southeast Asia and the Indian and Pacific Oceans over time. This will change the location of the resulting sulfate aerosols. Anthropogenic sulfate aerosols mainly cause scattering or reflection of sunlight and are estimated to have a radiative forcing of - 0.4±0.2 Wm−2 (Forster et al. 2007; Haywood and Boucher 2000), cooling at the Earth’s surface and in the atmosphere (Ramanathan 2001). Anthropogenic sources of sulfate make up 65% of the total sulfate, 63% from fossil fuel burning and 2% from biomass burning (see table 3.7; Haywood and Boucher 2000). There are several natural sources of sulfate aerosol as well: DMS sulfate 12% and volcanic sulfur 10%. Marine phytoplankton release DMS (dimethylsulfide), which oxidizes in the air to form SO2, part of which is converted to sulfate aerosol. As we have already mentioned, sulfate aerosol can increase the number of CCN (cloud condensation nuclei) and therefore presumably the number of cloud droplets, increasing the cloud albedo and cooling the Earth. Remote marine stratus clouds, which cover 25% of the oceans, are particularly sensitive to this increase in albedo. The amount of natural sulfate aerosol from DMS is estimated in 2000 to be around 20 Tg Sy−1 (Brimblecombe 2003; and see chapter 3). The DMS flux from the ocean would increase by 3% for the warming from a

doubling of CO2. This increased flux would produce increased sulfate aerosols and increased cloud droplets, resulting in a change of cloud albedo of - 0.05 Wm−2 (Bopp et al. 2004). Volcanic debris accounts for a small part of total global particle formation. but the production of volcanic material tends to be episodic and, therefore, dramatic right after major eruptions. For example, the eruption of Krakatau in the East Indies in 1883 ejected an estimated 25,000 Tg of material into the atmosphere, or around three hundred times the estimated normal yearly production of volcanic material (Goldberg 1971). The airborne particles produced by volcanoes consist of finely divided ash made up of silicate minerals and sulfuric acid aerosols, the latter originating from the oxidation of SO2 from the volcanic plume. The sulfuric acid aerosols can reach the stratosphere during unusually violent eruptions, and because of a much longer residence time for particles in the stratosphere due to a lack of rainout, the sulfuric aerosols can bring about appreciable global cooling for a few years until they fall out. This was the case for Krakatau (1883), El Chichón (1982), and more recently for the eruption of Mount Pinatubo (1991). Hansen et al. (1992) calculate a mean global surface temperature change of - 0.5°C as a result of the input of aerosols from Mount Pinatubo. An additional effect of sulfuric acid particles in the stratosphere is to participate in reactions that destroy ozone (Solomon et al. 1993; Tolbert 1994; and see section Ozone and the Ozone Hole below). Volcanic aerosols may also absorb solar radiation, warming parts of the stratosphere and changing stratospheric wind patterns (Kerr 1993).

Black Carbon Aerosols Black carbon in soot is a primary aerosol that comes from incomplete combustion in fossil fuel or biomass burning. Black carbon refers to the absorbing components of soot, which are elemental carbon and condensed organics. The total global emission of black carbon in 2000 was estimated by Bond et al. (2007) to be

about 8 Tg Cy−1. The sources of this black carbon are fossil fuel— diesel fuel (1.9 Tg Cy−1) and coal (1.0 Tg Cy−1)—and biofuel, including animal waste, charcoal, agricultural residues and wood, (1.5Tg Cy−1). There was also 3.4 Tg Cy−1 produced by open biomass burning (savannah burning, crop residues, and forest fires) (Bond et al. 2004). Black carbon in the atmosphere absorbs solar radiation reflected by the surface-atmosphere-cloud system. It also absorbs direct solar radiation, thus preventing it from reaching the Earth’s surface. These two processes cause warming in the lower atmosphere that amounts to 2.6 Wm−2 (see fig. 2.16c). However, since black carbon reduces the radiation reaching the surface, it cools the surface by - 1.7 Wm−2. The combined effect of the atmospheric warming (2.6 Wm−2) and surface cooling (- 1.7 Wm−2), means that, overall, black carbon warms the Earth by 0.9 Wm−2 (fig. 2.16c) at the top of the atmosphere (Ramanathan and Carmichael 2008; Ramanathan and Feng 2009). The black carbon forcing of 0.9 Wm−2 is the second largest cause of global warming after CO2, which produces 1.6 Wm−2. This is shown in figure 2.16 (and see fig. 2.11. and table 2.3). Hansen and Nazerenko (2004) get 0.8 Wm−2, including black carbon on snow of 0.3 Wm−2 (indirect effect).

Figure 2.16 Comparison of the global mean forcing due to greenhouse gases (GHGs) with that of atmospheric brown clouds (ABCs); (+) warming; (-) cooling. The number at the top of the atmosphere box (top box) is the top-of-the-atmosphere (TOA) forcing; the number within the atmosphere box is the atmosphere forcing; and the number within the lower box is the forcing at the surface. The TOA forcing is the sum of the forcing of the atmosphere and the surface. The forcing values represent the change in radiative forcing due to increase in gases for the year 2005, which is the same as the forcing from preindustrial to the present. The TOA (top of the atmosphere numbers) are taken from Solomon et al. (2007), and the atmosphere and surface numbers are derived from an atmospheric transfer model run by Ramanathan and Feng (2009), using the Chung et al. (2005) analysis. The uncertainty in forcing is ±20% for a and b and ±50% for c, d, and e. (a) Forcing for all greenhouse gases (GHGs: CO2, CH4, N2O, halons, and ozone); (b) Forcing for CO2; (c) BC (black carbon) aerosol direct forcing obtained by running the Chung et al. (2005) aerosol model with BC (black carbon) aerosol and (d) with non-BC (nonblack carbon) aerosol (direct and indirect forcing); (e) forcing for all anthropogenic brown cloud aerosols (ABCs), which includes both BC aerosols and non-BC aerosols. Adapted from Ramanathan and Feng (2009, fig. 8, p. 44), modified from Ramanathan and Carmichael (2008, fig. 2).

Atmospheric brown clouds (ABCs) are made up of black carbon aerosols, plus other aerosols such as sulfate, nitrate, organics, dust, etc. They occur in India, East China, central Africa, Mexico, Central America, Brazil, and Peru. Black carbon in atmospheric brown clouds absorbs radiation and warms (0.9 Wm−2), and non-black carbon aerosols (sulfate, organic, and other aerosols) reflect solar radiation and cool (– 2.3 Wm−2); thus, the total top of the atmosphere forcing from all atmospheric brown cloud aerosols (black carbon and non-black carbon) is shown in figure 2.16 as – 1.4

Wm−2, thus cooling the Earth. Without aerosols the surface would be warmer by 1.3°C. (Ramanathan and Carmichael 2008; Ramanathan and Feng 2009).

Organic Carbon Aerosols Organic carbon aerosols are also produced by fossil fuel burning, in addition to biomass burning and from natural biogenic emissions. Organic aerosols are emitted as primary aerosols or formed as secondary aerosols from the condensation of organic gases (VOCs, or volatile organic carbon). The total anthropogenic organic carbon emissions are estimated at 33.9 Tg C. Open biomass burning produces 74% of anthropogenic organic carbon aerosols at 25.1 Tg Cy−1 (Bond et al. 2004). According to Bond (2007), the total amount of organic carbon aerosol produced in 2000 from biofuel and fossil fuel was 8.7 Tg Cy−1. The sources of this aerosol are biofuel (6.1 Tg Cy−1), coal (1.3 Tg Cy−1), and liquid fuel (diesel and lighter fuel) (1.3 Tg Cy−1). The radiative forcing from anthropogenic organic aerosols, which are either entirely scattering or only weakly absorbing, is a cooling of - 0.05±0.05 Wm−2(Solomon et al. 2007). Natural biogenic organic material (mainly hydrocarbons) is directly emitted to the atmosphere by vegetation or produced from volatile organic carbons (VOCs) as a secondary aerosol. Natural continental primary organic carbon emissions may be large but are not well known (Denman et al. 2007). Kanakidou et al. (2005) estimate natural biogenic secondary organic aerosol production of 30 Tgy−1. The oceans are also a source of biogenic organic matter, especially during plankton blooms (O’Dowd et al. 2004).

Biomass Burning Aerosols Biomass burning aerosols are composed of a mixture of organic carbon, black carbon, and inorganic compounds like nitrate and sulfate. Biomass burning emissions are uncontrolled and variable.

Biomass burning is seasonal in southern Africa and in South America and includes savannah burning, crop residue burning, and forest burning. Studies have been made of pyrogenic and biogenic burning aerosols in southern Africa. The radiative forcing of biomass burning aerosols in clear skies is negative but when the aerosols are lifted over the clouds, the reflection of solar radiation is reduced over cloudy areas and the radiative forcing of biomass burning is positive. Overall, the radiative forcing of biomass burning aerosols is estimated to be +0.03±0.12 Wm−2 by Forster et al. (2007).

Nitrate Aerosols Ammonium nitrate aerosols form if sulfate is fully neutralized and there is excess ammonium. Thus, the direct RF due to nitrate aerosols is dependent upon the atmospheric concentrations of ammonium as well as NOx concentrations. The present and future decline in emissions of SO2, which forms sulfate aerosols in industrial areas of Europe and North America, will make nitrate aerosols more important in these areas (Feng and Penner 2007). Because of its sources (see chapter 3), nitrate aerosols have higher concentrations in industrial areas than in rural areas. There has been increased production of NOx due to increased fossil fuel combustion (including automobiles), increased fertilizer use, and intensification of agriculture. Nitrate and ammonium aerosols are highly hygroscopic and can absorb water to form aqueous solutions under typical atmospheric conditions. Nitrate is also taken up by dust aerosols. The radiative forcing of nitrate aerosols is estimated to be -0.10±0.10 Wm−2 (Forster et al. 2007).

Mineral Dust Aerosols On a global scale, most of the mineral dust comes from the Northern Hemisphere “dust belt” (Prospero et al. 2002), which includes the west coast of North Africa, the Middle East, central and southern

Asia, northwestern India (Indes River valley), and China’s loess area. This area is around 15°–30°N latitude, where the atmospheric circulation produces sinking dry air, which is warmed at the surface and causes strong evaporation, forming the subtropical belt of deserts (see chapter 1). There is little dust produced in the Southern Hemisphere. The dust sources are associated with arid areas with rainfall less than 20–25 cm, in topographic lows adjacent to a topographic high, with ephemeral streams and playas or salt flats. These areas usually have been previously flooded and have deep fine-alluvial deposits that can be transported by the wind. The Bodele depression in the Lake Chad region of Africa is the strongest dust source in the world. Lake Chad, which is at the south end of the area was much larger six or eight thousand years ago. Dust from this area is responsible for the huge tropical North Atlantic dust plume in winter, which transports dust to South America. This area is relatively uninhabited. Soil or mineral dust is assumed by Forster et al. (2007) to have an anthropogenic component of only 0–20% maximum. The radiative forcing from this anthropogenic dust is estimated as -0.1±0.2 Wm−2. Some anthropogenic sources of soil dust in the United States are farming, overgrazing, unpaved roads, cement production, and construction. Tegen et al. (2004) estimated that agricultural dust is 6.0, even reaching 6.4 in northern Utah. In eastern U.S. rain, where abundant CaCO3 dust is not available, H+ concentrations vary with SO4−−. Neutralization of natural acidity in unpolluted rain over land can also take place by reaction with ammonia gas (NH3). In regions where ammonia is emitted to the atmosphere from biological decay,

agricultural activity, etc., there may be enough to bring about a slight rise in pH via the reaction NH3 + H+ → NH4+ For example, Charlson and Rodhe (1982) calculate that, in the absence of sulfate aerosol, NH3, at the lowest concentrations found in continental areas (0.13 µg/m3), could raise the pH of CO2−containing rain from 5.7 to 6.2. By contrast, in the presence of sulfate aerosol, there is usually insufficient NH3 to effectively neutralize acidity. In the remote north Pacific Ocean (Quinn et al. 1990) the ocean emission fluxes of NH3−N and DMS-S are in a mole ratio of 1.2:1, which is lower than the 2:1 ratio needed to neutralize all of the H2SO4 acidity derived from the DMS. Also, when pollutive NH3 and H2SO4 are involved, there is normally insufficient ammonia for complete neutralization. The relative importance of NH3 and soil dust in neutralizing acidity varies from area to area, but on the average in the United States about one-third of the acid neutralization is due to NH3 (Munger and Eisenreich 1983). However, because NH3 gas or fine aerosol NH4 can travel further than coarse soil dust containing CaCO3, the neutralization effect of NH3 may affect areas farther from the source. In Beijing, China, where there was a lot of CaCO3 dust and NH3 along with a high sulfate concentration (4.39 mg/L), the pH was 6.2 rather than about 3.5 estimated if there were no neutralization (Galloway et al. 1987). One would expect such relatively high pH rains in other arid areas where there are alkaline soils with greater NH3 release and more CaCO3 dust. (See Rodhe et al. 2002). Over the oceans, sea-salt aerosol, which is alkaline (from bicarbonate and borate), may neutralize acids in marine precipitation to a small extent. When the concentration of sea-salt aerosol in rain

is large (Na > 3.0 mg/L) it is possible to raise the original pH by about 0.05 pH units (Galloway, Knap, and Church 1983; Pszenny, MacIntyre, and Duce 1982).

Acid Rain from Pollution Acid rain is defined here as that having a pH less than 5.7 due to reactions with acidic gases other than CO2. The acidic gases are SO2, NO2, NOx, and (to a lesser extent) HCl, and they result in the formation in the atmosphere and in rain clouds of sulfuric, nitric, and hydrochloric acids respectively. (See also sections on sulfate and nitrate in rain.) Overall reactions are SO2 + OH → … H2SO4 (sulfuric acid) SO2 + H2O2 → H2SO4 (sulfuric acid) NO2 + OH → HNO3 (nitric acid) HClgas → HCl (hydrochloric acid) followed by the dissociation of these acids in rainwater to form H+ ions: H2SO4 → 2H+ + SO4−− HNO3 → H+ + NO3− HCl → H+ + Cl Thus, as more and more of the precursor gases are added to the atmosphere by human activities, more and more hydrogen ions are produced and the pH of rainwater drops. This helps explain why the pH of rainfall at many locations has been decreasing over time.

Acid Rain in Europe Acid rain was first noted in northwestern Europe in the early 1950s. Barrett and Brodin (1955) found that precipitation in southern Sweden had a pH between 4 and 5 and that the pH was lowest in the winter when the air flow is from the south, bringing pollution from central and western Europe. A region of high acidity (pH 4.0 to 4.5) that was centered in the Benelux countries in the late 1950s had spread by the late 1960s to Germany, northern France, the eastern British Isles, and southern Scandinavia. By 1974 most of northwestern Europe was receiving acid precipitation (pH