Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 Restructuring the Climate Change Science Program S ocietiesâ ability to respond to climate change depends in part on the magnitude and speed of changes in the climate system and on the resilience of human and environmental systems in the face of these changes. Air and ocean temperatures are increas- ing, resulting in widespread melting of snow and ice and rising sea levels. This global warming has been occurring over the past cen- tury, but has greatly accelerated in the past few decades, driven by the addition of greenhouse gases, especially CO2, to the atmos- phere at an ever increasing rate. A warming in excess of 3ÂºC is possible (cf., Figure 2.1) and could push components of the climate system past various tipping points (e.g., Schneider and Mastran- drea, 2005; Lenton et al., 2008), including the possible loss of the major ice sheets and glaciers. The bell-shaped curve of the warm- ing with a wide range of 1.5ÂºC to 4.5ÂºC and a âfat tailâ shown in Figure 2.1 illustrates the large uncertainty in our understanding of the response of the climate system to human perturbation. It also suggests that we cannot entirely dismiss the possibility of irre- versible changes in the way Earthâs climate operates and how human and ecological systems respond. 21
22 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.1 Probability distribution of the predicted increase in global mean surface temperature due to a 3 Wm-2 radiative forcing from in- creases in greenhouse gases from preindustrial times to 2005. The probability density of the expected warming adopts the IPCC (2007a) climate sensitivity of 3Â°C warming due to a doubling of CO2, with a 90 percent confidence level of 2Â°C to 4.5Â°C warming. The realized warming is the warming from 1750 to 2005 that has been attributed to greenhouse forcing. Because of the small amount of warming that has been realized to date and the presence of strong cooling by aerosols, temperature in- creases above 2Â°C are likely not imminent but could be very large before the end of the century. The temperature thresholds for various climate tipping points are marked by the blue words. The ranges, taken from Len- ton et al. (2008), are not shown, but are 0.5Â°C to 2Â°C for the melting of Arctic summer sea ice; 1Â°C to 2Â°C for radical shrinkage of the Greenland Ice Sheet and 3Â°C to 5Â°C for shrinkage of the West Antarctic Ice Sheet; 3Â°C to 4Â°C for the dieback of the Amazon rain forest due to drastic reduc- tions in precipitation; 3Â°C to 6Â°C for persistent El NiÃ±o conditions; and 3Â°C to 5Â°C for a shutoff in the North Atlantic deep water formation and the associated thermohaline circulation. The tipping point of Himalayan- Tibetan glaciers is based on the IPCC (2007a) finding that these glaciers may suffer drastic melting when warming exceeds 1Â°C to 2Â°C above pre- industrial levels. SOURCE: Ramanathan and Feng (2008).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 23 What measures society should, can, and will apply to slow the growth of greenhouse gases and/or reduce the dangers posed by the expected large climate system changes are still far from settled. Changes in greenhouse gas emissions reflect behavioral patterns, energy consumption, population growth, and societal responses to climate change. These changes are happening in the context of complex socioecological systems in which nature and society are mutually dependent and are constantly affecting one another (posi- tively and negatively) across space and time (Folke, 2006). The fundamental dilemma faced by policy makers is how to forge ef- fective strategies both to mitigate further climate change and to adapt to the changes already under way, in view of the uncertain- ties in our knowledge about how climate affects humans and vice versa and of the political difficulties of taking costly action now for benefits that accrue in the future. The fat tails of the distribu- tion of climate sensitivity (Figure 2.1), rather than the average, may drive the economic trade-offs associated climate change (Weitzman, 2009). Policy and decision makers must have better information that meets their needs (NRC, 2009). Improving understanding of the interactions and feedbacks of the physical climate system with human and environmental sys- tems, improving predictions of longer-term causes and trends, and preparing the nation for future climate changes are grand chal- lenges. They are particularly difficult to tackle if we do not understand the system as a whole. Under the Climate Change Sci- ence Program (CCSP), much has been learned about components of the natural climate system, including the composition of the at- mosphere, the water and carbon cycles, and changes in the land surface (NRC, 2007c). It is now time to take a more holistic ap- proach and integrate across natural and social science disciplines and across the science and policy worlds to find solutions to cli- mate change-related problems that are of major concern to society. This chapter provides seven examples of societal issues that motivate the need for an integrated approach to the research pro- gram. Two are current issues stemming from changes in the climate system (weather and climate extremes, sea level rise and melting ice) and five focus on impacts of climate change (avail- ability of freshwater, agriculture and food security, managing ecosystems, human health, and impacts on the economy of the
24 RESTRUCTURING FEDERAL CLIMATE RESEARCH United States). The examples connect societal issues widely rec- ognized as essential to the well-being of the planet with high- priority science and application needs. Although not a comprehen- sive list, they show how the CCSP could be organized to yield both improved understanding of the climate system and the knowledge foundation needed to support sound decision making. EXTREME WEATHER AND CLIMATE EVENTS AND DISASTERS Extreme (severe) weather and climate events are the most visible manifestations of climate-related hazard. In the worst cases, such extreme events interact with socioeconomic, political, and ecological factors (e.g., food and water supply) to create economic or health disasters (Wisner et al., 2004). Especially at risk are the poor, uneducated, very old or very young, and the sick. How soci- ety deals with extreme weather events today provides an analog for understanding our vulnerability to hazard in a changing climate (Adger et al., 2003). The impact of climate-related hazard depends on two factors: (1) the level of exposure to the danger (e.g., storms, heat waves, droughts) and (2) the capacity of the vulnerable party to respond, cope, and adapt (Wisner et al., 2004; Tompkins et al., 2008). For example, Hurricane Mitch killed thousands when it struck Honduras in 1998, but had a much less devastating impact on Florida (Glantz and Jamieson, 2000). The reasons for the dispa- rate consequences relate both to the changing nature of exposure (Mitch started as a category 5 hurricane in the Caribbean and ended as a tropical storm in Florida) and to the high levels of pov- erty in Honduras, where many died because they did not have the means to flee or to âride out the storm.â Even in a country as wealthy as the United States, the growing frequency and cost of climate-related disasters have taken a toll. In the 1990s there were 460 presidential disaster declarations, nearly double the number of the previous decade, and 498 declarations were made from 2000 to October 2008.1 Of the 62 weather-related disasters that cost more than $1 billion between 1980 and 2004, one-quarter hap- 1 http://www.fema.gov/news/disaster_totals_annual.fema.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 25 pened after 2000 (DOC, 2005, cited by Burby, 2006: 172). Hurricane losses since 1990 have risen dramatically, both in absolute terms and as a fraction of gross domestic product (Nordhaus, 2006), mostly be- cause of increases in the population and the value of assets in exposed coastal regions (Pielke et al., 2008). Higher costs can be expected as climate continues to change (IPCC, 2007a). Research on climate vulnerability has identified many factors, both positive and negative, that shape the level of exposure and sensitivity of people and settlements (Eakin and Luers, 2006; see also Backlund et al., 2008; Gamble, 2008; Savonis et al., 2008). For example, changing demographics in U.S. coastal areas have likely increased overall vulnerability to storm-related flooding and damaging winds. Not only are more people living permanently (rather than seasonally) on coasts, they also are older (retirees), more racially and ethnically diverse, and more likely to have low- wage jobs (Cutter and Emrich, 2006). Approximately half of the U.S. population, 160 million people, lives in a coastal county (Gamble, 2008). By 2050, 86 million people in the United States will be 65 or older and potentially more sensitive to the effects of heat waves and flooding. Managing this vulnerability requires both short-term actions to prevent disasters and assist recovery efforts (e.g., evacuation; supply of clean water, shelter, and food; recon- struction of infrastructure) and longer term structural reforms to reduce peopleâs vulnerability to disasters (e.g., land-use regulation; Lemos et al., 2007). By definition, extreme events occur infrequently, typically as rare as, or rarer than, the top or bottom 10 percent of all occur- rences. A relatively small shift in the mean climate, caused by human activities or natural variability (e.g., changes in atmospheric circulation associated with the El NiÃ±o/Southern Oscillation [ENSO] phenomenon), can produce a larger change in the number of extremes. In a changing climate system, some extreme events will be more intense, some will occur more frequently, and others will occur less frequently (Karl et al., 2008). Yet building codes and insurance premiums are based in part on the occurrence of ex- treme events in the past. Over the past few decades, the number of heat waves and warm nights has increased in the inhabited continents, while cold days, cold nights, and days with frost have become rarer (Figure 2.2). The
26 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.2 Observed trends (days per decade) for 1951 to 2003 in the frequency of extreme temperatures, defined on the basis of 1961 to 1990 values, as maps for the 10th percentile, (a) cold nights and (b) cold days; and 90th percentile, (c) warm nights and (d) warm days. Trends were calculated only for grid boxes that had at least 40 years of data during this period. Black lines enclose regions where trends are significant at the 5 percent level. Below each map are the global annual time series of anomalies (with respect to 1961 to 1990). The orange line shows decadal variations. Trends are significant at the 5 percent level for all the global indices shown. SOURCE: From Trenberth et al. (2007), FAQ 3.3, Figure 1, Cambridge University Press. Adapted from Alexander et al. (2006).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 27 United States has experienced fewer severe cold episodes over the past decade than for any other 10-year period in the U.S. historical climate record, which dates back to 1895 (Kunkel et al., 2008). One of the adverse consequences of warmer winters (along with pro- longed drought stress and forest management practices) is the spread of the pine bark beetle, which has decimated forests in the western United States (NegrÃ³n et al., 2008). Global warming also influences changes in precipitation. Air holds more water as it warms (Dai, 2006; Santer et al., 2007), result- ing in more moisture for storms and thus heavier rainfalls or snowfalls and greater potential for flooding. For the contiguous United States, statistically significant increases in heavy (upper 5 percent) and very heavy (upper 1 percent) precipitation have been observed over the past three decades (Kunkel et al., 2008), and heavy rain events are contributing more to the total precipitation (Klein Tank and KÃ¶nnen, 2003; Groisman et al., 2004; Alexander et al., 2006). At the same time, warmer air leads to greater evaporation and surface drying in some areas and thus contributes to drought and increased risk of wildfires. Over the past several decades, drought has increased, especially in Africa, southern Asia, the southwestern United States, Australia, and the Mediterranean region (Figure 2.3). The extent of very dry land across the globe has more than doubled since the 1970s (Dai et al., 2004) as a result of decreases in precipitation and the large surface warming. Like other climate- related impacts, the impacts of drought depend on a combination of stressors at different scales (Wilbanks et al., 2007). For exam- ple, populations already stressed by poverty, warfare, or AIDS are more vulnerable to drought (see âFreshwater Availability,â below). Understanding how these stressors combine and interact is essen- tial for informing policy. Intense extratropical cyclones can produce extremely severe local weather, such as thunderstorms, hail, and tornadoes. Such storms appear to be increasing in number or strength (e.g., Wang et al., 2006), and their tracks have been shifting northward in both the North Atlantic and North Pacific over the past 50 years (e.g., Gulev et al., 2001; McCabe et al., 2001). Climate models project these storms to be more frequent over the next century, with stronger winds and higher waves (Meehl et al., 2007).
28 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.3 (Top) Spatial pattern of drought for 1900 to 2002, as repre- sented by the monthly Palmer Drought Severity Index (PDSI), which measures the cumulative deficit (relative to local mean conditions) in surface land moisture. The lower panel shows how the sign and strength of this pattern has changed since 1900. Red and orange areas in the top panel are drier (wetter) than average and blue and green areas are wetter (drier) than average when the values shown in the lower plot are positive (negative). The smooth black curve shows decadal variations. Wide- spread drought is increasing in Africa, especially in the Sahel, while some regions are getting wetter, especially in eastern North and South America and northern Eurasia. SOURCE: Trenberth et al. (2007), FAQ 3.2, Figure 1, Cambridge University Press. Adapted from Dai et al. (2004).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 29 Of all extreme events, however, tropical cyclones cause the greatest property damage (e.g., Box 2.1), and so any changes in their frequency and intensity are vital to residents who live in their paths, state and local disaster preparedness organizations, and the insurance industry (Murnane, 2004). The number of tropical storms and hurricanes affecting the United States fluctuates from decade to decade, and data uncertainty is larger prior to 1965, when the satellite era began (Gutowski et al., 2008). Nonetheless, it is likely that the annual number of tropical storms and hurricanes in the North Atlantic has increased over the past 100 years, al- though there appears to be no trend in the proportions of major hurricanes or in overall intensity (Holland and Webster, 2007). When multiple storms hit the same region, as happened in Florida and Louisiana in 2005, communities have little time for recovery and resilience building. Since about 1970, and likely since the 1950s, Atlantic tropical storm and hurricane destructive potential has increased (Emanuel, 2005, 2007). The destructive potential is strongly correlated with tropical Atlantic sea surface temperatures. Model simulations sug- gest that for every 1ÂºC increase in tropical sea surface temperature, core rainfall rates will increase by 6 to 18 percent and the surface wind speeds of the strongest hurricanes will increase by about 1 to 8 percent (Gutowski et al., 2008). Other changes in the climate system (e.g., higher sea level) as well as growing populations and development in coastal zones will worsen the impacts of hurricanes and the associated storm surges and beach and wetland erosion. Research Needs Because humans both contribute to extreme weather and cli- mate events and suffer from their consequences, research is needed to understand the underlying physical and human processes and their interactions, feedbacks, and impacts, as well as to meet the information needs of stakeholders developing warning systems and response and adaptation options. For example, states need improved understanding and prediction of storm events with the potential to generate major regional flooding (CDWR, 2007). Research is also needed on how to account for changing socioeconomic conditions, including adaptation over time, to improve our understanding of
30 RESTRUCTURING FEDERAL CLIMATE RESEARCH BOX 2.1 Hurricane Katrina Hurricane Katrina was one of the worst disasters in U.S. history and offers important lessons on how U.S. coastal regions may be vulnerable to potential increases in hazard related to future climate change. The cate- gory 3 storm, which hit New Orleans in August 2005, caused $81 billion in total damage and $40.6 billion in insured losses. On the northern Gulf coast, 1.2 million people were evacuated from their homes and 1,833 people were killed, directly or indirectly. In its wake, 43 tornadoes touched ground in Florida, Georgia, Alabama, and Mississippi. The different levels of vulnerability of individuals and communities became painfully clear in the aftermath of the hurricane. Preventing similar disasters will require research from a wide range of disciplines, including atmospheric physics, biology, sociology, engineering, political science, economics, anthropol- ogy, and psychology (Gerber, 2007). However, science alone will not solve the problem if integrated approaches and better communication, disaster management, and policy capacity are not in place (e.g., Waugh, 2006).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 31 FIGURE Peopleâs ability to flee or to recover from the negative impacts of Hurri- cane Katrina revealed the many social, physical, structural, and political dimensions of extreme weather or climate events. (Left) Vehicles leave New Orleans ahead of Hurricane Katrina on August 28, 2005. SOURCE: AP Photo/Bill Haber. (Right) Thousands wait to be evacuated from the Superdome in New Orleans, September 2, 2005. SOURCE: REUTERS//David J. Phillip/Pool. ____________________________ SOURCE: Weather Channel, http://www.weather.com/newscenter/topstories/060829katrinastats.html.
32 RESTRUCTURING FEDERAL CLIMATE RESEARCH losses associated with climate extremes. Specific research needs include the following (Gamble, 2008): â¢ Improved understanding of climate thresholds and vulner- abilities, impacts, and adaptive responses (including adaptation limitations) in a variety of different local contexts around the country â¢ Improved understanding of population changes and migra- tion, especially in areas of high vulnerability â¢ Improved understanding of vulnerable populations (e.g., the urban poor, native populations on tribal lands) that have limited capacities for responding to climate change. The results are key inputs to adaptation research that addresses social justice and envi- ronmental equity concerns Given high uncertainties regarding climate impacts, it may make sense to focus more on building adaptive capacity than on developing specific adaptation options for different types of ex- treme events (Pielke, 2007). Whereas adaptation is local, ways to build adaptive capacity can be generalized across individuals, communities, and countries (Eakin and Lemos, 2006). Research is also needed on the types of incentives that will encourage adapta- tion (Christopolos, 2008). Decision support tools are needed by disaster management agencies, first responders, city planners, and others responsible for hazard mitigation and management. Examples of the science needed to manage flood risk in the context of climate change in- clude (CDWR, 2008b): â¢ Updated flood frequency analyses of major rivers and streams â¢ Studies of forecast-based operations for major reservoirs â¢ Analysis of the costs and benefits of adjusting state water supply and flood control infrastructure to accommodate climate variability â¢ Assessment of innovative techniques for improving flood risk evaluation, including use of paleoflood reconstructions Finally, much of the research on the natural climate system and human contributions and responses relies on a good observational
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 33 record that enables trends in climate, including extremes, to be dis- cerned with a high level of confidence. However, large areas of the world, even large parts of North America, are underobserved. Moreover, most observations used for climate purposes are ob- tained from weather observing networks, although these data often reflect nonclimatic changes from station relocations, land-use changes, instrument changes, and observing practices that have varied over time. Only a few countries have developed true climate observing networks that adhere to the Global Climate Observing System (GCOS) climate monitoring principles.2 Research using observations for which nonclimatic changes have been removed, therefore, would provide a better understanding of climate system variability in extremes. Because of the presence of multidecadal modes of variability in the climate system, an understanding of natural and human effects on historical weather and climate extremes is best achieved through study of very long (century-scale) records. For many of the extremes discussed above, including temperature and precipitation extremes, storms, and drought, long-term, high-quality, homogeneous records are not available. Particular requirements to further improve our un- derstanding and detection of changes in weather and climate extremes include the following (Easterling et al., 2008): â¢ Research on how to quantify uncertainty in homogeneity- adjusted climate datasets, and the best adjustment methods â¢ Continued development and maintenance of high-quality climate observing systems that adhere to the GCOS climate moni- toring principles (e.g., U.S. Climate Reference Network3), including open exchange of data so more comprehensive analysis products can be produced â¢ Collection of higher frequency data, such as hourly pre- cipitation â¢ Collection of socioeconomic observations to inform im- pact, vulnerability, and adaptation research (e.g., cost-benefit data to analyze adaptation options; data on social networks, preferences, and adaptation resources and institutions; vulnerability indicators) 2 http://www.wmo.int/pages/prog/gcos/documents/GCOS_Climate_ Monitoring_Principles.pdf. 3 http://www.ncdc.noaa.gov/crn.
34 RESTRUCTURING FEDERAL CLIMATE RESEARCH â¢ Analysis of long-term observations by multiple, independ- ent experts to improve confidence in detecting past changes â¢ Creation of annually-resolved, regional-scale reconstruc- tions of the climate for the past 2,000 years to improve our understanding of regional climate variability â¢ High-temporal-resolution data from climate model simula- tions to improve understanding of potential changes in weather and climate extremes SEA LEVEL RISE AND MELTING ICE The reconstructed record of global sea level (1870 to 2001) re- veals an average increase of 1.7 Â± 0.3 mm per year (Church and White, 2006), primarily as a result of expansion of warming sea- water and discharge of ice from alpine glaciers, ice caps, and the Greenland and Antarctic ice sheets to the oceans. Although the rate of sea level rise varies on decadal scales, over this observational period global sea level exhibited an acceleration of 0.013 Â± 0.006 mm yr-2 (95 percent confidence; Church and White, 2006). Since 1993, tide gauge and altimetry data confirm the rate of sea level rise to be ~3 mm per year, although this rate was also attained briefly around 1950 and 1970. This recent acceleration is driven in part by increased thermal expansion and the melting of nonpolar glaciers (Meier et al., 2007). Increased ice discharge from Greenland also plays an im- portant role. Although measuring ice discharge and ice sheet mass balance is challenging (Cazenave and Nerem, 2004), available evi- dence suggests about a fourfold increase in Greenland ice discharge from 1993 to 2003 relative to the 1961 to 2003 period (IPCC, 2007b, Chapter 5). Observations using advanced technolo- gies point to accelerated ice losses since 1993 ranging from about 60 percent (1993 to 1998; Krabill et al., 2004) to a threefold in- crease (1993 to 1998 relative to 1998 to 2004; Thomas et al., 2006).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 35 The impact of melting sea ice on polar bear habitat is becom- ing iconic,4 but melting ice is also affecting human settlements. For example, the Inuit people of North America are having to change hunting and fishing practices and travel routes, and their cultural traditions and health are being adversely affected (Hassol, 2004). Rising seas added to high tides and storm surges will have pro- found effects on the built environment (e.g., Box 2.2) and ecosystems in coastal areas. Intrusion of saltwater will affect groundwater quality and supplies (Backlund et al., 2008), and is one of the most severe threats to long-term agricultural sustainabil- ity in the Pacific Islands (Shea et al., 2001). Higher storm surges will disrupt sewer systems and water treatment facilities and pro- mote rapid barrier island migration or segmentation, disintegrating wetlands (CCSP, 2009). Coastal and near-shore ecosystems such as coral reefs, mangroves, and sea grass communities as well as the coastal fisheries they support are particularly vulnerable to rising sea levels and increased storm surges. Predictions of how fast and how much sea level might rise are hampered by the scarcity of observations. Recent observations of increased ice discharge from Greenland (Rignot and Kanagaratnam, 2006; Howat et al., 2007), West Antarctica (Thomas et al., 2004; Rignot et al., 2008), and the Antarctic Peninsula (Scambos et al., 2004) were not included in Intergovernmental Panel on Climate Change (IPCC) projections. Thus, the IPCC projection of 0.18 to 0.59 meter of sea level rise by 2100 is likely an underestimate (IPCC, 2001b, Technical Summary; Rahmstorf, 2007; Pfeffer et al., 2008). Two recent studies using different approaches concluded that an increase of 1 meter by 2100 lies well with projected ranges (Rahmstorf, 2007; Pfeffer et al., 2008). A global sea level rise of 1 meter would affect 145 million people (most in Asia) at a cost of nearly 1 trillion U.S. dollars (IPCC, 2007c, Table 6.12). It would also inundate 65 percent of the coastal marshlands and swamps in the contiguous United States (Backlund et al., 2008), affecting habi- tat quality and triggering rapid nonlinear ecological responses (Burkett et al., 2005). Figure 2.4 shows the global and a local (San Francisco) area expected to be inundated by a 1 meter rise in sea level. 4 See news stories such as http://environment.newscientist.com/channel/ earth/climate-change/dn11656 and assessments of the population status of polar bears, such as Schliebe et al. (2006).
36 RESTRUCTURING FEDERAL CLIMATE RESEARCH BOX 2.2 Increasing the Adaptive Capacity of Transportation Systems on the Gulf Coast The Gulf Coast is one of the most climate-vulnerable regions in the United States. It is also one of the most critical for energy security since approximately two-thirds of all U.S. oil imports and 90 percent of domestic oil and gas extracted from the outer continental shelf are transported through this region (Potter et al., 2008). The oil and gas transportation networks as well as the regionsâ complex web of roads, airports, and wa- terways are vulnerable to sea level rise (and also to warmer temperatures, increased storm activity, and changed precipitation patterns; see Savonis et al., 2008). A sea level rise of 2 to 4 feet would place 27 percent of the major roads, 9 percent of the rail lines, and 72 percent of the ports at or below 4 feet in elevation at risk, despite protective structures such as lev- ees and dikes (Potter et al., 2008). Because the planning time frame of transportation managers is around 20 to 30 years, important decisions that will shape the regionâs adaptation options for the future are being made today. Although transportation managers are accustomed to planning un- der high levels of uncertainty (e.g., future travel demand, vehicle emissions, revenue forecasts, seismic risks) and environmental pressure, better climate change-related knowledge (e.g., levels of exposure, vulner- ability, resilience) are necessary to develop robust adaptation options. Research needs of interest to decision-makers include integrated climate data and projections, risk analysis tools, and region-based analyses. Sea levels will change noticeably only over decades, but such changes will continue for many centuries into the future. Knowing how much regional and local sea level is likely to rise would help improve the design and implementation of cost-effective measures to protect against coastal inundation, salinization of groundwater and estuaries, enhanced erosion, and ecosystem losses and for managing long-life infrastructure such as nuclear power plants. For example, Mount and Twiss (2005) estimated that it would cost at least $1 bil- lion to raise the California Central Valley levees just 0.15 meter. Sea level rise will have a greater impact in areas that are subsiding or that have gently sloping shorelines. The mid-Atlantic coast of the United States is an excellent example of a region with high potential for enhanced damage due to storm surges associated with extreme weather events (hurricanes, norâeasters; Najjar et al., 2000). A recent study (Kleinosky et al., 2007) highlights the vulnerability and in- creased risk of damage for 10 cities in the Hampton Roads, Virginia area due to hurricane storm surges superimposed on sea level rise, population growth, and poorly planned development.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 37 FIGURE 2.4 Areas potentially vulnerable to inundation as a result of a 1 meter rise in sea level, which is within the range expected by many scien- tists by the end of this century. The red areas in the top image show the global distribution of these low-lying shorelines. SOURCE: National Aeronautics and Space Administration, http://www.nasa.gov/topics/earth/ tipping_points_hiresmulti_prt.htm. The light blue areas in the bottom figure depict low-lying areas within the San Francisco Bay area, Califor- nia, based on U.S. Geological Survey elevation data and imagery from the National Agriculture Imagery Program. SOURCE: San Francisco Bay Conservation and Development Commission.
38 RESTRUCTURING FEDERAL CLIMATE RESEARCH Research Needs The threat of rising sea level raises a number of questions that cannot be answered with our current level of understanding. Among the most important are: At what degree of warming will the ice sheets of Greenland and West Antarctica be drastically affected? How will their marine-terminating glaciers and ice streams respond to warmer conditions? What volume of land-based ice might be dis- charged into the oceans and how rapidly? The controls on glacier flow are dominated by ice dynamical processes that are nonlinear (Howat et al., 2007), raising the possibility that glaciers and ice streams may become so unstable (pass a tipping point) that they will begin to rapidly discharge ice until a new steady state or equilib- rium condition is achieved. Once large ice streams and marine- terminating glaciers begin to move, they cannot be stopped by any form of intervention. The response of large ice sheets to warmer climate conditions has so far been difficult to quantify, model, and predict (Alley et al., 2005; IPCC, 2007b, Technical Summary; Rahmstorf, 2007). The controls on ice flow (dynamics), including the possible influence of meltwater and basal lubrication on glacier discharge, are poorly un- derstood, in part because of scanty observations (Das et al., 2008; Joughin et al., 2008). After the breakup of the Larsen B Ice Shelf on the eastern side of the Antarctic Peninsula, the affected outlet gla- ciers began to flow two to six times faster, whereas those flowing into the remaining intact parts of the ice shelf did not accelerate (Scambos et al., 2004). Projections of how these ice sheets are likely to respond requires the development of coupled ice sheetâoutlet gla- cierâocean models that can be nested within global climate system models. In situ data are needed on mass balance components (pre- cipitation, sublimation, blowing and drifting snow), changes in glacier dynamics and subglacial drainage systems, and the thermo- dynamic interactions of marine-terminating glaciers (Holland et al., 2008) and ice shelves buttressing land-based ice with warm water intruding underneath. Remotely sensed observations (e.g., laser al- timeter, synthetic aperture radar [SAR], gravity field differences) are needed to understand the drivers of mass balance changes. Repeat SAR images make it possible to estimate the volume of ice dis- charge per unit time. Realistic projections of sea level rise demand
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 39 better ice sheet models and, although progress has been made, no model including all relevant forces yet exists (Alley et al., 2005), nor does one appear imminent. Despite the predicted negative impacts, U.S. coastal policy often does not take sea level rise into consideration (CCSP, 2009). Research priorities should focus on tools, datasets, and land management in- formation to support and promote sound coastal planning, including better data and resources provided via platforms (e.g., geographic in- formation systems) that improve their usability by decision makers. The research should also link physical vulnerability with economic analysis, planning, and assessment of adaptation options. Specific re- search needs include the following: â¢ Understanding of increased risks of and damages from coastal storm surge flooding â¢ Developing risk management approaches for coastal devel- opment and local land-use planning â¢ Developing âplanned retreatâ strategies, such as the demoli- tion of large structures near the shore if sea level rises by a specified amount (Titus, 1990) or the prohibition of reconstruction of coastal property severely damaged by repeated flooding (Yohe and Neu- mann, 1997), or coastal protection strategies that factor in sea level rise and climate change, such as those planned in the Netherlands.5 FRESHWATER AVAILABILITY Climate change poses a grave threat to the availability of freshwa- ter in the United States and around the world. Large populations concentrated in cities and suburbs as well as our entire agricultural base are dependent on, and accustomed to, safe, reliable sources of freshwater. The availability of freshwater involves both supply and demand. Already there is growing demand (e.g., by people, agricul- ture, industry), unequal distribution (UNDP, 2006), and declining sources (Figure 2.5; IPCC, 2007a; Bates et al., 2008). On the supply side, the availability of freshwater depends not only on the global wa- ter cycle, which describes the flows and storage of water in the natural 5 http://www.deltacommissie.com.
40 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.5 Declining water availability per capita from 1950, projected to 2025. Developing countries in arid regions are expected to be hardest hit by water scarcity, but even developed countries can expect significant reductions in water availability. SOURCE: UNDP (2006). Adapted from Pitman (2002), data copyright World Bank. system (precipitation, lakes, river flow, groundwater, snowpack, gla- ciers, and water vapor), but also on our technical capacity to store and adapt freshwater systems to societal needs (e.g., by building dams and canals or by developing restoration and clean up technologies). On the demand side, freshwater availability depends on governance (indi- viduals, institutions, communities, and organizations), behavior, and the values shaping water use and sustainability (e.g., consumption, conservation, valuation and equitable distribution). Both supply and demand are expected to be affected by a changing climate.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 41 How much change has already occurred and how much is likely to occur in the future is uncertain because of the large natural variabil- ity of basic components of the water cycle (e.g., precipitation, stream flow) and the difficulty of predicting many of the social and behav- ioral processes affecting water (e.g., consumption, conservation). For example, climate is potentially a transformation factor in water gov- ernance (Bates et al., 2008). As with many climate variables, uncertainty in both past and projected trends is smallest for large (e.g., global) spatial averages, and largest at the regional scales (e.g., river basins and groundwater reservoirs), where the water cycle most di- rectly affects society and where water managers most need information (Beller-Simms et al., 2008; Lemos, 2008). Average global precipitation and evaporation are expected to in- crease, based on theory and confirmed by many models. The model projections of changes in precipitation are not uniform around the globe; generally the increase in precipitation is expected to be most pronounced in the extra tropics, accompanied by a drying of the tropi- cal land from 10S to 30N (Bates et al., 2008). Some observational analyses have suggested that precipitation increase associated with increasing temperatures may be underestimated by current models (Lambert et al., 2008). The physical processes involved in precipitation and evapora- tion are complex, involving dynamic processes that occur over length scales smaller than those resolved by climate models, ranging from aerosols to clouds to weather systems. Radiation budgets are affected by both scattering aerosols such as sulfates and absorbing aerosols such as black carbon, which intercept sunlight before it reaches the surface (IPCC, 2007b, Chapter 2). The reduction of sunlight at the ground leads to a decrease in evaporation and a corre- sponding decrease in precipitation. Aerosols can also nucleate cloud drops and influence rainfall patterns locally and regionally (Rosenfeld et al., 2008). Changes in spatial gradients of sea surface temperatures, due to natural or anthropogenic forcing, also have a major influence on continental precipitation (Box 2.3). The socio- economic and political processes are also complex and feedbacks among water access, consumption, markets, ecosystems services, equity and gender distribution, security, development, and health are not well understood (UNDP, 2006).
42 RESTRUCTURING FEDERAL CLIMATE RESEARCH BOX 2.3 Sahel: Drought of Unprecedented Severity The Sahel region of Africa borders the Sahara Desert and is an area of low rainfall, frequent drought, and limited natural resources (top two figures). The Sahel region as well as the rest of western Africa face major challenges arising from climate variability and the effects of predicted cli- mate changes on food production, freshwater availability, and desertification. The last Sahelian drought from the early 1970s to the mid 1980s is among the worst on record and left about 100,000 dead and close to a million on food aid (Wijkman and Timberlake, 1984). FIGURE Desert landscape in Mali. SOURCE: Romano Cagnoni/Peter Arnold Inc. Scientists are still debating the causes of the devastating drought, but the current consensus is that the primary forcing term is de- cadal-scale changes in ocean temperatures. In particular, warmer temperatures in the Indo-Pacific warm pool and a combination of cooler- than-normal North Atlantic temperatures and warmer-than-normal South Atlantic temperatures are emerging as the dominant factors. Greenhouse warming can account for the warmer Indo-Pacific warm pool temperatures and aerosol cooling can account for the cooler-than-normal North Atlantic sea surface temperatures. But these human forcing terms by themselves cannot account for the amelioration of the drought. Natural variations in Atlantic and Indo-Pacific sea surface temperatures (e.g., ENSO induced) have to be invoked. Scientific uncertainty is hampering our ability to pre- dict future changes in this vulnerable region of sub-Saharan Africa. For example, simulations by two reputable climate models in the United States disagree on even the sign of the changes (bottom figure): one predicting a
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 43 40 to 60 percent decrease in rainfall by 2100 and the other predicting a 20 percent increase! FIGURE Observed precipitation trends from 1950 to 1999. SOURCE: Hoerling et al. (2006). Copyright 2006 American Meteorological Society.
44 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE Simulated time series of rainfall departures over the Sahel for July through September 1870 to 2099 from two different climate models. Reference climatology is 1870 to 1999. Both models were forced with estimated greenhouse gas and aerosol changes through 1999 and with the SRES A1B emissions scenario (IPCC, 2007b) thereafter. SOURCE: Adam Phillips, NCAR, based on results re- ported in Hoerling et al. (2006). The great Sahelian drought illustrates the many factors that influence the impact of water scarcity on ecosystems, communities, and social groups. Twenty years after the drought, research focusing on the interac- tions between environmental degradation, socioeconomic transformation, and climatic change has painted one of richest pictures of vulnerability and adaptation in the less developed world (Batterbury and Warren, 2001; see also the special issue of Global Environmental Change, 11, 2001). By examining social, political, and environmental change together, this re- search challenged well-established myths about desertification, competition between human settlements and livestock for land and water, and migra- tion, and showed how large-scale processes at the global level (greenhouse warming and anthropogenic pollution) connect with local-scale processes (livelihood adaptation and local knowledge) to produce drought. ___________________________ SOURCES: Zeng (2003) and Hoerling et al. (2006).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 45 Scientists are able to predict global average trendsâthe first- order response of the hydrological cycle to a warming planet. However, the regional changes that are most important for plan- ning water resources are poorly understood and less predictable. For example, changes in the distribution of temperature and water vapor in the atmosphere, as well as in circulation patterns, will change the amount and type of precipitation (snow or rain) that falls across the United States. Coupled with temperature and radia- tion changes at the land surface, such direct and indirect effects make prediction of critical resources such as the depth of seasonal snowpack in the Rockies difficult. Similarly, for areas dependent on groundwater aquifers, recharge is sensitive not just to total pre- cipitation, but also to changes in storm climatology (intensity, duration, and intermittency of storms, all of which change with climate warming) as well as near-surface weather parameters (e.g., air temperature, humidity; Levine and Salvucci, 1999; NRC, 2004b). Statistical analyses of hydrological and meteorological records have found evidence that such key aspects of the water cycle are changing. For example, observations around the world indicate an increase in frequency of intense rainfall. Models sug- gest this intensification will increase in the coming decades (Figure 2.6), leading to increased incidence of diseases such as diarrhea (UNDP, 2006), more flooding (Backlund et al., 2008), and, ironi- cally, to possibly less recharge. The IPCC Fourth Assessment predicts both increases (wet re- gions get wetter) and decreases (dry regions get drier) in annual average river runoff and water availability, as well as changes in the extent of areas affected by drought and flooding (IPCC, 2007a). Of particular concern is the vulnerability of mountain gla- ciers and snowpack and the risk of severe loss of water resources. A recent example is the 8-year drought in the Colorado River basin (1999â2007 water year), which is the most extreme in the meas- ured hydrological record (100 years). Within the past decade, many communities in southern California have experienced their single driest year on record (CDWR, 2008a). Even the southeast- ern United States is in a drought (Box 2.4). Some of the most alarming findings in the IPCC Fourth As- sessment, such as the estimate that dryland areas have doubled since the 1970s (IPCC, 2007b, Chapter 3), are not based on direct
46 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.6 Increase in the amount of daily precipitation over North America that falls in heavy events (the top 5 percent of all precipitation events in a year) compared to the 1961â1990 average. Various emission scenarios are used for future projections. Data for this index at the conti- nental scale are available only since 1950 (pink line). SOURCE: Karl et al. (2008). measures such as soil moisture, but rather on crude estimates based on running averages of precipitation and air temperature. The large uncertainties highlight the need for future investments in obser- vations, models, and process understanding. However, current climate information can be used to support decisions on water re- sources at a variety of geographic scales, even at the current skill levels of hydrological forecasts (Beller-Simms et al., 2008). A host of freshwater governance options have emerged, including mecha- nisms for making decisions under uncertainty and adaptive management to enable freshwater systems to respond to different kinds and magnitudes of impacts (Ivey et al., 2004; Olsson et al., 2004; Pahl-Wostl, 2007; Werick and Palmer, 2008).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 47 In summary, the twentieth century has witnessed fundamental changes in the hydrological cycle from global to watershed scales. From a physical perspective, coupled oceanâatmosphere models are able to account for some of these changes, but poor understand- ing of the forcing terms (e.g., aerosols, land surface changes), coarse resolution of the models (few hundred kilometers), and lack of data over the oceans limits the ability of models to capture re- gional and local changes. From a socioeconomic perspective, there has been considerable research on water but much less that also considers water governance and climate. An integrated approach that takes account of physical, social, and ecological factors affect- ing freshwater change is needed to understand the potential transformations and find solutions. BOX 2.4 Drought in the Southeast, a Wake-Up Call Parts of the southeastern United States have been experiencing drought conditions since 2005 or early 2006 (Figure). The southeastern drought is instructive for two reasons: it provides an example of drought in a humid part of the United States, where water scarcity is not typically seen as a major challenge; and it illustrates the regionâs relative lack of preparedness for drought, especially compared to the arid West, where drought is a prominent water management concern. The Southeast has experienced significant population growth, but has not invested in the ma- jor interregional water infrastructure and institutional arrangements that might have allowed it to respond to drought. Although decreasing rainfall was well monitored by state climatologists, the impacts to agriculture, fish- eries, and municipal water supplies may have been made worse because the involved states (Georgia, Alabama, and Florida) failed to act on the water resources compacts between them (Feldman, 2007, cited in Beller- Simms et al., 2008). The states could not agree on water allocation schemes and so let the compact expire. Faced with the tough decision of either relying on the forecasted above-average Atlantic hurricane season or being more conservative, the governor of Georgia instituted state-level outdoor water restrictionsa and declared a state of emergency for parts of the state in October 2007. Georgia then filed a motion in federal district court for a preliminary injunction to require the U.S. Army Corps of Engi- neers to reduce releases for downstream water demands, including mandated flows for aquatic species in Florida listed under the Endangered Species Act. In addition, Georgiaâs Senate Resolution 822, introduced in 2008, called for establishment of Georgia-North Carolina and Georgia- Tennessee boundary-line commissions to survey and settle disputed state boundary locations that, if settled in Georgiaâs favor, would place portions of waterways such as the Tennessee River within Georgia. This case shows that even when climate information is available, unresolved con-
48 RESTRUCTURING FEDERAL CLIMATE RESEARCH flicts between upstream and downstream user priorities constrain their use for mitigating negative impacts (Beller-Simms et al., 2008). FIGURE Low water levels in Lake Lanier, the main source of drinking water for Atlanta, shown on November 16, 2007. SOURCE: Ed Jackson, University of Georgia. _____________________________ a http://www.caes.uga.edu/topics/disasters/drought/. Research Needs Specific research needs include the following: â¢ Prediction of changes in water supply (runoff, groundwa- ter, snowpack) and the reliability of the water supply, which requires improvements in decadal modeling, regional modeling, and understanding and modeling of the land surface hydrological sensitivity to climate change (Graham et al., 2007) â¢ Understanding the causes and predictability of extreme events (e.g., droughts, floods) â¢ Improve understanding and predictability of updated wa- tershed-level rainfall-runoff relationships that account for increased precipitation intensity for flood forecasting purposes, especially for locations prone to rain-on-snow flood events
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 49 â¢ Prediction of changes in water demand, which require demographic models that incorporate climate change impacts and models that consider the effects of climate change on natural and agricultural landscape water use â¢ Research on water governance, including adaptive man- agement models, adaptive capacity building, and water systems sustainability â¢ Research on the economics of water supply, demand, and conservation; and on human perceptions and valuations of impacts, cost of adaptation, and equity, which are needed to inform adaptive action â¢ Development of long-term observations and tools for pre- dicting hydrological variables of most value to water resource managers (e.g., timing of snowmelt, groundwater recharge rates) from climate model output Better datasets are needed for determining decadal- to longer scale trends in regional forcing terms because of aerosols (spe- cifically absorbing aerosols, which would be measured by the Aerosol-Cloud Ecosystems mission recommended by the National Research Councilâs Decadal Survey [NRC, 2007b]) and land- surface modification (specifically land cover change, which would be measured by the Landsat Data Continuity Mission). Precipita- tion measurements over land and the oceans are critical for both basic climate science and water resources applications and would be made by the Global Precipitation Mission, recently given high priority in the Decadal Survey. Global measurements of stream flow, soil moisture, and evaporation are also needed. For example, the Soil Moisture Active and Passive Mission, also recommended by the Decadal Survey, would provide data for drought monitoring and for driving predictive models of water balance. Socioeconomic data needs include water demand, consumption patterns, scarcity, equity, distribution, and adaptation costs. A comprehensive review of operational ground-based monitoring networks (e.g., Snowpack Telemetry Network) would reveal whether they are adequate to detect climate change impacts.
50 RESTRUCTURING FEDERAL CLIMATE RESEARCH AGRICULTURE AND FOOD SECURITY In 2007, an estimated 923 million people were seriously undernourished, 75 million more than in 2005.6 Climate change is expected to alter the global food supply, with implications for global and regional agricultural production and food security. Indeed, the ef- fects of climate change on food availability and the stability of the food system are already being felt, especially in rural locations where crops fail or yields decline, and in areas where supply chains are disrupted, market prices increase, and livelihoods are lost (FAO, 2008). Some important agricultural areas appear to be experiencing sig- nificant deviations from the average climatic conditions under which the current farming systems developed, causing considerable hardship (e.g., Box 2.5). A record-setting severe winter in Central Asia in 2007 followed by large snowfall and severe flooding have threatened food security in the region, particularly in Tajikistan and Kyrgyzstan.7 Af- ghanistan and Iraq are currently experiencing the worst drought in 10 years, adding to those nationsâ woes.8 Such problems can also stress less vulnerable countries. The extreme heat wave of 2003 in France and Italy resulted in uninsured economic losses of EUR 13 billion for the agricultural sector (IPCC, 2007c, Chapter 5; see also Box 2.9). Understanding and predicting regional climate trends and their impact on agriculture and the national food supply is a high pri- ority for governments.9 Climate change affects both commercial farming, which is often an integral part of national economies, and subsistence farming, which is common in developing countries and determines the livelihood of millions of people. This latter group is by far the most vulnerable to climate variability and change (Parry et al., 2005), and it is well recognized that poor, natural resource- dependent, rural households will bear a disproportionate burden of the adverse impacts of climate change (Mendelsohn et al., 2007; Agrawal, 2008). 6 http://www.fao.org/newsroom/en/news/2008/1000945/index.html. 7 http://www.fao.org/giews/english/shortnews/casia080408.htm. 8 http://www.pecad.fas.usda.gov/highlights/2008/09/mideast_cenasia_ drought/. 9 http://www.earthobservations.org/cop_ag_gams.shtml.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 51 BOX 2.5 Catastrophic Long-Running Agricultural Drought, Murray Darling Basin, Australia The Murray Darling Basin provides 85 percent of the water used for irrigation in Australia and has traditionally produced 40 percent of the countryâs fruit, vegetables, and grain. Whereas drought typically occurs once every 20 years or so, over the past 7 years it has become an annual occurrence. The Australian government has spent about $2 billion dollars in the past few years in "exceptional circumstances" payments to help the affected farmers. It also offered $150,000 to farmers who decide to leave their land. Over the past 5 years, extreme drought conditions have forced more than 10,000 farmers off the land. Many ranchers have had to sell off their stock, the remaining farmers have had to use water more efficiently, and severe water restrictions have been introduced in urban areas in the region. The competing demands for water for domestic use, irrigation, and ecosystem preservation far exceed the recent flow of the three main rivers. A U.S. $3.6 billion emergency water conservation plan is being put in place, but it is uncertain whether this plan will be enough if the drought continues or if it will be implemented in time to make a significant difference. FIGURE Dead trees and cracked earth on a farm near Kerang, a district in the Murray Darling Basin about 360 km north of Melbourne, August 24, 2007. SOURCE: REUTERS/Tim Wimborne. Australia can experience strong ENSO events with devastating events on rangelands and agriculture. Severe droughts have occurred throughout the countryâs history, for example in 1900, 1942, 1982, and 1992. The IPCC Fourth Assessment projects that there will be up to 20
52 RESTRUCTURING FEDERAL CLIMATE RESEARCH percent more droughts in the region by 2030 and a decrease in annual Murray Darling River flow by 10 to 25 percent by 2050. The La NiÃ±a conditions in 2006 failed to bring its usual heavy rains to Australia, highlighting the need for improved seasonal to interannual re- gional climate forecasting to better predict rainfall and temperature over the next season and likely trends over the next few years. A better under- standing of climate trends would place federal and regional governments in a better position to manage the resulting economic impacts and popula- tion displacement, to help their most vulnerable citizens, and to promote effective adaptation strategies. ________________________ SOURCES: http://news.bbc.co.uk/2/hi/asia-pacific/7499036.stm, http://www.independent.co.uk/news/world/australasia/australias-epic-drought-the- situation-is-grim-445450.html, http://www.bom.gov.au/climate/drought/livedrought.shtml, http://www.environment.gov.au/water/mdb/index.html, http://news.nationalgeographic.com/news/2007/11/071108-australia- drought_2.html. Tropical crop production is likely to suffer under a warming climate, whereas mid- to high-latitude regions could benefit ini- tially from a small amount of warming (IPCC, 2001a). In its Fourth Assessment report, the IPCC noted that the large majority of climate models predict a decrease in precipitation in the subtrop- ics by the end of the century and an increase in precipitation extremes in southern and eastern Asia, east Australia, and northern Europe (IPCC, 2007c, Chapter 5). Declines in water availability are projected for the Mediterranean Basin, Central America, sub- tropical Africa, and Australia. The Southwest and mid-continental agricultural areas of the United States are also expected to have droughts, reducing crop production and/or increasing demands for water in an area that is already beginning to experience water con- flicts. The last U.S. national assessment (Reilly et al., 2000) concluded that the net effect of the climate scenarios studied on the agricultural sector over the twenty-first century is generally posi- tive. A more recent and detailed assessment of the impacts of climate change on U.S. agriculture, crops, rangelands, and live- stock (Hatfield et al., 2008) presents a less optimistic picture. Although the Food and Agriculture Organization projects about a 60 percent decrease in the growth rate of food production, an 80 percent increase in agricultural production by 2050 is re- quired to feed a growing population (FAO, 2008). This need, in
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 53 turn, will require new croplands to be cultivated, many of which will replace tropical woodlands and forests in sub-Saharan Africa and Latin America (Ramankutty et al., 2002). Further increases in cultivated area may be needed if an increased frequency of climate extremes lowers yields below projections (Easterling et al., 2007). Shrinkage of mountain glaciers will decrease the water available, but increasing evaporation will increase the need for irrigation. The stress of climate variability and change on the global food supply will be exacerbated by population growth, increased wealth in developing countries, rising cost of fertilizer, political instabil- ity, national policies, and pests and invasive species. For example, wheat fields in the United States are now being planted with corn for ethanol, driven by the increased cost of gasoline, the demand for alternative energy sources (in this case biofuel), and govern- ment subsidies (USDA, 2007). Crop failures around the world in 2007 led several countries to meet national needs by restricting crop exports, thus reducing global supply (Trostle, 2008). Research Needs Basic and applied research, supported by modeling, field stud- ies, and satellite observations, are needed to provide an improved understanding of global agricultural land use, productivity, and food supply in the context of a changing climate. The research needs fall into two broad categories: climate modeling for agricul- ture and global and integrated modeling of agricultural land use and associated mitigation and adaptation options. Climate Modeling for Agriculture Current predictions are largely inadequate for food security systems. Improved climate models are needed (1) that generate output at regional scales with improved timeliness and skill (Muk- hala and Chavula, 2007) targeted for specific agricultural needs (Meinke and Stone, 2005), and (2) that include inputs to key proc- esses in crop models related to climate change (i.e., temperature, water stress, and their interaction with elevated CO2; Tubiello and Fischer, 2007; Tubiello et al., 2007). Matching the spatial and tem- poral scales of climate and crop models is a necessary integrative
54 RESTRUCTURING FEDERAL CLIMATE RESEARCH step (Challinor et al., 2007). Adjustments in rain-fed and irrigated agriculture depend on projections of water supply and demand over the next few decades, particularly for vulnerable semi-arid areas (see Box 2.5 and the âFreshwater Availabilityâ section). These projections should quantify the rates of glacial retreat and changes in precipitation for mountain systems which feed irrigated lands (e.g., in Central Asia). Recent efforts to enhance drought monitoring in the United States will need to be replicated for drought-prone regions of the world with populations and liveli- hoods at risk. Global and Regional Integrated Monitoring and Modeling of Agri- cultural Land Use and Associated Mitigation and Adaptation Options A new generation of integrated dynamic Earth system models that incorporate both physical and socioeconomic factors is needed to better project changes in regional food supply and demand re- sulting from a changing climate and to inform mitigation and adaptation options (Howden et al., 2007; Ingram et al., 2008). Ef- fective adaptation will require an integrated view of climate change issues, including climate variability and market risk in the context of regional economic and sustainable development (Adger et al., 2007). It will require effective institutions for determining agricultural production and implementing adaptation measures at a range of scales. Tariffs and subsidies that strongly influence the global supply of food will inevitably change as governments re- spond to shifting markets and changes in global agricultural supply and demand, resulting in part from changes in regional climate (Tubiello and Fischer, 2007). At the local scale, institutions and institutional partnerships will have to be strengthened to increase access to adaptation methods (Agrawal, 2008). Regional, spatially explicit, process models of land-use change are needed to project agricultural expansion, intensification, and abandonment and to model the potential impacts of these changes on the major biogeochemical cycles, landâatmosphere exchange of water and energy, and human population dynamics. Place-based models should explore societal vulnerability and the various autonomous and planned adaptation pathways and coping strate-
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 55 gies. Climate change at the low end of the anticipated range over the next few decades is likely to have only modest economic im- pacts on U.S. agriculture (Mendelsohn et al., 1994; Kelly et al., 2005; Schlenker et al., 2005), but the impacts in other areas of the world could be larger. Similarly, the economics of adaptation in agriculture is poorly understood. End to end, coupled biogeochemical, hydrological, and eco- nomic models should address the impacts, feedbacks, and costs of different agricultural land use (e.g., extensification, abandonment) and different land-use mitigation options to sequester carbon (e.g., conservation agriculture, no-till agriculture, shade cropping), to establish the impacts of agricultural intensification and increasing fertilizer use (e.g., see Box 2.7). Similarly, improved scientific un- derstanding is needed to examine the trade-off between using crops for food or for biofuel and the impacts on food prices, secondary land use, and soil erosion. Field experiments will be needed to parameterize these models and to quantify the net carbon seques- tration and the water-use and -quality implications associated with different mitigation and alternative energy options (e.g., NRC, 2008d). Priority Infrastructure Needs The infrastructure needed to support the aforementioned sci- ence includes increased computational capacity to run higher resolution climate models with regional specificity (see âEarth System Modelingâ in Chapter 3) and improved land surface obser- vations.10 Continuous satellite measurements, such as following the Moderate Resolution Imaging Spectroradiometer with the Visible Infrared Imaging Radiometer Suite, are needed for monitoring agriculture. Long-term moderate resolution (i.e., Landsat class) observations will be needed, but with an increased temporal fre- quency (i.e., 3- to 5-day coverage) to monitor changes in cropland and crop area and to drive crop production models and famine 10 A summary of the observational infrastructure needed for agricultural monitoring can be found at http://www.earthobservations.org/cop_ag_ gams.shtml.
56 RESTRUCTURING FEDERAL CLIMATE RESEARCH early warning systems.11 The inadequacy of U.S. spaceborne ob- servations in this respect has led the U.S. Department of Agriculture to become the single largest purchaser of Indian satellite data, which are now used for monitoring U.S. crops. Targeted high- resolution (1 to 3 m) imaging is also needed to monitor crop con- ditions in subsistence agricultural regions, to improve national agricultural production estimates, and to help monitor the agricul- tural aspects of carbon management. Recent advances in microwave remote sensing for agricultural monitoring also warrant further investigation. MANAGING ECOSYSTEMS Humans actively manage ecosystems to provide food, water, timber, and other resources. We rely on ecosystems to regulate local climate conditions and remove pollutants from the air and water (Millennium Ecosystem Assessment, 2005). Although many of these services are already under stress due to pollution, overuse, land-use change, and other anthropogenic factors, climate change will further affect the ability of ecosystems to sustain these services and natural resources (IPCC, 2007c, Chapter 4). Consequently, it will be important to understand the linkages between ecosystems, societies, and climate; to assess human and ecosystem vulnerabili- ties to climate change; and to devise management strategies that mitigate climate change (e.g., decreasing deforestation rates or planting forests to sequester carbon) while preserving ecosystems and their services. The current distribution of plant and animal species over large areas of the globe reflects human appropriation of primary produc- tion (Haberl et al., 2007), alteration and fragmentation of habitat, and modifications of the energy, nutrient, and water cycles. Mod- ern ecosystems are also responding to observed climate changes, as recorded by changes in the timing of phenological events (e.g., leaf-out, flowering), migration patterns, and the ranges of fish and marine mammals (IPCC, 2007c). The rapid rate of climate change, combined with human-induced stressors (e.g., poor land manage- 11 http://www.earthobservations.org/documents/cop/ag_gams/20070716_ geo_igol_ag_workshop_report.pdf.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 57 ment practices), may outpace the ability of ecosystems to adapt, leading to steep declines in biodiversity and ecosystem resilience (IPCC, 2007c). This issue was recognized in Article 2 of the United Nations Framework Convention on Climate Change, which states that stabilization of greenhouse gases âshould be achieved within a timeframe sufficient to allow ecosystems to adapt naturally to climate change.â The documentation of large-scale mortality events, such as those observed for pinyon-pine forests in the southwestern United States (Breshears et al., 2005) or rapid declines of coral reefs (Box 2.6) suggests that this goal is not being achieved. BOX 2.6 The End of Coral Reefs? Coral reefs provide critical habitat to support fisheries and marine biodiversity. They benefit humans by supporting fishing and tourism, sup- plying natural products, and forming a breakwater that helps protect coastal property from storm and wave damage. The global net economic benefit of reefs has been estimated at $30 billion per year (Cesar et al., 2003), including several billion dollars per year in Florida and Hawaii (Johns et al., 2001; Cesar et al., 2002). The past several decades has seen a dramatic increase in coral mortality and reef degradation (Pandolfi et al., 2003). A third of reef species are cur- rently in danger of extinction (Carpenter et al., 2008). The reasons for this decline include coastal development, increased disease, overfishing, pollution, and climate change (Buddemeier et al., 2004; see also figure). Climate change affects coral reefs in several ways. Elevated atmospheric CO2 de- creases ocean pH and carbonate ion content, reducing the ability of corals to form the calcium carbonate skeletons that form the reef structures, and ulti- mately undermining reef structures and their ability to support biodiversity (Kleypas et al., 2006). Warmer ocean temperatures may increase the geo- graphic range that coral reefs can develop, but also increases coral bleaching, caused when the coral expels its algal symbiont. Sea level rise, which changes the intensity of coastal storms and coastal erosion regimes, will also negatively affect coral reef structures. In heavily populated regions where most coral decline is observed, existing stresses weaken the ability of coral reefs to adapt to climate change. However, even corals in âpristineâ areas are affected by climate change and elevated CO2 levels. No analogs of ocean chemistry exist in the historic record to help us predict the long-term response of coral reefs to climate change. However, coral reefs are already in steep decline, creating the potential for tremen- dous consequences for marine ecology, biodiversity, and local economies. Research is needed on the basic biology of coral reefs, especially the causes of observed declines and the effects of elevated water temperature and CO2. Human behaviors driving changes in reef ecosystems, as well as consequences of the loss or alteration of coral reef ecosystem services have to be understood to design effective management strategies.
58 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE The interactions of multiple stressors related to coral reef decline. Stress- ors that are direct results of climate change are in white boxes; stressors that are related to human use of corals or that may be altered by climate change (e.g., ex- treme weather events) are in blue boxes. Note that there are feedbacks between coral reef decline and some human activities (e.g., tourism), which in turn affect reefs and how they are managed. Climate change will affect ecosystems through a number of mechanisms, such as by altering patterns of temperature, rainfall, ocean stratification, upwelling, and mortality rates caused by ex- treme events, such as storms, fires, and coastal hypoxia (Box 2.7). Interactions among individual organisms responding to climate changes are complex and may lead to threshold responses (or tip- ping points) with rapid changes in ecosystem productivity, composition, or location. For example, warmer winter tempera- tures have been linked to increased prevalence and intensity of shellfish diseases (Cook et al., 1998; Soniat et al., 2008), and ear- lier spring blooms of marine plankton which affect fishery production (Harrison et al., 2005). Interactions between tempera- ture and metabolic demand by marine organisms, coupled with the effects of temperature on oxygen solubility, can lead to nonlinear responses in fish productivity (Del Toro-Silva et al., 2008). Cli- mate-mediated physiological stress resulting from warming temperatures can also compromise disease resistance of marine and terrestrial organisms and result in emergence of new diseases, in-
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 59 creased frequency of opportunistic diseases, and exposure of pre- viously uninfected host populations to pathogens (e.g., Harvell et al., 1999, 2002, 2004). Even when climate change is minimal, ecosystems will re- spond to elevated CO2 levels in the atmosphere, through either altered photosynthesis on land (Norby et al., 2005) or reduced cal- cification rates in the oceans (Orr et al., 2005; Kleypas et al., 2006; Box 2.6). The effects of these changes on individual organ- isms and ecosystems are only beginning to be studied, and their interactions with climate responses are not well known. For exam- ple, increased productivity in forests from elevated CO2 may add resilience of forests to storm or insect damage (NegrÃ³n et al., 2008). Excess CO2 absorbed into the ocean may increase calcifica- tion by some phytoplankton species, thereby enhancing carbon export to the deep ocean (Iglesias-Rodriguez et al., 2008), but cause irreparable damage to other organisms (Box 2.6; Orr et al., 2005). Other effects of atmospheric composition on marine and terrestrial ecosystems include altered sunlight for photosynthesis with increased aerosol loading and effects of pollutants (especially nitrogen deposition and ozone) on productivity. In many locations, responses to climate or atmospheric changes will add to other human-related stressors such as pollu- tion, land use, and the introduction of invasive species to affect ecosystem resilience. For example, increases in storm runoff and warmer, more stratified coastal ocean conditions may exacerbate âdead zonesâ (Box 2.7). Ecosystem responses to climate change or large-scale man- agement will in turn modify climate on a variety of spatial scales. Land and ocean ecosystems are a key component of the climate system. They process energy, water, carbon, and nutrients, and mediate the fate of incoming sunlight, which in turn influence fac- tors such as cloud formation and greenhouse gas fluxes. Some of these mechanisms may amplify or dampen climate change through altered surface energy balance and/or greenhouse gas emissions. Management strategies that involve manipulation of ecosystems to mitigate climate change (e.g., sequestering carbon by planting trees, managing forests, or fertilizing the oceans) will involve a spectrum of humanâclimateâenvironment interactions. In evaluating such strategies, not only the target impact (e.g., sequestration) but
60 RESTRUCTURING FEDERAL CLIMATE RESEARCH BOX 2.7 Enter the Dead Zone When excessive amounts of nutrients (usually from fertilizer) flow into coastal waters, massive amounts of organic matter (e.g., algae) are pro- duced, which consume oxygen as it decays and thus create âdeadâ zones. Dead zones in coastal oceans have expanded exponentially since the 1960s, with the number doubling each decade. There are currently 400 dead zones, covering a total area of more than 245,000 km2 (Diaz and Rosenberg, 2008). The largest in the United States is at the mouth of the Mississippi River (Figure), which carries nitrogen and phosphorus from chemical fertilizers used in agriculture. In 2008 this dead zone reached its second largest extent (21,000 km2, approximately the size of New Jersey), fed by high nitrate loads (37 percent higher than 2007 and the highest recorded since measurements began in 1970) from the Mississippi and Atchafalaya rivers. Its extent would have been greater if not for aeration of the waters caused by the passage of Hurricane Dolly.a FIGURE Mississippi River plume (brown water at left) meets the Gulf of Mexico (blue water at right) at Southwest Pass. SOURCE: N. Rabalais, Louisiana Universi- ties Marine Consortium. Used with permission. Hypoxia (very low levels of dissolved oxygen) is one of multiple stressors affecting aquatic ecosystems (Figure, below); others include overfishing, habitat loss, toxic algal blooms, and climate change. Although not the direct cause of coastal hypoxia, climate change affects environ- mental conditions that can affect the extent and/or likelihood of dead zones. For example, projected increases in rainfall will increase river dis- charge, nutrient delivery, and stratification (via freshwater influx) of the upper water column, thereby possibly expanding coastal regions impacted by dead zones (Justic et al., 1997; Rabalais et al., 2002; Diaz and Rosenberg, 2008). On the other hand, more frequent storms may mitigate the development of dead zones. Changing circulation patterns may con- tribute to formation of regions of coastal hypoxia. The dead zone reported
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 61 off the coast of Oregon in 2006, which extended 3000 km, is an example of the contribution of atmospheric and coastal circulation to the develop- ment of a dead zone. Wind patterns over this region intensified upwelling of shelf water with low dissolved oxygen, exacerbating the existing bio- logically-produced hypoxic conditions in the area (Chan et al., 2008). Wind-driven coastal upwelling is a response to large-scale atmospheric and oceanic circulation patterns, which in turn are influenced by global warming. The Oregon hypoxia event is troubling because eastern bound- ary current systems such as this one are among the most productive marine ecosystems in the world. Warmer waters resulting from a warming climate will dissolve less oxygen, which may enhance low-oxygen regions. An ocean biogeo- chemical model driven by the âbusiness as usualâ emissions scenario predicts reduced oxygen levels and increased extent of oxygen mini- mum zones in the oceans in the next two centuries (Schmittner et al., 2008). Ocean temperatures will also affect O2 metabolic demand by fish (Del Toro-Silva et al., 2008), which may increase fish mortality rates and expand dead zones. These examples illustrate how climate change may become more of a critical contributor to low-oxygen re- gions. Understanding the occurrence and extent of dead zones requires research on the complex interactions between land use, coastal ecosystems, environmental conditions, and climate change. Such knowledge is needed to develop strategies to minimize and miti- gate the effect of these events. FIGURE Oxygen levels (in ppm) in the bottom waters of the Gulf of Mexico dead zone, from July 21 to 27, 2008. The red area enclosed by the black line is in a state of hypoxia. SOURCE: N. Rabalais, Louisiana Universities Marine Consortium. Used with permission. ________________________ a http://www.gulfhypoxia.net/research/shelfwidecruises/2008/Press Release08.pdf.
62 RESTRUCTURING FEDERAL CLIMATE RESEARCH also the consequences for ecosystems, non-CO2 climate effects (e.g., heat and water budgets, other greenhouse gases), the vulner- ability of carbon storage over longer timescales, and economic trade-offs (e.g., value of sequestration versus avoided emissions) need to be considered (Dilling et al., 2003). For example, private corporations are making plans for large-scale releases of iron to the oceans to generate carbon offsets, despite scientific uncertainty about the efficacy and timescale of iron fertilization for carbon sequestration and the ecological consequences of such additions (Buesseler et al., 2008). Strategies such as afforestation not only change carbon budgets but also affect water fluxes and stream lev- els across the United States (Jackson et al., 2005). Tree planting programs in boreal and tropical ecosystems will not have the same net climate effect because of the disproportionate effect of boreal trees on reflectivity of the land surface (Bala et al., 2007). Forests planted to sequester carbon will be vulnerable to extreme weather events such as hurricanes that can cause large-scale mortality (Chambers et al., 2007). Feedbacks between drought, forest mor- tality, fire, and land clearing in tropical forests may lead to a critical loss of biodiversity and to a shift to overall drier climates in tropical regions (Bonan, 2008). There will inevitably be trade-offs between use of land for climate mitigation and for other priorities. For example, a broad mitigation strategy for sustained reduction of emissions from deforestation and degradation will have to consider the livelihoods of the people living in the forested regions (Malhi et al., 2008). Despite the sustainable development requirement em- bedded in the Kyoto clean development mechanism, ecosystem management programs have fallen considerably short of their original goals (Bozmoski et al., 2008). All lands and coastal areas in the United States are managed, even if management takes the form of a decision to leave lands âwild.â However, changes in climate, atmospheric composition, and pollutant deposition will affect even set-aside areas. Consider- ing climate change in decisions on federal land management poses both scientific and regulatory challenges (Julius et al., 2008), in- cluding the ability of current regulatory frameworks such as the Endangered Species Act (ESA) to incorporate projected climate change impacts into permitting decisions (see Box 2.8).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 63 BOX 2.8 Climate Change and the Endangered Species Act FIGURE The polar bear was listed as threatened by the U.S. Fish and Wildlife Service in May 2008 and a Special Rule under the Endangered Species Act providing for its conservation was issued in December 2008. SOURCE: Dave Olsen, U.S. Fish and Wildlife Service. Concern over the future of the polar bearâthe first species listed as threatened due to climate concernsâcenters on loss of habitat associated with warming of the Arctic and sea ice melting. However, climate change will clearly affect the ranges of many individual species, the future suitability of areas set aside to preserve habitat, and plans for managing discrete popula- tions of species of concern. As presently authorized, the ESA is focused on single-species management rather than ecosystem management, an ap- proach that does not readily facilitate adaptation. Reauthorization of the ESA to address this and other aspects of how the federal government manages species of concern has been a subject of proposed legislation and much commentary over the past decade.a Experience to date with ESA admini- stration and compliance has demonstrated the Actâs strong influence on land and water management decisions and the corresponding costs of re- covery plan implementation for the regulated community, even absent the further complications associated with anticipated climate change. Options for ecosystem adaptation could include changes to the existing regulatory framework to incorporate climate change into species recovery planning and to focus funds available for recovery plan implementation on strategies that provide resiliency. Policy decisions such as thisâhow to amend current legislation or to create a new legal or administrative regulatory framework for including climate change in key legislationâwill frame needs for future sci- entific research on key species and ecosystems and their responses to climate change. ____________________________ a See, for example, http://www.publicland.org/endangerSpecies.htm.
64 RESTRUCTURING FEDERAL CLIMATE RESEARCH Research Needs An enhanced program of basic and applied research is needed to improve understanding of the responses of marine and terrestrial ecosystems to climate change, the major impacts to and vulner- abilities of ecosystems and the services they provide, and the role of human actions and nonclimate stressors in facilitating or amelio- rating the potential for rapid ecosystem responses to climate change. CCSP Synthesis and Assessment Product 4.4 (Julius et al., 2008) discusses the design of climate change adaptation strategies for a subset of federally managed lands that balance resource needs with preservation of biodiversity and key ecosystem services under a changing climate, as well as the evaluation of the societal costs of specific management actions (or inaction). Research is also needed to assess ecosystem and land management options that can help in preserving biodiversity and sequestering carbon under a changing climate and to evaluate the societal costs of specific management actions (or inaction). Specific research needs include the following: â¢ Assessment of key vulnerabilities of ecosystems to climate and other human stresses, including the compound effects of mul- tiple stresses and the potential impacts of extreme or abrupt events (e.g., heat waves, extended drought, increased severe weather or flooding, changes in sea ice extent, changes in ocean circulation). â¢ Mechanisms and timescales for adaptation of ecosystems and ecosystem management to climate and other changes, and the possibility of thresholds leading to, for example, ecosystem col- lapse, extinctions, or regional-scale mortality events. Long-term datasets are required to investigate paleovegetation shifts and rates of ecosystem migration. New research is needed to separate the effects of climate from other changes, such as those caused by inva- sive species, pollution, land-use change, landscape fragmentation, or the loss of high-level predators. â¢ The net impact of increased CO2 levels, including ocean acidification, in combination with other stressors (pollutants, nutri- ent deposition) on ecosystems, especially for important ecosystems such as tropical forests and coral reefs. â¢ The consequences of changing ecosystems on climate feedbacks and human vulnerabilities. This requires assessment of
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 65 the full suite of climate feedbacks, including carbon, surface en- ergy and water balance, aerosols, and clouds. Key uncertainties are associated with areas vulnerable to rapid change in land cover, in- cluding the Arctic and tropical regions. â¢ The human behaviors associated with natural resource use, such as harvesting and land-use change, that affect trophic struc- tures and change ecosystems, and how these behaviors change in response to ecosystem stress or change. The effects of management strategies on climate, ecosystem services, and the resilience of eco- systems to climate change will need to be assessed. Field experiments and models can be designed to learn about coupled human- and environmental systems and to test different manage- ment interventions (i.e., adaptive management). â¢ The valuation of ecosystem services, including the eco- nomic and other costs associated with impacts of climate and other environmental changes. â¢ How managed ecosystems function, including those asso- ciated with growing urban areas, and how to improve the provision of ecosystem services in human-dominated landscapes. We need to understand how to manage the trade-offs between services with direct contributions to local human livelihoods and those with more indirect or global contributions, such as carbon sequestration and biodiversity preservation. â¢ Adaptive approaches and institutional and governance mechanisms for addressing the regulatory aspects of special status species management. To make significant progress, the research will have to be sup- ported by observation networks to document ecosystem changes over time and across types of ecosystems and human interactions through continuous measurement of variables such as greenness, land cover, and ocean color. An ocean observatory network (ORION Executive Steering Committee, 2005) and a continental- scale land observatory network (Keller et al., 2008) have been planned and are moving into implementation, although it may be years before operational systems are created that can be used to forecast environmental changes and their effects on biodiversity, coastal ecosystems, and climate. Improved models with better process understanding are needed (1) to evaluate the combined
66 RESTRUCTURING FEDERAL CLIMATE RESEARCH impacts of climate and other environmental stressors on a range of ecosystems at regional scales, and (2) to integrate and evaluate human drivers of and responses to environmental change in dy- namic feedbacks with ecosystem models. Finally, tools are needed to inform adaptive management strategies and estimate the resil- ience of various ecosystems under scenarios of climate, nutrient, water, and human systems change. HUMAN HEALTH Climate change has been called the greatest regressive tax in history, with the populations imposing the least stress through their greenhouse gas emissions experiencing the greatest health impact and vice versa (Figure 2.7). This relationship exists because, for the most part, climate change does not create new diseases and other health risks, but exacerbates existing ones. Poor health not only amplifies vulnerability, but also reduces the ability of com- munities and individuals to cope with or adapt to climate and other stresses (IPCC, 2007c, Chapter 8). Substantial inequalities in cop- ing capacity exist, and perhaps are growing, worldwide, including in the United States, where poor, elderly, uninsured, and minority populations are much more vulnerable. In the United States, a ro- bust public health infrastructure, such as sanitation and wastewater treatment facilities, has proven the best defense against adverse health effects from climate change (Gamble, 2008). The most authoritative assessment of the impacts of climate change on health to date was done in conjunction with the Com- parative Risk Assessment (CRA) project of the World Health Organization (Ezzati et al., 2004; McMichael et al., 2004). It found that for only five outcomes (malnutrition, diarrhea, malaria, flood injuries, and cardiovascular disease), 160,000 premature deaths annually could be attributed to climate change in 2000, or 0.4 percent of the global burden of disease. Some 88 percent of the attributable burden fell on poor children because of their exist- ing vulnerability to diarrheal diseases, malnutrition, and malaria. Climate change impacts in the second CRA assessment are ex- pected to be significantly larger, even though the base year will only be 5 years later (2005).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 67 FIGURE 2.7 Cartogram of climate-related mortality from malaria, mal- nutrition, diarrhea, and inland flooding per million people in the year 2000. The sizes of the regions are proportional to the increased mortality. SOURCE: Patz et al. (2007), Figure 1b. Reproduced with kind permission from Springer Science and Business Media. Climate change health impacts are divided into five categories: 1. Direct impacts through changing weather patterns (e.g., storms, floods, temperature extremes) 2. Indirect impacts through changes in water supply, water quality, and air pollution, and in ecosystems leading to shifts in disease vectors 3. Systemic impacts through shifts in food supplies, refugee patterns, coastal and agricultural livelihoods, and societyâs re- sponses to climate change, such as geoengineering, carbon taxes, and biofuel production 4. Low-probability high-consequence impacts, such as ex- tremely rapid climate change or sea level rise 5. Cobenefit impacts (sometime called âno regretsâ strate- gies), in which climate mitigation efforts are chosen to help protect health by reducing health-damaging air pollution emissions, low- ering the vulnerability of poor populations, improving the built environment, and other means In general, the ability to quantify the size and distribution of health impacts using standard biomedical tools such as validated exposure models and epidemiology is highest for category 1 and declines for
68 RESTRUCTURING FEDERAL CLIMATE RESEARCH categories 2 and 3. Category 3 impacts may be the most important for health over the long run. Few attempts have been made to quantify effects in category 4. The fifth and more positive type of impact is also of substantial research and policy interest. Direct Health Threats Changes in weather and storm patterns will engender many of the most serious climate change threats to the health of the U.S. population. Heat waves are expected to become more intense, more frequent, and to last longer (IPCC, 2007b, Chapter 10). High temperatures and humidity can cause death or chronic illness from the after-effects of heat stress. Outdoor workers in the construc- tion, agriculture, forestry, and fishing industries appear to be at particular risk. A heat wave in the midwestern United States led to approximately 600 heat-related deaths in Chicago over a period of 5 days in July 1995. Tens of thousands died in the 2003 heat wave in Europe (Box 2.9). These events have taught us much about hu- man vulnerability and highlighted the need both for factoring climate information into public health systems and for integrating climate, social, and health research. Climate change is expected to increase the risk of intense pre- cipitation events and flooding (IPCC, 2007b, Chapter 10). Floods can overwhelm preparatory and coping systems, even in regions with long experience in flooding, as the 2008 floods in the mid- western United States revealed. In 2003, 130 million people were affected by floods in China alone. Rarely included in such data are the secondary deaths that follow from the unsafe and unsanitary conditions in the wake of floods. Hurricanes are likely to become more intense with climate change, with impacts over broader scales (IPCC, 2007b, Summary for Policy Makers). Water supplies were contaminated with oil, pesticides, and hazardous wastes in the aftermath of Hurricane Katrina in 2005 (Manuel, 2006). Contamination of water supplies with fecal bacteria has led to diarrheal illness and some deaths fol- lowing several hurricanes. As with heat waves, richer and poorer communities have different vulnerabilities (Adger et al., 2005). The insurance and reinsurance industry will be challenged to ade- quately cover the risk of such major events.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 69 BOX 2.9 2003 European Heat Wave In August 2003, a heat wave raised temperatures to more than 40Â°C (104Â°F) in France for 7 days (Figure), causing more than 14,800 deaths. Other European countries, including Belgium, the Czech Republic, Germany, Italy, Portugal, and Spain, also reported excess heat-related deaths, with total deaths of approximately 35,000 (IPCC, 2007c, Chapter 8). Older people were especially vulnerable, with about 60 percent of the deaths in France occurring in persons 75 years of age and older (Vandentorren and Empereur-Bissonnet, 2005). The lack of air-conditioning, inexperience coping with very high temperatures (e.g., need for hydration), and absence of nearby relatives were contributing factors. The extreme heat also caused other harmful exposures, such as increased tropospheric ozone and par- ticulate matter. A French parliamentary inquiry found that existing systems for surveillance of heat wave stress were inadequate as were deficiencies in public health systems. Since that event, European governments have improved risk management systems, including better health warning and care of the elderly. Such heat waves are likely to increase with climate change and to further stress public health systems in a number of countries. FIGURE Differences in daytime land surface temperatures in Europe in 2003 from temperatures measured by the Moderate Resolution Imaging Spectroradiometer in 2000, 2001, 2002, and 2004. SOURCE: Image by Reto StÃ¶ckli, Robert Simmon, and David Herring, NASA Earth Observatory. Available at http://www.iac.ethz.ch/staff/ stockli/europe2003/. ______________________________ SOURCES: Lagadec (2004), IPCC (2007c).
70 RESTRUCTURING FEDERAL CLIMATE RESEARCH Droughts lead to regional water scarcity, salinization, disrup- tion of food systems, and increased plant infectious diseases or pests, and affects human health through malnutrition, infectious diseases, and respiratory diseases. In the United States, the main health risks from drought depend on (1) the effects of temperature on the incidence of diarrheal disease (IPCC, 2007c); (2) linkages between water availability, household access to improved water, and the health burden associated with a range of diseases; and (3) the effects of temperature and runoff on the microbiological and chemical composition of water supplies. Indirect Health Threats Indirect effects of climate on health are linked with changes in ecosystems and include air quality, allergens, and vectorborne in- fectious and parasitic diseases. Concentrations of ground-level ozone associated with climate change are increasing in many re- gions and are implicated in pneumonia, asthma, other respiratory diseases, and premature mortality. Concentrations of other air pol- lutants, particularly fine particulate matter, may increase in response to climate change. Climate change may increase the fre- quency and severity of fire events, releasing toxic gaseous and particulate air pollutants, and possibly increasing the long-range transport of air pollutants such as aerosols, carbon monoxide, ozone, mold spores, and pesticides. Climate change has already caused an earlier onset of the spring pollen season in the northern hemisphere. Changes in the spatial distribution of natural vegeta- tion may favor the growth of invasive plant species that cause allergies, such as ragweed. Shifts in the natural reservoirs of disease vectors, such as mos- quitoes, rodents, marine algae, birds, and deer, are already resulting in greater human exposures in some parts of the world, perhaps including the United States. Diseases such as malaria may return, although careful maintenance of the U.S. public health in- frastructure may prevent them from becoming significant (Gamble, 2008). Dengue fever and Lyme disease may also rise in the country due partly to climate change. Climate change can shift the distribu- tion of tick and mosquito vectors of disease (IPCC, 2007c, Chapter 8).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 71 Diseases transmitted by rodents may also increase during heavy rainfall and flooding events. Systemic Health Impacts Although not well suited to standard assessment methods, sys- temic impacts are the most worrisome because of their potentially pervasive impacts on health. Changes in agricultural production due to climate change will likely lead to an increase in malnutri- tion, which is a major health risk globally. Malnutrition was the largest category of ill-health related to climate change in the 2004 CRA. The linkage is not straightforward, however, as malnutrition is also influenced by economic, social, and governmental factors, which vary in time and space. Similarly, shifts in refugee and other migration patterns due to sea level rise, persistent droughts, changes in agriculture, and other climate-related stressors are sig- nificantly mitigated or enhanced by specific social, economic, political, security, and geographic circumstances. Migrants and refugees exhibit substantially different health patterns and needs for public health services and medical care than do stable popula- tions. The sustained ability to earn a livelihood is an important determinant of family health in all societies. Health may also be affected by efforts to mitigate or adapt to climate change. For example, although carbon taxes may reduce emissions, they also can create âenergy povertyâ in the developed world, in which poor people are not able to afford to heat or cool their residences, and drive households in poor countries back to polluting solid fuels. Burning solid fuels is already responsible for 1.6 million premature deaths annually, twice as many as all urban outdoor air pollution. Biofuel subsidies, which can contribute indirectly to malnutrition, and geoengineering schemes such as injecting aerosols into the atmosphere are likely to have wide- spread impacts on health. Reliable assessments of these impacts, however, require cooperation and development of methods and databases across a number of disciplines.
72 RESTRUCTURING FEDERAL CLIMATE RESEARCH Low-Probability High-Consequence Impacts Several threshold changes might be triggered by climate- changing pollutants, including runaway methane emissions from the ocean floor or tundra, rapid melting of large ice sheets, and shifts in major ocean currents. Given the potential speed and mag- nitude of their impacts on climate and sea level, the resultant health impacts and those of societyâs responses could be large. Systematic assessments would require adapting probabilistic risk assessment methods used in other realms to these large complex systems. Cobenefits Impacts Given the substantial resources that may be required for cli- mate mitigation, research aimed at understanding the impacts of reducing both greenhouse and other health-damaging pollution would be beneficial. Examples of research topics that may lead to measurable and cost-effective cobenefits for human health include: â¢ The effects of improved combustion methods on air qual- ity and health â¢ The effects of different modes of transportation (e.g., walking, public transportation, driving) on air pollution, traffic, and obesity risks â¢ Ways in which the use of energy efficient materials and design affect household (and urban) environmental risks and en- ergy demand â¢ The relationship between food prices, greenhouse gas emis- sions, and the health impacts of dietary choices â¢ The effects of alternative land use practices (e.g., refores- tation, cultivation of biofuel plants) on human welfare and disease â¢ How changes in the use of contraception may affect health, population, and resource consumption Showing which greenhouse mitigation efforts can yield short- term health and other benefits, even if they are intended primarily for protection from climate changes decades in the future, would improve the attractiveness and political viability of these in- vestments.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 73 Research Needs Systematic assessments of current and future health risks from climate change are needed to help understand the total impact of climate change and thus to guide mitigation and adaptation efforts. Of particular need is a more complete understanding of the uneven pattern of health risks, both within and between populations, which are expected to have highly unequal impacts. High-priority research and health impact assessment activities include: â¢ The readiness of the nation to predict and avoid public and occupational health problems caused by heat waves and severe storms â¢ The potential U.S. health threats from changes in the pat- tern of disease vectors, such as birds, rodents, and mosquitoes under different scenarios of climate-induced ecosystem change â¢ Characterization and quantification of relationships be- tween climate variability (trends or fluctuations in temperature, precipitation, or other weather parameters), health outcomes, and the main determinants of vulnerability and equity within and be- tween populations (location, age, health status, etc.) â¢ Development of reliable methods to connect climate-related changes in food systems and water supplies to health under differ- ent conditions â¢ Estimation of disease burdens in all the main categories (direct, indirect, systemic) attributable to current climate change, from the global to the subnational level â¢ Prediction of future risks in response to climate change scenarios and of reductions in the baseline level of morbidity, mor- tality, or vulnerability â¢ Development and application of systematic standardized methods for assessing cobenefits, including associated economic evaluation â¢ Development of robust and sophisticated assessment meth- ods for evaluating the health cobenefits and/or adverse impacts of mitigation measures such as biofuels, multimodal transportation, and geoengineering â¢ Identification of the available resources, limitations of, and potential actions by the current U.S. health care system to prevent,
74 RESTRUCTURING FEDERAL CLIMATE RESEARCH prepare for, and respond to climate-related health hazard and to build adaptive capacity among vulnerable segments of the U.S. population Risk assessments are needed to address these aims, including well-established methodsâsuch as time-series studies to describe the current relationships between meteorological variables and health risksâand rapidly developing fields, such as empirical and biological modeling of climatic and other factors affecting the dis- tribution of infectious diseases. Of particular difficulty and importance are hybrid models or protocols that effectively bring these two types of assessments into a common framework. IMPACTS ON THE ECONOMY OF THE UNITED STATES The Kyoto Protocol set binding targets for 37 industrialized countries and the European Community to reduce greenhouse gas emissions. President Bush did not support signing the agreement in 2001 because it âwould cause serious harm to the U.S. econ- omy.â12 The United States has a new president, and economic impacts of climate change are of high near-term policy relevance. The economic impacts from greenhouse gas mitigation policies are among the most important unknowns in the climate policy debate. Broadly speaking, there are two mechanisms by which climate change has a fundamental impact on the economy of the United States (and on the world economy). First, economic activities that depend on climate (e.g., agriculture) are affected by a change in the climate, and for large climate changes, that effect will undoubtedly be negative. Whatever the damages, they are expected to rise rap- idly as the magnitude of climate change increases. For example, one study (Nordhaus, 2008) estimates that the annual economic damages from a 2.5Â°C temperature increase are only 20 percent of the annual damages from a 6Â°C temperature increase. There is, of course, a great deal of uncertainty in the likely damage caused by climate change, as discussed below. These damage estimates typi- 12 Letter from G.W. Bush to Senators Hagel, Helms, Craig, and Roberts, March 13, 2001, available at www.whitehouse.gov/news/releases/2001/ 03/20010314.html.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 75 cally assume that adaptation will be pursued to soften the potential impacts of climate change. Adaptation itself will involve costly actions, but if adaptation is not pursued, damages would be higher. For larger increases in temperature, the bulk of the damage is ex- pected to occur through unanticipated and abrupt change. One of the biggest impacts from climate change on the econ- omy of the United States will be through coastal flooding. The damage and disruption that accompanies hurricanes and other se- vere weather events will be magnified by a rise in sea level. Nearly every sector of the economy as well as the welfare of individuals will be affected in some way by climate change. The second way that climate change will affect the U.S. econ- omy is the cost of reducing greenhouse gas emissionsâmitigation. Although there may be pleasant surprises as emissions of green- house gases are reduced (such as energy-saving innovations or companies that do better than expected in achieving reductions), there will be costs and those costs will be borne by everyone. Higher prices for energy and energy-intensive goods are usually needed to reduce consumption. People will reduce their energy consumption and carbon generation, but not entirely painlessly. The more slowly emissions are reduced, the easier it will likely be. For instance, allowing more time to reduce emissions avoids pre- mature retirement of energy-inefficient capital. Of course, the down side is the delay in reducing emissions. A primary tool for evaluating policy options is integrated assessment models (Box 2.10). Research Needs Despite work cited here, in IPCC reports, and elsewhere, our knowledge of the economics of climate change is surprisingly incomplete and imprecise. Given that we are making decisions on trillion-dollar investments to control greenhouse gases based on what we know now, it seems clear that gaining a better under- standing of the economics of climate change should be a high social priority. Many of the economic research problems associ- ated with climate change can be categorized into five broad issues: mitigation of greenhouse gases, regulatory response, im- pacts of climate change, incidence, and adaptation. Other issues
76 RESTRUCTURING FEDERAL CLIMATE RESEARCH BOX 2.10 Integrated Assessment Models of the Climate and the Economy Integrated assessment models have become one of the most use- ful and well-developed approaches to examining the climate problem and what to do about it. The concept is simple. A pure climate model represents how the climate will evolve given exogenous drivers from the economy, where emissions originate. A pure economic model of climate treats the consequences of emissions as exogenous. An inte- grated assessment model captures in a compact fashion how the climate evolves in response to emissions, how the changed climate impacts economic activity in the world, and how those impacts in turn are combined with mitigation costs to affect policy and the evolution of the economy. Integrated assessment models differ in the level of detail on climate and/or the economy and in the level of closed feedback between climate evolution and economic evolution. One of the earliest integrated assessment models was developed in the 1970s by Edmonds and Reilly (1983). Other early examples in- clude models developed by Manne and Richels (1991) and by Nordhaus (1977, 1991). The mid 1990s saw major progress on this front, with the development of the Dynamic Integrated Climate Econ- omy (DICE) model (described in Nordhaus, 1994) and other more advanced models (see review in IPCC, 1995). In the DICE model, the atmosphere is represented by a two-box dynamic model and the econ- omy is represented by a single sector. The decision variables are capital investment and investment in mitigation, and all else flows from them. Over the past 15 years, a good deal of progress has been made in developing more sophisticated integrated assessment models. The primary advance has been to better represent regional differences in the models, rather than view the world as a single economy with aver- age climate impacts. Furthermore, the number of integrated assessment models has multiplied. A recent comparison of global climate-economy models involved 19 different models and modeling groups (Weyant et al., 2006). Although there are many dimensions on which integrated assessment models can be improved, one of the most important is improved data and understanding related to underlying costs, benefits, and economic processes. that touch on climate, such as discounting and uncertainty and risk, are not discussed here. A review of some of the issues in the economics of climate change can be found in Kolstad and Toman (2005) and Heal (2009).
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 77 Mitigation The cost to reduce carbon emissions in particular sectors by particular amounts is subject to a great deal of uncertainty, in both the short run and the long run, and for consumers as well as busi- nesses. Although there have been a number of studies of this problem, a great deal of uncertainty remains. Some analysts suggest a low or negative cost for significant reductions; others suggest sig- nificant positive costs (see the discussion in Fischer and Morgenstern, 2006). For example, the state of California estimated that costs of reducing emissions would be negative (there would be a savings due to mitigation), a result roundly criticized by peer reviewers.13 A re- cent comparison of economic models found that the cost of controlling an extra ton of carbon in 2025, assuming policies to limit greenhouse gas concentrations to double preindustrial levels, ranged from $2.8 per ton to $482 per ton (Weyant et al., 2006). Estimates of the costs to reduce carbon emissions have been produced for the economy as a whole (e.g., Nordhaus, 2008) as well as at the sectoral level, particularly by the IPCC. For example, IPCC (2007d) suggests that substantial emission reductions can be obtained in the building sector at negative cost; other negative cost opportunities exist in other sectors. However, the IPCC estimates are neither specific to the U.S. context nor comprehensive, and they do not deal with the rate of change of mitigation as it affects costs. The Economics of Climate Change attempted to quantify both the costs and benefits of mitigation (Stern, 2006). However, the data underlying the analysis are sparse (see Symposium on Stern Review in the Winter 2008 issue of the Review of Environ- mental Economics and Policy). The United States has successfully reduced emissions of air pol- lutants such as SO2 (Ellerman et al., 2000). However, the policy challenge of reducing CO2 emissions is economically different in two ways: (1) CO2 emissions come from many diverse sources throughout the economy, whereas the bulk of SO2 emissions came from a few hundred electric power plants, and (2) behavioral change and technological innovation are both likely to play a more profound role with CO2 reduction because carbon is integral to fossil fuel, 13 See http://www.arb.ca.gov/cc/scopingplan/economics-sp/peer- review/peer-review.htm.
78 RESTRUCTURING FEDERAL CLIMATE RESEARCH whereas sulfur is a contaminant that can be removed. Therefore, it is important to develop a better understanding of the determinants of behavioral change and technological innovation. A research program including cost engineering studies, econometrics, and field experi- ments (e.g., artificially changing rate structures and observing how behavior changes) would seek to answer questions such as the short- run and long-run marginal cost of reducing CO2 emissions by vari- ous levels for the automobile industry, and ways to accomplish the reduction most effectively (e.g., by changing vehicle design or the CO2 content of fuels, reducing miles traveled per vehicle). Regulatory Response Emission reduction responses depend on policies such as fuel efficiency standards, fuel taxes, feebates,14 technology-push regu- lations,15 and cap-and-trade systems. Experience with cap-and- trade systems is limited to the European Trading System for Car- bon (see papers in the Winter 2007 issue of the Review of Environmental Economics and Policy), the U.S. sulfur trading sys- tem (e.g., Ellerman et al., 2000), and a number of localized trading systems. We have learned a great deal about these economic incen- tive systems, although our experience with economy-wide trading systems is limited. Other economic incentives are in use as well as prescriptive regulation (Freeman and Kolstad, 2006) and regula- tions that rely on voluntary actions (e.g., Morgenstern and Pizer, 2007). Little experience exists for carbon regulation (Box 2.11), which will be fundamentally different from many previous regula- tory regimes in that behavior as well as technology will be affected. For example, if electric utilities face a price of carbon permits equal to $100 per ton of CO2, what investments in renew- able energy can be expected? How will drivers and automobile manufacturers respond to an upstream (regulation at the energy producer) cap-and-trade system versus a carbon tax or a down- stream (regulation at the energy consumer) cap-and-trade system with a similar carbon price? 14 A feebate involves a rebate to above-average performers, financed by a fee on below-average performers, so that no net revenue is collected. 15 A technology-push regulation is one designed to spur innovation and expand the menu of technological options.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 79 BOX 2.11 The Energy Price and Economic Effects of Reducing U.S. Carbon Emissions A number of bills have been introduced in the U.S. Congress to limit the emissions of greenhouse gases over the coming decades (e.g., see Appendix A). As of January 2009, none have passed, in part because of questions about how much reducing greenhouse gas emissions will cost and what will happen to energy prices as a result. Virtually all the pro- posed legislation relies in large part on a cap on emissions of greenhouse gases nationwide, implemented through a system of tradable emissions allowances. A leading proposal in the most recent (110th) Congress was the Lieberman-Warner Climate Security Act of 2007. One of its key fea- tures was a cap-and-trade system, capping greenhouse gas emissions 7 percent below 2006 levels beginning in 2012, gradually tightening to 29 percent below 2006 levels by 2030. A detailed analysis by the U.S. Energy Information Administration found that most of the emission reductions would come from the electric power sector via changes in the way electric- ity is generated (EIA, 2008a). The effects on price would be too modest for consumers to strongly reduce energy consumption. Gasoline prices were assumed to be 10 to 20 percent higher in 2020 and 20 to 40 percent higher in 2030 than in the reference case. Although these are not trivial increases, they are within the variation in prices consumers experienced in 2008. Losses in total national economic output (gross domestic product) would be less than 1 percent in 2030. The EIA (2008a) analysis reports precise dollar figures for the conse- quences of reducing greenhouse gases, but there is considerable uncertainty regarding many of the assumptions and conclusions emerging from this report and others like it. The critical nature of the potential impacts on the U.S. economy illustrates the importance of research to better understand the economic impact of greenhouse gas regulations. Damage from Climate Change Costly damages to society are among the consequences of cli- mate change. These costs are poorly understood from a physical point of view, let alone an economic point of view. Figure 2.8 summarizes several studies of the damage to the overall economy from a change in the global mean temperature. It is important to emphasize that the figure suggests more precision in these estimates than is warranted. For instance, for moderate temperature changes (e.g., less than 3Â°C), the estimates are similar, suggesting consen- sus. However, there is little consensus regarding the damage from modest climate change. In fact, the degree of uncertainty of climate impacts on the economy is generally considered to be very large.
80 RESTRUCTURING FEDERAL CLIMATE RESEARCH FIGURE 2.8 Some estimates of the global damage from a change in the global mean temperature, as used in several integrated assessment models of climate policy. SOURCE: Dietz and Stern (2008). Reproduced by permission of Oxford University Press. Adapted from Smith et al. (2001), Figure 19-4. A number of studies have focused on impacts for individual economic sectors. Agriculture costs have been studied most, but many important sectors of the economy have received virtually no attention (Mendelsohn et al., 1994; Mendelsohn and Neumann, 1999). The effects of warming can be mixed, bringing benefits in some cases and costs in others. If the temperature rises when it is cold, there can be less crop damage from freezing, less energy needed for heating, and fewer deaths from cold. However, if the temperature rises too much when it is already hot, crop damage can be severe, more energy is needed for air-conditioning, and some will die from heat waves. The net impact of a given climate change scenario can therefore be quite ambiguous. There is some evidence that substantial damages from climate change may be associated with extreme weather events, but such events (by definition) are rarer and less well studied in both the natural and social sciences (see the âExtreme Weather and Climate Events and Disastersâ sec- tion). Understanding the damages from temperature extremes is a crucial issue that will likely require more refined spatial and tem- poral detail than exist in most datasets.
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 81 Incidence Aggregate net benefits (market plus nonmarket environmental benefits minus market plus nonmarket costs) is not the only metric to use in evaluating greenhouse gas policies. The distribution of net benefits and costs from controlling greenhouse gas emissions or from the impacts of climate change and adaptation also has ramifications for environmental justice. For instance, with a cap- and-trade system covering the entire U.S. economy, what income groups end up paying for the costs of greenhouse gas regulation and where do job losses and gains occur? Although some work has been done on who ultimately pays and/or benefits (the incidence) from environmental regulations generally (Metcalf, 1999; West and Williams, 2004), this topic remains largely uninvestigated. This literature generally finds carbon taxes to be moderately re- gressive. Greenhouse gas regulation will reduce energy consumption and thus, in all likelihood, emissions of associated non-greenhouse- gas pollutants. The levels of changed emissions of these copollut- ants are poorly understood as are the monetary benefits of the decreased levels of copollutants (the cobenefits). For instance, what reduction in conventional air pollutants can be expected in urban areas as a result of greenhouse gas regulations? Although some work has been done on this question (e.g., Wier et al., 2005), research is needed to better understand the interplay between co- pollutants and greenhouse gases from a regulatory perspective. Adaptation Economic analyses outside the climate arena consider adaptation primarily in the context of price changes. When the price of gasoline goes up by $1, people may adapt by driving a more fuel-efficient car, moving closer to work, or modifying their driving habits. One of the earliest papers on the economics of adaptation focused on investments in irrigation as a way of adapting to uncertainty over precipitation (McFadden, 1984). Such defensive expenditures can blunt the damage from climate change. The nature of the adaptation depends on the speed of the change.
82 RESTRUCTURING FEDERAL CLIMATE RESEARCH People and businesses will similarly adapt to climate change (e.g., Reilly and Schimmelpfennig, 2000; Kelly et al., 2005; IPCC, 2007c; Mansur et al., 2008), although the magnitude and speed of that adaptation are not well understood. When farmers perceive a changed climate, they will change their agricultural practices; when individuals see their local climate become less hospitable, they may migrate to better climes. This aspect of adaptation is autonomous, since it will occur naturally without government in- tervention. In contrast, changing power lines, water systems, levees, and other public infrastructure to withstand climate change may require complex government action and thus governmental planning and decision making. This sort of adaptation can be called public adaptation and it will not occur automatically. Re- search is needed to better understand both the private and public adaptation processes so we can better estimate the costs and dam- age from climate change policies. Furthermore, the timing of public adaptation is important for public policy. WHERE DO WE GO FROM HERE? Climate change is having an impact on basic human require- ments, such as water, food, and health. These impacts will become larger in the coming decades. A research program that integrates across the many dimensions of this issue is needed (1) to guide the nation in the multiple choices it faces to reduce the costs and risks of these impacts, and (2) to provide early warning of changes that are abrupt and large enough to push climate and human systems past tipping points. The nation must prepare itself for the possibil- ity of warming in excess of 3Â°C by the end of the century, followed by the disappearance of most alpine glaciers, the rapid disintegration of the Greenland Ice Sheet, and a rise of sea level of up to several meters (cf., Figure 2.1). It must also prepare for in- tense severe weather and heat waves, which stress the nationâs ability to provide needed water supplies. Such stresses need to be considered in the context of other stresses almost certain to be oc- curring, such as economic changes, changes in the global market, and potential international conflicts. Preparations will require the integration of models and observations at a much more advanced
RESTRUCTURING THE CLIMATE CHANGE SCIENCE PROGRAM 83 level than is possible now, as well as the knowledge that comes from linking research on the natural climate system with research on hu- man drivers and responses, and factoring in the needs of decision makers in the research agenda. This in turn requires maintaining a strong natural science research component while strengthening hu- man dimensions research and developing more fruitful interactions with decision makers. The societal issues discussed above provide a framework for human dimensions research. But given the historic emphasis of the program on the natural sciences, a focused effort on key aspects of the human dimensions is also needed to speed pro- gress and further develop the research priorities. The key elements of a research program aimed at understanding climate change and supporting climate-related decisions are discussed in Chapter 3.