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.
4 Information Needs for Decision Support T he goal of the U.S. Global Change Research Act (USGCRA) of 1990 is to âassist the Nation and the world to understand, assess, predict, and respond to human-induced and natural processes of global change.â This language makes it clear that the intent of the act is to foster both fundamental scientific investigations on global change and applied re- search designed to support appropriate responses to it. For climate change, the latter covers a range of mitigation and adaptation responses. Providing decision support to those who are in charge of the responses is essential for carrying out the purposes of the act, and to provide a sci- entific basis for this support, the nation needs to develop the science of climate change response, as a complement to the science of climate change processes. Understanding the physical dimensions of climate is a necessary but not sufficient condition for supporting climate change responses. Also needed are contributions from a wide range of disciplines including behav- ioral and social science disciplines that are not currently well represented in scientific programs on climate and its impacts. Chapters 2 and 3 address the process aspects of developing scientific support for climate-affected decisions. An important principle developed there is that decision support processes should take priority over informa- tion products, because unless attention is paid to process issues, particularly two-way communication between the likely producers and users of infor- mation for decision support, the products that are generated are unlikely to address decision makersâ needs. Of course, information content is also critical for sound decision making. Decision support processes need to yield understanding of what decision makersâ key information needs are and to 91
92 informing decisions in a changing climate lead to the development of information that is capable of supporting high- quality decisions. This chapter focuses on information needs for decision support, seen from the perspectives of decision makers. It emphasizes the need for re- search for decision supportâthat is, research that provides various types of decision-relevant information not currently provided by U.S. climate science programsâand basic and applied research on decision support. It highlights challenges associated with providing use-relevant data across spatial and temporal scales and across sectors, along with ways of overcoming those challenges. Short case studies are used throughout to illustrate information needs and approaches that are successfully engaging decision makers at local and regional scales. INFORMATION FOR DECISIONS Individuals and organizations facing climate-sensitive decisions are not often concerned with climate change per se, but rather with how it may affect their responsibilities, commitments, and priorities. Thus, information for climate-related decision support must be salient to their priorities, or it is unlikely to be helpful. It follows that decision support strategies should be built on an under- standing of decision makersâ values and priorities, as well as the constraints under which they operate. As highlighted in Chapter 2, this type of under- standing is best developed through interaction between the decision makers and those who would inform them. Usersâ needs are diverse and their data and information requirements are similarly diverse. In particular, they need information matched to the spatial and temporal scales of their agencies or organizations and concerning climate parameters that are meaningful to them; for an example, see Box 4-1. The types of information required for climate responses are many and varied, ranging from climate data to data on affected populations and eco- systems. Agencies and organizations that are responsible for responding to extreme climate events need to know what types of events to prepare for and the likely occurrence of such events as well as the potential effects on human populations, economic activity, and built and natural systems. Understanding these effects in turn requires knowledge about population characteristics, current and future settlement patterns, social vulnerabil- ity, trends within national, regional, and local economies, and ecological variables. Information is required for a wide range of potential mitigation and adaptation strategies. Mitigation decisions may center on ways of reducing greenhouse gas emissions, decreasing atmospheric greenhouse gas concen- trations, and changing land cover. On the adaptation side, decisions focus
INFORMATION NEEDS FOR DECISION SUPPORT 93 BOX 4-1 The Pileus Project The Pileus Project, conducted by researchers at Michigan State University, began as part of the U.S. National Assessment/Great Lakes Regional Assess- ment, with funding from the U.S. Environmental Protection Agency. Its objectives were to identify, with stakeholder assistance, the influence of climate on Michigan agriculture and tourism industries; create models to quantify the impacts of past and projected climate variability and change; and develop decision-support tools for climate-related risk management. The project focused on one agricultural productâtart cherries, a crop that is extremely vulnerable to temperature extremes and also very important to Michiganâs agricultural economy and to the nation, since Michigan provides more than 70 percent of the U.S. supply. Stakeholders provided input on assessment goals, identified information needs, provided expertise and data, and evaluated the decision support tools developed by the project. A suite of web-based tools was developed that included a historical climate tool, downscaled precipitation scenarios, a future scenarios tool, and tools to aid decision makers with respect to their future crop investments (see http://www.pileus.msu.edu/agriculture/tc_tools. htm). The Pileus Project officially ended in August 2007, but work continues with support from the National Science Foundationâs Human and Social Dynamics Program. The discussions with stakeholders revealed specific kinds of information they wantedâfor example, the expected date of the last spring frostâthat was not available from existing climate models. A key lesson of the project was that ad- dressing decision makersâ needs frequently requires the development of new forms of data. SOURCE: Presentation by Jeffrey Andresen and Julie Winkler, Department of Geography, Michigan State University; available at http://www.pileus.msu.edu/. on reducing the climate-related vulnerability of human systems and activi- ties, improving the ability to respond to damage caused by extreme climate events, and encouraging people to take the future impacts of climate change into consideration in their own decision making. Decision makers also face choices with respect to the design and implementation of institutions and policies to enhance both mitigation and adaptation activities. Those kinds of decisions require information about climate, but they also require a wide range of other types of information. Mitigation strate- gies designed to reduce greenhouse gas emissions from motor vehicles, for example, may need information on the most effective incentives for automobile manufacturers and purchasers, on appropriate urban design approaches, and on how to combine incentives, regulations, and infor-
94 informing decisions in a changing climate mation into effective policies. Also important are public opinion data on environmental concern and attitudes about fuel-efficient vehicles. Decisions regarding changes in agricultural practices depend on detailed information on how climate change affects growing seasons and cropsâthe kind of information sought by Pileus Project investigatorsâalong with knowledge concerning both more robust and alternative crops. Decisions on infrastruc- ture improvements for flood protection require information from sources as diverse as civil and structural engineering, infrastructure life-cycle analysis, environmental impact assessment, demography, public finance, and law. Example: Natural Hazards Loss Estimation Experience with natural hazards illustrates how diverse information sources are often needed for decision support. Hazard impact and loss modeling uses data on characteristics of the natural and built environment, provided by environmental scientists, engineers, and community building and planning departments; data on populations at risk, provided by demog- raphers, geographers, urban planners, and other social scientists; algorithms developed by modelers; and data on direct and indirect economic and social effects, provided by economists, public health researchers, and other social scientists. The hazard-related decision support software tool that is most widely used in U.S. communities is HAZUS-MH (Hazards United States, Multi- Hazard Version), which was developed by the National Institute of Building Sciences with funding from the Federal Emergency Management Agency. HAZUS tools and modules enable users to anticipate the physical, social, and economic effects of earthquakes, floods, and wind hazard events, including building damage, earthquake-induced fires, lifeline failures, the hardest hit geographical areas and population groups, direct losses, indi- rect economic losses, and the size of populations displaced by such events. Geographic Information Systems (GIS) provide an integrating platform for simultaneously analyzing different information inputs. HAZUS findings can be used to support decisions related to land use, building codes, evacua- tion planning, disaster response, and predisaster planning for postdisaster recovery (for more information, see http://www.hazus.org). HAZUS was developed with federal government funding primarily for use by public entities, but private firms also engage in extensive modeling efforts, particularly for use by insurers and reinsurers. Some of these firms have moved into modeling the impacts of terrorist events and large-scale catastrophes and are now focusing their modeling efforts on the climate- related events. Hazard loss modeling provides several lessons that have implications for the development of climate change decision support strategies and tools.
INFORMATION NEEDS FOR DECISION SUPPORT 95 First, successful models seek to assist diverse decision makers by answering a wide range of questions, such as: â¢ In the next hurricane, how soon must evacuation orders be issued, when might evacuation routes become blocked by flooding, and what seg- ments of the population will need evacuation assistance? â¢ How many residents will require shelter after a disaster, for how long, and what can be expected in terms of the demographic composition and needs of shelter populations? â¢ How much will a particular utility lose or save over the next 30 years by mitigating earthquake hazards in a high-hazardâor lower- hazardâregion? â¢ In the next earthquake, how many people are likely to die and how many will require hospitalization? â¢ What is the magnitude of a particular insurance companyâs portfo- lio risk for wind hazards, both globally and in particular regions? Second, modeling efforts are inherently multidisciplinary. For ex- ample, most California decision makers who try to reduce earthquake hazards have little interest in earth science and geophysics per se, but considerable interest in how the physical processes associated with earth- quakes interact with vulnerable environments and how they affect valued assets and human populations. California has experienced many large earthquakes that were not disasters because they did not hit population centers or disrupt important economic activities. Data on physical earth- quake effects become meaningful only in the context of data provided by other disciplines. Third, models enable both decision makers and the public to visualize how disasters will affect valued assets. In 2006, for example, a model of the recurrence of the 1906 San Francisco earthquake, developed to coincide with the 100th anniversary of the event, illustrated for various audiences the range of effects that would result today. In 2008, a similar impact modeling scenario was released for an earthquake on the Southern San Andreas Fault, which would affect a large region in Southern California. The scenario serves as the basis for extensive disaster exercises and public education ef- forts. The scientific details of how the earthquake will propagate along the San Andreas are less important for decision makers than information on the eventâs effects on hospitals, schools, power lines, transportation networks, hazardous materials sites, and populations at risk. Fourth, even though all elements in loss models contain uncertainty, and even though many modeling tools are quite crude by scientific stan- dards, the tools help decision makers anticipate and act to reduce hazard impacts. Tools such as HAZUS became widely used because they were
96 informing decisions in a changing climate marketed to decision makers and planners and because user groups were created and sustained through governmental action. Finally, loss estimation projects have been designed specifically to encourage action to reduce disas- ter losses and impacts and not to fund basic scienceâeven though scientists provide crucial data inputs. Other Examples As the above discussions show, useful information for responding to climate change requires climate information and many other kinds of infor- mation as well. The examples below illustrate the many types of data and information required to assess both climate impacts and the effectiveness of efforts to respond to a changing climate landscape. Citiesâ Efforts to Reduce Greenhouse Gas Emissions Approximately 700 mayors have endorsed the U.S. Mayors Climate Protection Agreement, and many cities have initiated large-scale climate change mitigation and sustainability programs. Such efforts require infor- mation to assess program effectiveness, costs, benefits, and both intended and unintended consequences of programs, as well as to set priorities among various mitigation and adaptation activities. Chicagoâs actions since 2000 include providing grants for plantings on rooftops and roadway medians, enhancing alternative transportation opportunities, retrofitting city buildings for energy efficiency, and encouraging energy efficiency in commercial, industrial, and residential buildings. Similarly, initiatives of the GREEN LA Program in Los Angeles range from producing electrical power from renewable sources to creating green space, implementing smart growth strategies, and reducing water consumption. Coping with Climate Change in New York City In New York, PlaNYC (see Appendix A) involves numerous mitigation and adaptation decisions by households, public and private organizations, and diverse economic sectors and authorities, spanning approximately 1,600 different governmental units. Climate-related information contained in global climate change models and regional climate scenarios based on downscaled data are needed to support those decisions. Decision makers also need other types of information, such as sociodemographic, economic, transportation, and building stock data; vulnerabilities of health, energy, coastal, and water systems; cross-sectoral interactions; and information on the effectiveness of a range of mitigation, adaptation, and sustainability strategies. PlaNYC activities also have monitoring and assessment com-
INFORMATION NEEDS FOR DECISION SUPPORT 97 ponents that require program evaluation data to encourage learning and improve program effectiveness. Adaptation in the Great Lakes Region Climate change predictions for the Great Lakes Basin point to warmer, dryer summers; shorter winters; more winter precipitation falling as rain; less ice; and irregular, higher intensity storms. This information becomes useful mainly as it is linked with other information to project how the physical climate changes will affect economic and other activities, includ- ing: the recreational infrastructure in the region (e.g., docks too high for use and navigational hazards from low lake levels); commercial shipping (e.g., ships will have to carry smaller quantities of cargo so that they can float higher); and the drying of wetland areas, which will affect wild rice crops and fisheries These consequences will in turn have an impact on jobs, livelihoods, and costs in a variety of economic sectors. New data will be needed to trace the effects of physical climate alterations on the economic and social activities affected by those alterations. Western Water Management Climate change confronts water managers in several western states with the prospect of serious droughts and decreased winter snowfall, leading to reductions in snowpack, which accounts for about 35 percent of Californiaâs usable annual surface water (California Department of Water Resources, 2006). Managers are considering major new investments in water storage and distribution infrastructure and policies to reduce demand. Some are asking how much reduction in water demand can be expected at what level of increase in water prices and as a result of public education programs. To consider these options, they need more careful monitoring of precipitation and snowpack, as well as better information about consumer response to incentives and information. Some managers also need information about the potential for saltwater intrusion into groundwater due to sea-level rise, and the ability of freshwater-bearing sediments to repulse intrusion. Wildfire Management Wildfire management strategies, such as decisions about where to al- locate and pre-position resources for fire prevention, prescribed burns, and fire suppression, rely on a similarly wide range of information. Some needed information is climate related, including ambient temperatures; precipitation amounts, frequencies, and timing; amounts and timing of snowpack melt; changes in speciation that affect land cover; changes in high wind frequency
98 informing decisions in a changing climate and severity; and changes in the probability of fire ignition by lightning. So- cial information is equally needed, including about intensive development in the wildland-urban interface, which increases fuel loads; policies related to the management of public and private lands in wildfire risk areas; and social perceptions of land value as influenced by human habitation, recreational uses, species richness, and aesthetic and cultural attributes. Ecosystem Management Climate change is expected to affect ecological systems in many ways (National Research Council, 2008a). Organizations that manage conserved land require information on how terrestrial ecosystems will change as cli- mate changes and what their conservation value will be after some species and habitats disappear and others replace them. Managers of marine mam- mals and fish concerned with determining sustainable rates of commercial, recreational, and subsistence harvest may soon require information about how species reproduction, growth, physiology, and migrations respond to changes of ocean water temperature, acidity, primary production, predator and prey populations, and change in hypoxic or anoxic zones (e.g., Chan et al., 2008). For anadromous species, changing water temperatures, water levels, flow rates, and seasonal timing of flows in streams and rivers cascade into changes in the availability of riparian habitat; water column stability and mixing; pollutant, nutrient, and oxygen concentrations; populations of other riparian species; and the prevalence of, and resistance to, diseases (National Research Council, 2004b). In northern latitudes, losses of stable ice cover may reduce the availability of refuge habitat for juvenile fish (National Research Council, 2004a). Climate-related changes can affect hu- man uses of riparian shorelands and water, which can then produce further impacts on anadromous populations (National Research Council, 2004e). Transportation Transportation decision makers find it difficult to obtain climate-related information relevant to planning and design in usable formats and at the appropriate spatial and temporal scales (National Research Council, 2008b). Issues include changes in winter weather, which accounts for 40 percent of highway operating budgets in northern states: in the frequency of hur- ricanes on the Gulf Coast; and in spring melting and permafrost in Alaska, which affect bridges and oil pipelines. Decision makers need locally specific information about such variables to select materials and designs for foun- dations, subsurfaces, and drains. They need accurate digital elevation maps in coastal areas to forecast effects of flooding and storm surges. Including climate change will require recalculations of innumerable transportation
INFORMATION NEEDS FOR DECISION SUPPORT 99 engineering standards, and this effort in turn will require extensive and costly research and testing. At present, transportation planners generally incorporate projected land use patterns into their decision making, but not proposed climate adaptation and mitigation efforts that could dramatically alter land use, which would require corresponding changes in transportation plans (National Research Council, 2008b). In many instances, transporta- tion professionals have not yet engaged with the scientific and agency com- munities that might develop and provide the needed information. Because the transportation sector produces the fastest growing rate of carbon dioxide emissions, it is important to consider not only the effects of climate change on transportation infrastructure, but also the effects of the infrastructure on climate change. In the long term, better community and transportation infrastructure planning can reduce vehicle-miles trav- eled, thus slowing climate change and facilitating adaptation to a carbon- constrained world. Heat Wave Warnings Heat waves cause substantial mortality and sufferingâmore than 700 deaths in Chicago in 1995, and perhaps 70,000 in the deadly 11-day 2003 heat wave in Europe. Effective warning systems can reduce heat-related mor- tality: The system in Philadelphia saved an estimated 117 lives in a 3-year pe- riod (Ebi and Schmier, 2005). However, the most useful weather parameters for predicting danger are still debated and may be location specific. It is not yet clear whether a high nighttime or daytime temperature is more dangerous, and different cities use different weather criteria for health decisions (Bernard and McGeehin, 2004). These include temperature-humidity indices, the num- ber of consecutive hot days, temperature combined with time of year (a heat wave early in the summer season is generally more lethal than one in mid- or late-summer), and parameters based on analyses of air mass parameters in relation to historical evidence of mortality rates (Kalkstein and Tan, 1996). The most effective heat health warning systems require reliable local weather forecasts and known doseâresponse relationships between climate conditions and health outcomes to allow appropriate activation and deactivation of re- sponse plans, as well as involvement and coordination of the proper agencies (Kovats and Ebi, 2006). Anticipating West Nile Virus Outbreaks Above-average temperatures are linked to transmission of West Nile Virusâespecially the more lethal strain that emerged in 2002âthrough increased replication in the major mosquito vector, Culex pipiens (Dohm and Turell, 2001; Dohm, OâGuinn, and Turell, 2002; Reisen, Fang, and
100 informing decisions in a changing climate Martinez, 2006; Kilpatrick et al., 2008; see Institute of Medicine, 2008, for more details). Thus, climate warming is expected to lead to increased outbreaks. Human and equine infection follows a known causal chain that determines the factors that require monitoring for anticipating outbreaks. These factors include early-season weather conditions (especially heat and dryness), mosquito abundance, mosquito infection, avian host populations and infection rates, and equine and human cases. Reducing Household Greenhouse Gas Emissions Homes and private motor vehicles account for nearly 40 percent of national carbon dioxide emissions in the United States and are therefore a major target for mitigation. The relevant decision makers include govern- ment policy makers at all levels, manufacturers of vehicles and appliances, builders, retailers, lenders, and households. Their decisions all need infor- mation, though the information is of different kinds. For example, house- holds need information on where the greatest potential savings lie, how much they will need to invest to meet mitigation goals, and how to assess whether the claims of those providing energy-saving equipment and services are credible and verifiable. Some of this information is available in appli- ance and vehicle certification and labeling programs and from metering and feedback systems, but it is not always available in easily usable forms, from credible sources, or targeted to the choices at hand. Some needed informa- tion is not available at all. Summary These examples illustrate the needs of many kinds of climate-sensitive de- cision makers for many different kinds of information, as well as for related basic understandings of processes that affect the results of their decisions. It is important to emphasize again that despite the language of the USGCRA, these and many other information needs are not being addressed in the cur- rent U.S. climate research effort, which focuses overwhelmingly on under- standing physical processes related to climate change and underemphasizes the various ecological, economic, and social conditions and processes that, together with climate processes, shape the consequences of human responses to climate change. (We discuss specific research and data needs below.) Models for Meeting Information Needs The eight Regional Integrated Sciences and Assessments (RISA) centers are explicitly problem focused and exemplify a promising approach to providing user-driven integrated scientific information at regional scales.
INFORMATION NEEDS FOR DECISION SUPPORT 101 RISA goals include characterizing the state of knowledge regarding climate variation and change at appropriate scales for decision making; understand- ing knowledge gaps and elucidating the linkages that characterize climateâ environmentâsociety interactions; providing a framework for responding to climate-related risks; and establishing priorities for research that can address the needs of decision makers (Pulwarty, Simpson, and Nierenberg, 2009). Regional assessments, the precursors to RISAs, began over a decade ago; the new name reflects the notion that science and assessments should be âintegrated,â both in the sense of being interdisciplinary and in terms of their fit with regionally specific knowledge requirements. Significant features of RISA projects include the use of participatory strategies in problem framing and problem solving; strong involvement on the part of stakeholders who represent a wide range of perspectives (academics, regional, state, and local agencies, extension networks, govern- mental bodies); an emphasis on assigning projects to scientists who live in the regions in which they are conducting assessments; team-building efforts designed to integrate physical and social science expertise; and the devel- opment of pilots and prototypes that serve as vehicles for collaborations among scientists and decision makers. Fundamental to RISA programs is the notion that better science does not necessarily lead to better decisions. Rather, as discussed in Chapter 2, they seek to improve decisions both through the incorporation of scientific information and by developing and sustaining knowledge-action networks. The Climate Assessment for the Southwest (CLIMAS) is a RISA that was established in 1998 and is based at the University of Arizona. CLI- MAS works with stakeholders on issues related to climate change and water availability: It does so in a context that includes significant ecologi- cal change, increasing population and urbanization, and specific economic trends. Like other RISA centers, CLIMAS develops information that is directly relevant to decision makers in the region and that spans a very broad range of sectoral and disciplinary concerns. For example, CLIMAS anthropologists have conducted research to better understand the historical, social, and economic roots of climate-change vulnerability and the specific needs of groups, such as ranchers and farmers, whose livelihoods are highly climate-sensitive. Because the health effects of climate change were deemed important by some stakeholders, CLIMAS researchers worked with the Arizona Department of Health Services, physicians, and other scientists to obtain data and create a model that enables health officials to better understand the potential for future disease outbreaks. CLIMAS personnel also worked with air quality managers on such issues as dust abatement at construction sites and ozone pollution rates and traffic congestion, as well as with water managers on reservoir-level projections.
102 informing decisions in a changing climate In each case, different kinds of information were required. Based on their experiences, CLIMAS scientists observed (Jacobs et al., 2005:9): A first step in developing useful products and services involves understand- ing the context in which they will be used. With a worldview strongly influenced by the boundaries of their own research, scientists may not recognize that the new information they produce may be only a very small consideration in a managerâs âdecision spaceâ although scientists might perceive that climate information is crucial to the management of a water system, they might fail to realize that multiple institutional, political, and legal issues dominate the decisionmaking process. Other models are being developed throughout the country in response to decision makersâ needs. Box 4-2 describes one example that addresses water management issues in Southern California. BOX 4-2 Climate Change Vulnerabilities and Water Management in California Only a few of the water agencies in California have begun to include climate- change projections in their planning activities. One exception is the Inland Empire Utilities Agency (IEUA) in Southern California, which has been working with the RAND Corporation on a study funded by the National Science Foundation to identify how climate change will affect its long-range urban water management plan. RAND customized a tool called Water Evaluation and Planning for the IEUA region, which made it possible to evaluate the performance of the agencyâs water management plan under a range of planning assumptions that took into account plausible future weather conditions; groundwater hydrology; increases in the intensity of water use resulting from future population projections; future water supply imports; future costs of different types of water supplies; and the effects of water-use efficiency programs. The National Center for Atmospheric Research developed weather data based on the best available climate projections for the region. Multiple future climate and weather scenarios were run, showing that the utilityâs current plan is appropriate if the regionâs climate does not change, or if it becomes wetter, but also that the plan could result in significant water shortages should the climate become drier. Other decision support activities centered on exploring the utilityâs options with respect to future planning, taking into account information on the costs associated with alternative scenarios and with different water-management strategies, such as replenishing groundwater, recycling water, and introducing efficiencies. The water agency is using results from this decision- support exercise in communicating with stakeholders and partner agencies. SOURCE: Information from Groves et al. (2007, 2008).
INFORMATION NEEDS FOR DECISION SUPPORT 103 THE SCIENCE BASE The examples above illustrate the types of information that are needed to support a wide range of sectoral and problem-based decisions. Managing the risks associated with a changing climate requires greatly enhanced efforts to meet these information needs. Doing so will require engagement on the part of a broader range of disciplines than is cur- rently represented in climate science activities. As the examples discussed throughout this report show, understanding climate processes is necessary but in no way sufficient to provide a sound foundation for climate-related decision making. A long series of National Research Council (NRC) reviews (e.g., 1988, 1990, 1992, 2004d, 2007b) has pointed out the ways in which increased at- tention to the human dimensions of climate and other global environmental changes can provide both new research discoveries and practical strategies for climate-related mitigation and adaptation activities. The relevant sci- ence base has been reviewed in multiple studies since the 1980s, which also offer recommendations for future development in the area (e.g., National Research Council, 1984, 1985, 1989, 1992, 1994, 1997, 1999a, 1999b, 2002a, 2002b, 2004d, 2005a, 2008c). Although not all these studies ex- plicitly address the question of climate-related decision support, they are valuable for mapping the scientific area, assessing its status and progress, and identifying key research needs. Most of these studies were recently reviewed for the purpose of identifying fundamental needs for knowledge on the human dimensions of climate change in the context of a forthcom- ing NRC study to offer strategic advice to the U.S. Climate Change Science Program (Stern and Wilbanks, 2008). This group of reports and a new study (National Research Council, 2009b) provide detailed guidance on research needs. The latest study calls for an expansion of the scientific ef- fort from a focus on the climate system and its components to encompass the end-to-end climate problemâfrom understanding causes to supporting actions needed to cope with the social impacts of climate change. It also calls for better integration of natural and social science and of basic and applied research; see Box 4-3. These priorities are fully compatible with those in this report. This section draws on those works and briefly identifies the key ele- ments of the science base for climate-related decision support and related research needs. We begin with a discussion of broader research needs re- lated to producing knowledge to inform climate-related decisionsâtermed science for decision supportâand then consider the science of decision sup- port, which treats decision support as a distinct field of inquiry. We then turn to associated needs for data, observations, and human resources.
104 informing decisions in a changing climate BOX 4-3 Priorities for Expanded Scientific Effort Restructuring Federal Climate Research to Meet the Challenges of Climate Change recommends restructuring the climate change research program to ad- dress the climate problem in an end-to-end way by better integrating natural and social science, as well as basic research and practical applications. It recom- mends six top priorities: 1. Reorganize the program around integrated scientific-societal issues to facilitate cross-cutting research focused on understanding the interactions among the climate, human, and environmental systems and on supporting societal re- sponses to climate change. 2. Establish a U.S. climate observing system, defined as including physical, biological, and social observations, to ensure that data needed to address climate change are collected or continued. 3. Develop the science base and infrastructure to support a new generation of coupled Earth system models to improve attribution and prediction of high-impact regional weather and climate, to initialize seasonal to decadal climate forecasting, and to provide predictions of impacts affecting adaptive capacities and vulner- abilities of environmental and human systems. 4. Strengthen research on adaptation, mitigation, and vulnerability. 5. Initiate a national assessment process with broad stakeholder participation to determine the risks and costs of climate change impacts on the United States and to evaluate options for responding. 6. Coordinate federal efforts to provide climate services (scientific information, tools, and forecasts) routinely to decision makers. Science for Decision Support Climate change effects are the result of the conjunction of physical and biological events and their interactions with social and economic forces. As the cited NRC assessments of U.S. climate-related research have noted, the research effort has focused overwhelmingly on improving understanding of biophysical events in the climate system and very little on understanding the human and environmental processes on which the outcomes of practical decisions about climate response depend. What is now needed is an inte- grated effort that includes fundamental and applied climate research as well as fundamental and applied research on the social, economic, ecological, and cultural conditions that determine the human consequences of climate change and of responses to it. On the basis of our review of past research assessments and of the information needs of climate-sensitive decision makers, we conclude that
INFORMATION NEEDS FOR DECISION SUPPORT 105 additional research in five focused areas is essential for providing critical missing pieces of the information these decision makers need: vulnerability, mitigation, adaptation, the interaction of mitigation and adaptation, and emerging opportunities from climate change. Fundamental research on hu- man processes and institutions that interact with the climate system (e.g., risk-related judgments and decision making, environmentally significant consumption, institutions governing resource management) should also be part of the national research effort under USGCRA, even if this research is not tied directly to decision support needs. Priority setting for this kind of research is beyond the scope of this report (but see National Research Council, 1999a, 2005a, 2009b; Stern and Wilbanks, 2009). Climate Change Vulnerabilities Many climate-sensitive decisions require an improved ability to es- timate, analyze, and project human vulnerabilities to climate change in particular regions, sectors, or communities. There is a need to build linked time-series databases covering these variables as a basis for the modeling needed to make projections on the future characteristics and geographic distribution of vulnerable populations (National Research Council, 2007d). Such efforts can benefit from research on social vulnerability to hazard events, as exemplified by the work of the Hazards and Vulnerability Re- search Institute at the University of South Carolina, which has developed indices of social vulnerability for counties nationwide (see http://webra.cas. sc.edu/hvri). Future research is needed to examine the vulnerability of people, places, and economic activities on several dimensions: by type of climate-driven event (e.g., storm surge, crop failure, heat wave, changing ecology of dis- ease); by location and scale; by relevant characteristics of affected popula- tions (e.g., socioeconomic characteristics, age, disability); and by sector (e.g., market and subsistence agriculture, water supply and quality, insured and uninsured property, large-scale public works). Research that takes demographic and economic projections into account can yield scenarios of vulnerability that can be integrated with climate scenarios to produce improved projections of the future impacts of climate change (National Research Council, 1998b, 1999b, 2009a). The Potential for Mitigation The needed research would seek improved understanding of (a) the hu- man sources of climate change (e.g., global warming potential associated with specific human actions, driving forces of activities with significant climate consequences); (b) phenomena relevant for evaluating policy op-
106 informing decisions in a changing climate tions (e.g., the potential to change actions that drive climate change with particular kinds of interventions, limits of and barriers to change); and (c) the results of mitigation policies (including costs, effectiveness, and noncli- mate consequences). Discussions of options for climate change mitigation are often rooted in policy targets and models in which the behavior of individuals, organi- zations, and economies is extrapolated from past trends or derived from simple assumptions, rather than from empirical studies of how this be- havior develops and responds over time to education efforts, policies, and regulations. Highly aggregated models of some of the drivers of climate change, such as energy and land use, have often been far off the mark in predicting future trends. Integrating into such models data on population dynamics, economic activity, energy and resource demand, and other social indicators has the potential to yield improved forecasts based on better understanding of the underlying processes (National Research Council, 1984, 1992, 1997, 1998b, 2005c). The social changes engaged by climate change will be difficult to predict with precision, however, and the iterative approach using analysis and deliberation suggested in Chapter 3 is an es- sential complement to modeling. Efforts to mitigate climate change by altering anthropogenic driving forces depend on encouraging social and behavioral change among indi- viduals, organizations, and institutions. Much of the needed change takes the form of inducing innovation and adoption of technologies for energy efficiency and low-carbon energy production and for the design of com- munities and other physical infrastructure, as well as changes in the use of existing technology and infrastructure. Change can be accomplished by various combinations of regulatory action, standard setting, provision of infrastructure, public education and social marketing, financial incentives, and voluntary action. Research is needed to identify effective initiatives and assess the efficacy of policy alternatives. Adaptation Contexts and Capacities Adaptation to climate change is the result of how regions, sectors, populations, and their governing institutions cope with their vulnerabilities (Adger et al., 2007). Adaptation strategies take place in and across multiple sectors and span a range of time periods, from long-range efforts, such as strengthening protections against riverine and coastal flooding and con- trolling development in areas prone to sea-level rise; through preparatory activity, such as planning and mobilizing resources to respond to extreme climate-related events; to planning and carrying out recovery activities following such events. Adaptation also involves the development of early warning systems for climate-related societal effects. The list of phenomena
INFORMATION NEEDS FOR DECISION SUPPORT 107 for which early warning systems are needed is a long one that includes disease outbreaks, drought, wildfire, landslides, famine and famine-induced migration, and potential conflicts over resources. Research is needed to develop indicators of adaptive capacity that can address the diversity of types of disruptive events; assess effects by region, sector, human activity, and time scale; incorporate assessments of coping capacity (e.g., emergency preparedness and response systems, insurance s Â ystems, disaster relief capabilities); and consider diverse types of impacts (e.g., on life and health, economic systems, business organizations, gov- ernments, and communities; see Yohe and Tol, 2002; Brooks and Adger, 2005). Another need is to assess various generic and event-specific adapta- tion Âoptions in terms of their ability to reduce unwanted consequences of climate change while taking advantage of the opportunities such changes present. Interactions of Mitigation and Adaptation Along with the importance of improving the scientific understanding of mitigation and adaptation as separate research priorities, there is a rapidly emerging need to improve the ability to consider mitigation and adaptation as joint contributors to an integrated approach to climate change responses (Wilbanks et al., 2003; Klein, Sathaye, and Wilbanks, 2007). Few decision-making entities choose only one set of mitigative strate- gies or adaptive measures; many communities are responding in multiple ways to climate change (see below). This type of research will be challeng- ing, both because of the higher degree of emphasis that has been placed to date on mitigation than on adaptation and because researchers tend to specialize in one area or the other. The benefits of a more balanced and inte- grated approach include a more realistic and comprehensive understanding of climate response options, their interrelationships, and their joint effects on the human consequences of climate change. Many ongoing activities provide test beds for research, sources of re- search questions, and potential audiences for and users of research results. For example, hundreds of cities and at least 30 states currently have some form of climate action plan. Most of the plans focus mainly on the mitiga- tion of greenhouse gases; only six states have adopted climate change adap- tation plans. Some states have not yet begun to consider either mitigation or adaptation, and many have plans that have not yet been implemented (Pew Center for Global Climate Change, 2008). Research is needed on these efforts, focusing not only on their effectiveness, but also on factors that influence program adoption, implementation, and content and on les- sons learned. Interactions between researchers and public officials may help focus
108 informing decisions in a changing climate research efforts on questions of pressing practical importance. Many com- munity climate response programs already involve partnerships with uni- versities, RISAs, and federal agencies, such as the Environmental Protection Agency, and some programs have explicit research components. These research efforts are just a beginning. The federal government can recognize the large and varied suite of climate response programs currently under way as a major research opportunity and develop mechanisms for funding research on and with such programs, including research with a specific decision support focus. Emerging Opportunities Research efforts are also needed for understanding and taking ad- vantage of emerging opportunities associated with climate variation and change. As societies acknowledge the end of climate stationarity, there is an understandable tendency to focus on the negative aspects of future climate variability and change. While recognizing those challenges, it is also impor- tant to highlight the many opportunities that are emerging as a consequence of changing climate. Thus, paralleling the need to understand the risks and decision support requirements associated with future climate projections is an equally compelling need for information to support climate-related decisions that can be beneficial and profitable. Climate change mitigation and adaptation strategies present numerous opportunities for entrepreneurship, creativity, investment, and economic growth. Opportunities for creative solutions to the challenges associated with climate change abound: examples include new climate-informed tech- nologies for farming, effective measures for ecosystem restoration and carbon sequestration, green technologies, and alternative energy develop- ment. New information will also be required to support engineering, design, construction, and land-use measures to adapt to both climate change itself and new climate-related hazards. At least some businesses and industries have come to recognize both the importance of adopting measures to mitigate and adapt to climate change and the business opportunities that are provided by such change. For example, in 2005, Goldman Sachs became the first investment bank to adopt a comprehensive environmental policy that included a commitment to invest in activities that preserve and enhance ecosystem services. The Constancy of Change Two other points concerning information and science for decision sup- port warrant emphasis here. First, as understanding of climate change and its consequences develops, and as decision makers identify new information
INFORMATION NEEDS FOR DECISION SUPPORT 109 requirements and pose new questions regarding mitigation and adaptation strategies, the research agenda of science for decision support will develop and evolve. This chapter offers examples of information needs and related data requirements, but we emphasize that it is impossible to anticipate in advance all the kinds of information decision makers may require or how those needs will change over time. Indeed, the specification of information needs is one key phase in the processes of engagement and collaboration discussed in Chapter 2. Providing decision support in a dynamic environ- ment of change in climate and in climate understanding requires openness to decision makersâ evolving information requirements. New and unan- ticipated problems and scientific questions can be expected to continue to emerge, sometimes with great urgency. Second and in a related vein, the need for learning and adaptation in developing well-informed climate responses, discussed in Chapter 3, gen- erates its own set of information needs. Decision support needs to draw on data on the outcomes of past and ongoing mitigation and adaptation choices, including both their intended and unintended consequences. Learn- ing lessons and diffusing innovations in response to climate change requires a body of data and evidence, so that innovations that can be shown to be effective are widely known. New information will be needed to make such determinations. The Science of Decision Support The scientific base for informing and improving human decisions, about climate change as in other arenas, lies in the decision sciences, including seminal work and new fields, such as cognitive and brain science (e.g., Kahneman et al., 1982; Raiffa, 1968; Kahneman and Tversky, 2000; Trepel, Fox, and Poldrack, 2005; Platt and Huettel, 2008). Some of this work has been linked to climate change through the decision making under uncer- tainty research centers started in 2004 by the National Science Foundation (NSF). For example, the Decision Center for a Desert City at Arizona State University focuses on research related to urban growth and water resource management in the Phoenix area, and the Center for Research on Environ- mental Decisions at Columbia University conducts research on such topics as media effects on climate change risk assessment and the role of affect and direct experience in processing uncertain climate information. The Climate Decision Making Center at Carnegie Mellon University focuses on the chal- lenges faced by insurance, forest, fisheries, ecosystem, and utility managers and on complex decision making in the Arctic region. A recent review of research needs for improved environmental decision making (National Research Council, 2005a) emphasized the need for re- search on the kinds of decision support activities and products that are most
110 informing decisions in a changing climate effective, for which purposes and with which audiences. It recommended studies focused on assessing decision quality, exploring decision makersâ evaluations of decision processes and outcomes, and improving formal tools for structuring decisions. These needs exist in the specific case of improving climate-related decision support. Research is needed on information needs, characterizing risk and uncertainty, communication processes, decision sup- port product development, and decision support âexperiments.â Information Needs Both fundamental and applied decision research assume that a key step in applying the science of decision support to all types of decisionsâ including those related to climateâis to understand diverse stakeholdersâ information needs. The kinds of engagement processes discussed in Chapter 2 provide important insights, but there is also a place for formal research to identify the kinds of information that would add greatest value for climate- related decision making and to understand information needs as seen by decision makers. Such research would improve understanding of the kinds of information that can improve climate-sensitive decisions. Climate Risk and Uncertainty Informing responses to climate change means providing information about continually changing environmental conditions and about projected risks and benefitsâand the information is always incomplete and uncertain. Peopleâs normal ways of understanding may not be adequate for processing this kind of information, with the result that they may sometimes misper- ceive useful information as too unreliable to support action, while at other times putting too much weight on a single recent event as an indicator of a major underlying change. Research on how people understand uncertain information about risks and on better ways to provide it, given knowledge about human understanding, can improve decision support processes and products. For example, although most risk analysts think of these issues in terms of presenting consequences and probabilities in terms of given values, most nonspecialists also consider qualitative aspects of risks, emotional reactions to risk information and risk-related decisions (Slovic et al., 2004; Rottenstreich and Hsee, 2001; Loewenstein et al., 2001), tradeoffs among values (Gregory, 2003; Tetlock, 2000; Keeney and Raiffa, 1993), and the context of choices (Arvai et al., 2006b). Insights from the literature on risk characterization suggest the need to develop and test novel ways of framing and presenting information for climate-sensitive decisions.
INFORMATION NEEDS FOR DECISION SUPPORT 111 Decision Support Processes Research is needed on processes for providing decision support, includ- ing the operation of networks and intermediaries between the producers and users of information for decision support. This research should include attention to the most effective channels and organizational structures to use for delivering information for decision support; the ways such infor- mation can be made to fit into individual, organizational, and institutional decision routines; the factors that determine whether potentially useful information is actually used; and ways to overcome barriers to the use of decision-relevant information (National Research Council, 1999b, 2005a, 2008d). It should also include research on processesâsuch as workshops, focus group discussions, stakeholder elicitation techniques, and decision framing exercisesâfor increasing mutual understanding between informa- tion producers and users. Another important topic is research on methods for organizing decision support processes, including the roles of boundary organizations, and on process-based approaches for improving understand- ing, such as those that combine deliberation with analysis (see Chapter 3; see also National Research Council, 1996b, 2008c). This research can help make decision support processes more effective by building on basic social science knowledge and past experience. Decision Support Products Research on the science of decision support includes studies of the design and application of decision tools, messages, and other products that convey user-relevant information in ways that enhance user understand- ing and decision quality. These products may include models and simula- tions, mapping and visualization products, websites, and applications of techniques for structuring decisions, such as cost-benefit analysis, multi- attribute decision analysis, and scenario analysis, among others (Keeney and Raiffa, 1993; Lempert et al., 2003; van Asselt, 2000; von Winterfeldt and Edwards, 1986). The products may also include tools to help decision makers manage uncertainty by identifying choices that are robust across multiple climate futures. Research can help clarify which tools, developed in which ways, can effectively link to decision makersâ preferences for dif- ferent types of decision aids. Decision Support âExperimentsâ Efforts to provide decision support for various decisions and decision makers are already under way in many cities, counties, and regions. These efforts can be treated as a massive national experiment thatâif data are
112 informing decisions in a changing climate carefully collectedâcan be analyzed to learn which strategies are attractive, which ones work, why they work, and under what conditions. The efforts already under way also present an opportunity to learn more about how information of various kinds, delivered in various formats, is used in real- world settings; how knowledge transfers across communities and sectors; and many other aspects of decision support processes. Future efforts can benefit greatly from systematic assessment of past and present ones. NEEDS FOR DATA AND OBSERVATIONS The variety of climate-sensitive decisions and decision makers generates great variety in needs for information and for data that can become the basis for information. As noted elsewhere in this report, the scale at which data are provided can present serious challenges for usefulness. Although climate change projections have typically focused on global or continental scales, the vast majority of decision contexts require information on the consequences of climate change and of responses to it at national, regional, and local, and other appropriate scales (e.g., ecosystems, watersheds). Various downscaling methodologies are currently in use, and useful climate information is now available at mesoscale levels. For example, the Lawrence Livermore National Laboratory has used statistical downscaling approaches in order to develop climate change datasets for use by research- ers and decision makers. The Northeast Climate Impact Assessment uses downscaling methods to provide information on climate change impacts that are useful for decision making in communities in the northeastern states. Analyses show, for example, how temperatures will be affected over time as a function of different approaches to emission reduction. RISAs also rely extensively on downscaled climate information, since their activities focus on specific regions affected by climate change. Scale is equally important for nonclimate data for climate-related deci- sions, and addressing inconsistencies in scale is a major challenge. Utility service and water management areas, for example, may or may not be con- sistent with city and county boundaries, which can in turn be inconsistent with scales at which climate data can be provided. Census data exist at various levels of aggregation, are standardized nationwide, and generally are of high quality, but the same cannot be said for other types of data that can be important to decision makers, such as building inventories. Input- output, computable general equilibrium, and other types of models used by economic geographers, regional scientists, and planners generate results for economic âregions,â but definitions of what constitutes a function- ing regional economy vary widely and economic regions are typically not consistent with political jurisdictional boundaries. Downscaling to county, city, or smaller levels of aggregation results in the loss of important data
INFORMATION NEEDS FOR DECISION SUPPORT 113 on economic interrelationships. Here again, data collected for particular purposes and at different levels of aggregation may or may not be consistent with usersâ decision requirementsâor with downscaled climate data. Providing data at appropriate time scales can also present a major chal- lenge. Decision makersâ time horizons differ, and the time scales for which they require information may vary from the scales for which climate and other forms of data are currently available. For those planning long-term infrastructure investments, decadal and longer term climate projections, combined with other long-term trend information, such as data on popu- lation growth, may be sufficient. In contrast, ranchers, farmers, fisheries managers, and emergency managers may require information on a seasonal scale. In sectors such as public health and disaster management services, information on climate trends, climate variation (e.g., El NiÃ±o/Southern Oscillation [ENSO] cycles), and weather may be needed and may need to be integrated with other relevant data (e.g., Glass et al., 2000; Westerling et al., 2006; Kitzburger et al., 2007). At the same time, as with reducing spatial scales, shrinking time scales introduces new uncertainties, increas- ing the need for appropriate ways of communicating uncertainty (National Research Council, 2006a). Data Availability and Data Linkage In some cases, data exist that are unavailable to the research com- munity. An obvious example is data collected from respondents under promises of confidentiality (e.g., census data). Publicly held data of this type are typically made available to researchers only in aggregated form. Although specific locational information about respondents might be useful for some research purposes, such as identifying locations of greatest human vulnerability to coastal storms, providing it would violate confidentiality guarantees. Ways to resolve conflicts between the objectives of access and confidentiality with spatially explicit data are still being developed (see, e.g., National Research Council, 2007c). Another example is data collected by businesses and held as proprietary for competitive or other reasons. Although some such of these may also have value for informing climate- related decisions, there are no established ways to identify such data or to balance public interests in its use against the private interests of the data owners. Nevertheless, a climate regime that is no longer stationary makes it more pressing to find ways to make fuller use of existing data for sup- porting climate-related decisions. In other cases, datasets on human activities are not usefully linked to environmental data, a challenge that has been noted in previous reviews (National Research Council, 1992, 2004d, 2005c). These include census data, economic input and output data, data on the natural and built envi-
114 informing decisions in a changing climate ronments, data on property values and insured and uninsured losses from past extreme weather-related events, the vulnerabilities of different popula- tions in the United States, data on health and well-being (often survey-based data), and ecoregion assessments. All these data are relevant to climate- sensitive decisions and are collected in different units of analysis. These issues have been noted before. More than 15 years ago, a Na- tional Research Council (1992:249) review recommended: The federal government should establish an ongoing program to ensure that appropriate data sets for research on the human dimensions of global change are routinely acquired, properly prepared for use, and made avail- able to scientists on simple and affordable terms. There is a national need to (i) inventory existing data sets relevant to the human dimensions of global change, (ii) critically assess the quality of the most important of these data sets, (iii) make determinations about the quality of data required for research on major themes, (iv) investigate the cost-effectiveness of various methods of improving the quality of critical data sets, and (v) make decisions regarding new data needed to underpin a successful program of research. These recommendations, written with an eye to research needs, are appropriate today for the practical purposes of building a scientific base for informed decisions about climate response. A more recent report, Deci- sion Making for the Environment, reiterated the point and called for an intensive effort on the part of natural and social scientists to develop sets of indicators capable of characterizing ânot only states of the biophysical environment but also human influences on nature and the impact of the physical world on humansâ (National Research Council, 2005a:87â88). The nation is currently far from reaching this goal. Similar needs exist in the area issues of climate and human health. A recent report in the Annual Review of Public Health notes that research on the relationship between climate change and health currently focuses mainly on impacts on infectious diseases, rather than on âindividual, fam- ily, social, and nutritional risks to the populationâ (Jackson and Shields, 2008:66). Citing a lack of collaboration between agencies concerned with climate change and those that focus on health and public health issues, the report calls for the development of monitoring systems for increased data collection on such conditions as asthma and other respiratory diseases; research on how climate variation and change affect human health; studies to determine the extent to which climate change is already having an impact on health outcomes; and research on future health risks under different climate-change scenarios. Significant data limitations also exist for climate science itself, and some existing and planned observing systems in the physical sciences have been cancelled or delayed or are deteriorating. A recent report (National
INFORMATION NEEDS FOR DECISION SUPPORT 115 Research Council, 2007b), typical of many about climate modeling, em- phasizes oceanic, terrestrial, and atmospheric observing systems and the need to ensure adequate coverage and reliability and linking these observing systems throughout the world by means of collaborative efforts, such as the Integrated Global Observation Strategy (Trenberth, Kark, and Spence, 2002; Trenberth, 2008). These needs are real. Our emphasis here is on observations relevant to linking human and environmental phenomena, an enterprise at an early stage of development. Existing Data or New Data An important data-related issue concerns deciding when new data have to be collected for a particular purpose, rather than using or custom- izing available data. Many decision support efforts involve some blend of newly collected and existing data. California provides a good example: The state used a combination of new and archived data in its efforts to chart the stateâs climate future. As part of these efforts, the state established the California Climate Change Center to facilitate research that would be responsive to decision makersâ needs. It developed a 5-year research plan and, through this process and the ensuing research, identified some needs for new data. The state funded a nonprofit organization, the California Climate Action Registry, to collect emissions data from state organizations. It established linkages with a RISA program, the California Applications Program at the Scripps Institution of Oceanography, which conducted stud- ies of climate change impacts in such areas as water resources, wildfires, and public health (Franco et al., 2008). Research on the effects of climate change in California also uses exist- ing climate models and data from existing national and regional assessment reports. For example, a report on the health, economic, and equity impacts of climate change (Redefining Progress, 2006) developed projections on fu- ture health impacts by combining data from the California Health Interview Survey (which focused on differences in insurance coverage statewide across different income and ethnic groups, a major factor in health outcomes); historic data on heat wave mortality, also by race and ethnicity, for the Los Angeles area; and data on the relationship between ozone levels and health conditions, such as asthma, along with asthma incidence rates for differ- ent social groups. These were existing data used in new ways to address concerns related to climate impacts. Even with extensive efforts to use existing data and information sources, it is important to recognize that the very nature of the phenomena for which decision support is requiredâspanning climate, ecological, and societal processes and impactsâcreates a continual stream of new information needs. Meeting those needs will necessitate new and diverse research activi-
116 informing decisions in a changing climate ties, ranging along a continuum from basic research to narrowly focused and context-specific investigations, and create needs for new data. Indicators for Climate Impacts and Responses Social indicators research involves the systematic collection, analysis, and archiving of data from the social, economic, behavioral, and policy sciences, reflecting such concepts as quality of life, human health and well- being, social inequality, and political processes (see Land, 1983; also see the World Handbook of Political and Social Indicators, published by the Inter-University Consortium on Political and Social Research; the journal Social Indicators Research; and the social indicators used by the National Association of Planning Councils for examples of measures and applica- tions). Integrated with climate data, such indicators can provide a sound basis for climate-related decision support. Some relevant work has already been done. For example, a number of projects have sought to develop indicators of sustainability that can be ap- plied at different levels of aggregation. Social indicator-based indices have also been used to measure population vulnerability to natural, technologi- cal, and socially generated hazards such as arise on humanitarian crises (National Research Council, 2007d). Vulnerability indicators include mea- sures of income, education, poverty status, and household composition. Other indicatorsâfor example, of societal, regional, and community- level capacity to respond to climate changeâare also needed, especially if they can be linked to actual climate change response efforts. For example, in the hazards area, a number of activities are currently under way to de- velop measures of resilience that can support local and regional decision making. Oak Ridge National Laboratories is currently leading the Com- munity and Regional Resilience Initiative, which, in partnership with local communities, seeks both to develop resilience indicators for extreme events and to enhance local resilience. The National Oceanic and Atmospheric Ad- ministrationâs (NOAAâs) Coastal Services Center is engaged in a multiyear effort to provide coastal communities with a suite of hazard, vulnerability, and resilience assessment tools to support community decision making to reduce the impacts of coastal hazards. Developing existing and new social indicators relevant to climate change will make it possible to conduct assessments that are consistent across communities and countries and over timeâassessments that are cru- cial for a deeper understanding of the relationships among climate change, society, and ecological systems. Such indicators will also yield important baseline and milestone measures. Indicator development can also be a vehicle for engaging decision makers. Climate-relevant social indicators are also useful for comparing the climate-related decision making across
INFORMATION NEEDS FOR DECISION SUPPORT 117 decision types, communities, and sectors and therefore for helping decision makers learn from the experiences of others. They could become central to a multidisciplinary observational system that integrates existing and new social indicators and data on the broad range of social experiments taking place throughout the nation in responding to climate change. NEED FOR A MULTIDISCIPLINARY WORKFORCE To develop the science of and for decision support, to produce useful decision support information, and to get it used, it is critical to build mul- tidisciplinary and interdisciplinary teams whose members interact and work together to better integrate data for use by decision makers. Strong teams consist of members who have high levels of expertise in their own fields, but who are also willing and able to engage with counterparts from other fields and to cross the divide between science and its uses. Many well-documented challenges exist, including overcoming the transaction and opportunity costs associated with cross-discipline collaborations, attempting to launch multidisciplinary efforts within organizations that reward discipline-based work, training the needed workforce, and enabling scientists to develop careers in interdisciplinary science. This need has long been recognized with regard to research on human-environment interactions (e.g., National Research Council, 1992:Chapter 7). In one formulation (National Research Council, 2004d:28): In both the social sciences and the natural sciences there is considerable knowledge that has the potential to make major contributions to the current and long-term goals of the CCSP [U.S. Climate Change Science Program], however that knowledge has not yet been fully applied to these goals, nor has the broad set of interfaces between these disciplines been ad- dressed. The necessary personnel to execute an enhanced level of research cannot be assumed to exist, particularly for research problems that cross disciplinary boundaries. In a number of fields, particularly in the social sci- ences, there are relatively few researchers in the position to undertake cli- mate research. Furthermore, it takes years to increase workforce capacity. The achievement of these capacity-building goals will require systematic investments over a long period of time. This overall assessment remains valid, even though some promising programs exist. One example is a new NSF-funded interdisciplinary gradu- ate education and training program called C-CHANGE (Climate Change, Humans, and Nature in the Global Environment), based at the Center for Research on Global Change at the University of Kansas (see http://web. ku.edu/~crgc/IGERT). Other universities at which RISA centers and NSF-
118 informing decisions in a changing climate funded centers on decision making under uncertainty are located also rep- resent test beds for cross-disciplinary educating and training. Development of the needed workforce will not occur without sustained efforts on the part of NSF, NOAA, the Environmental Protection Agency (EPA), and other scientific and mission agencies concerned with climate and its impacts. Science agencies in the U.S. Global Change Research Program (GCRP) (created by the USGCRA), including NSF, should expand current programs and initiate new programs aimed at supporting the develop- ment of a well-balanced, multidisciplinary climate response science/human dimensions workforce. This expansion is needed in order to address the climate-related decision support needs of the future. Efforts should target several levels simultaneously: undergraduate and graduate programs, junior and senior faculty, and scientists in the public and nonprofit sectors. A particular need is for development of the scientific workforce at the interface of the environmental and social sciences. This area lags behind current needs for many reasons, including a dearth of federal research sup- port over several decades and resistance in several social science disciplines to interdisciplinary work, applied research, and collaboration with natural science and engineering. In light of this history, developing the needed scientific workforce will require changes in academia as well as in govern- ment. A long-term commitment to supporting research for and on decision support, including funds for ongoing research centers, projects, data devel- opment efforts, and training, would provide critically important incentives for changes in academia. However, such investments may not be sufficient. We encourage the new Americaâs Climate Choices project at the National Research Council to take up this issue. Workforce enhancement initiatives also need to build capacity within sectors, organizations, and institutions that will increasingly need to use climate, climate response, and human dimensions information in their de- cision making. The need is not so much for researchers as it is for trained personnel who can help link research to its potential users. To achieve this goal, federal agencies might, for example, support training for people work- ing in or with climate-affected constituencies to increase their ability to understand and interpret climate-related scientific knowledge, information, and data. And agencies might support scientists from the climate research community to spend time in climate-affected organizations so they can bet- ter understand the organizationsâ information needs. Agencies might also support the development of career paths for climate researchers to work in applied settings in both the public and private sectors. â At the time of this writing, these agencies were the participants in the CCSP and the Cli- mate Change Technology Program, the interagency groups responsible for implementing the USGCRA.
INFORMATION NEEDS FOR DECISION SUPPORT 119 Workforce development can be achieved through a number of mecha- nisms, including enhanced funding to ongoing research and training pro- grams; the expansion of interdisciplinary graduate programs; increased student involvement in research applications and demonstration projects; support for scholarships, fellowships, and internships; and training experi- ences for decision makers and other users of climate and climate response information. Workforce enhancement efforts should also extend to the training experiences that are routinely provided for upper-level federal em- ployees in relevant agencies. Such efforts could also contain an international component linking researchers throughout the world through collaborative research and training projects and student and faculty exchange activities. In a broad sense, the federal climate research enterprise needs to pay more attention to identifying and finding effective ways to address the challenges of workforce development for decision support. Two points deserve special mention with regard to the need for a changed workforce to provide climate-related decision support. First, re- search initiatives need to reflect the fact that climate-related decisions are made in complex decision contexts in which climate information constitutes only one set of inputs into decisions. Research to inform climate change responses requires a comprehensive, multidisciplinary approach and a com- mitment to the collection of data that both support decisions and enable learning through deliberation with analysis. Agencies that are part of the GCRP and other agencies concerned with energy and the environment can contribute to the needed changes by increasingly taking a multidisciplinary and decision-oriented approach to collecting data and information in their areas of responsibility. Second, agencies need to recognize the need to support scientific re- search in fields other than climate science, whose inputs are required for fully informed climate decision making. Advances in decision support re- quire the development of data and knowledge throughout the entire range of relevant disciplines and of a multi- and interdisciplinary science work- force focused on improving the quality of environmental and climate deci- sion making. There is also a need to create and sustain settings in which authentic cross-disciplinary collaborations and stakeholder relationships can evolve over time. CONCLUSIONS AND RECOMMENDATIONS Climate-related decisions require integrated knowledge and information that includes both possible future climate conditions and socioeconomic in- formation. Together, that information provides insights into climate vulner- abilities, impacts, and the costs and benefits of alternative mitigative and
120 informing decisions in a changing climate adaptive activities. Such data are particularly vital for long-term decisions that involve substantial investments. The best way to meet the decision support objectives of the GCRP is to redefine priorities with the aim of producing information that is useful for decision-making processes. Such redirection will help fulfill the legal re- quirements of the USGCRA. As stated in that 1990 law, the purpose of the GCRP is to âassist the Nation and the world to understand, assess, predict, and respond to human-induced and natural processes of global changeâ [emphasis added]. Given that anthropogenic climate change has been es- tablished as a significant threat to human well-being, a shift of emphasis is required toward research relevant to climate change responses. The central rationale for further development of the federal research effort under the USCGRA should be to ensure that decision makers at all levels have the information they need in order to address the opportunities and challenges arising from global environmental change, including climate change. Put another way, in light of the pressing need to move from science to action, decision support is the linchpin of the program. We note that this view is consistent with that in the National Research Council (2009b) review of the Climate Change Science Program. A broad range of basic and applied climate science is still needed in the program and change in focus should not come at the expense of that need. At the same time, a key point of this report is that science is also needed to understand, assess, and predict the consequences of possible responses to climate variation and change. In addition, science is needed to improve the processes by which science supports climate-affected decisions. All these kinds of science should be use inspired (Stokes, 1997)âthat is, they should contribute to informing or improving societal response to climate change. Both basic and applied science, and both natural science and social science, can meet this test. Conclusion 6: Achieving decision support objectives requires research to understand, assess, and predict the human consequences of climate change and of possible responses to climate change. That research should be closely integrated with basic and applied research on climate processes. Recommendation 6: The federal agencies that manage research activi- ties mandated under the U.S. Global Change Research Act (USGCRA) should organize a program of research for informing climate change response, as a component of equal importance to the current national program of research on climate change processes. This program should include research for and on decision support, aimed at providing decision- relevant knowledge and information for climate responses.
INFORMATION NEEDS FOR DECISION SUPPORT 121 The research for decision support should have five substantive foci: 1. understanding climate change vulnerabilities: human development scenarios for potentially affected regions, populations, and sectors; 2. understanding the potential for mitigation, including anthropo- genic driving forces, capacities for change, possible limits of change, and consequences of mitigation options; 3. understanding adaptation contexts and capacities, including pos- sible limits of change and consequences of various adaptive responses; 4. understanding how mitigation and adaptation interact with each other and with climatic and ecological changes in determining human system risks, vulnerabilities, and response challenges associated with climate change; and 5. understanding and taking advantage of emerging opportunities associated with climate variability and change. The research on decision support should have five substantive foci: 1. understanding information needs; 2. characterizing and understanding climate risk and uncertainty; 3. understanding and improving processes related to decision sup- port, including decision support processes and networks and methods for structuring decisions; 4. developing and disseminating decision support products; and 5. assessing decision support âexperiments.â Research to understand decision support processes should include as- sessments of the transferability of knowledge gained from experience out- side the United States, where in some cases decision support efforts have a longer and better documented history than in the United States. Requests for research support from this program should be reviewed for evidence that research results are likely to be useful to decision makers. This might include evidence that the research plan was influenced by actual researcher- user interactions, evidence of knowledge of the target decision makersâ information needs and decision contexts, or a value-of-information analysis showing how particular decision makers could expect to benefit from using the information to be developed. Proposals for research on decision support might be examined for their value for improving decision support tools or systems that have already demonstrated value. Claims that a new decision support product ought to be of value would carry less weight than evidence that specific users want it or would benefit in specific ways from using it.
122 informing decisions in a changing climate Recommendation 7: The federal government should expand and main- tain national observational systems to provide information needed for climate decision support. These systems should link existing data on physical, ecological, social, economic, and health variables relevant to climate decisions to each other and develop new data and key indica- tors as needed. The effort should be informed by dialogues among potential producers and users of the indicators at different levels of analysis and action and should be coordinated with efforts in other parts of the world to provide a stronger global basis for research and decision support. The expanded observational capability, by linking social and health data with climate and ecological data, will enable better forecasting and estimation of climate-related vulnerabilities and impacts, the costs of cli- mate change to human well-being, and the effects on future vulnerabilities and costs of various mitigation and adaptation responses. This expansion should include, but not be limited to, the following elements: â¢ geocoding existing social and environmental databases and indica- tors relevant to climate impacts and responses at all governmental levels and in the private sector, with a special emphasis on longitudinal datasets; â¢ developing methods for aggregating, disaggregating, and integrat- ing such datasets with each other and with biophysical data and reconciling inconsistencies, for example, through the use of spatial social science and GIS methods; â¢ creating new datasets to fill critical gaps in existing data; â¢ greater support for research (e.g., modeling and process studies) to improve methods for producing use-relevant information; and â¢ engaging decision makers at various levels and in governmental and nongovernmental sectors in the identification of critical data needs for climate-affected regions, sectors, and populations. Datasets should be developed and maintained at levels of aggregation suitable to the needs of decision makers and matching the degree to which disaggregation can reasonably benefit them. Requests for support for de- veloping datasets, observational systems, or indicators should be reviewed for evidence that the results are likely to be useful to decision makers or to contribute to international data development efforts. In evaluating such funding requests, plans to encourage the actual use of new data systems and indicators should also be considered.
INFORMATION NEEDS FOR DECISION SUPPORT 123 Recommendation 8: The federal government should recognize the need for scientists with specialized knowledge in societal issues and the sci- ence of decision support in the field of climate change response. There should be expanded federal support to enable students and scientists to build their capacity as researchers and as advisers to decision makers who are dealing with the changing climate. Encouraging multi- and interdisciplinary research on climate change impacts, decision making, and decision support is a daunting challenge, given the generally narrow focus of many scientists. A different approach is needed to meet the challenge of providing useful and timely decision support to the wide range of people and organizations that must take ac- tion to mitigate and adapt to climate change. Training and enabling a new generation of researchers on climate-change vulnerability, resilience, and response is key to meeting the challenges of climate change.