Strategic Plan for the
Climate Change
Science Program
Review draft, November 2002

PDF Version of Full Report

PDF Version of this Section

How to Submit Comments

Acronyms

Authors & Contributors

Chapter 4:
Decision Support Resources

The Climate Change Research Initiative (CCRI) will synthesize the results of the research conducted by the Climate Change Science Program (CCSP) to present critical information to decisionmakers and resource managers both within and outside of the US Government. Decisionmakers, as defined here, engage in the development of national policy such as setting national goals for greenhouse gas emissions and negotiating with other countries over international agreements. Along with resource managers in different regions and sectors, decisionmakers also are engaged in policy, planning, and operational decisionmaking issues related to the management and allocation of natural resources and the associated physical infrastructure. The science and decision support activities sponsored by the CCSP are designed to provide critical information about a number of the decisions and natural resource issues affected by climate variability and change. One major key element of the CCRI is the ongoing engagement of scientists, decisionmakers, resource managers, and other stakeholders in identifying issues and questions, and providing data and products that include characterizations of uncertainties and the level of confidence associated with this information.

One of the principal motivations behind the CCRI is enhancing the CCSP commitment to synthesizing scientific results and producing decision support resources responsive to national and regional needs. Decision support resources include a wide variety of mechanisms for creating and supporting a dialogue between scientists and decisionmakers to identify issues and questions of concern, and for framing the research agenda needed to answer the questions. They also include a variety of analytical techniques, including historical data analysis, scenarios, and applied climate modeling, that serve decisionmakers, and product development that arises from the strong interaction between the science and decisionmaking needs.

One component of the CCRI will focus on national-level challenges associated closely with the mitigation issues (improving understanding of the costs and benefits of particular strategies for reducing emissions) associated with long-term global climate change. In a parallel effort, the CCRI will accelerate the development of a structure and process for integrating science with decision processes to assist the development of regional and sectoral adaptation responses (actions to reduce vulnerability, seize opportunities, and enhance resilience) to variability and long-term changes in climate. These two efforts complement and reinforce each other with lessons learned about how the process of synthesizing and analyzing scientific information can inform policy and operational decisions. Although the actual process of making policy and resource management decisions should remain entirely separate from the research function, the establishment of a new class of working relationships will ensure that the sponsored research is well informed by an understanding of what information is timely and useful for decisionmakers, resource managers, and other stakeholders. Research will provide a continually stronger foundation to help decisionmakers evaluate the suite of alternative policy options and operational strategies.

This section of the Strategic Plan describes activities intended to initiate innovation in decision support resources that are particularly relevant to the driving forces and effects of climate change at a national and regional level, recognizing the need for continued progress in basic climate science questions. Because climate is not the only variable component in the decisionmaking process, and societal challenges rarely reveal themselves as neat, single-issue topics, this initial focus is nested within a commitment to integrate across temporal scales, spatial scales, and multiple effects (both positive and negative).

The following sections lay the groundwork for building decision support into the CCSP: the incorporation of science-based decision support research including scenario development; applied climate modeling; and the development and application of improved methods for dealing with scientific uncertainty in the decision process.

For the last decade, the primary focus of the development of climate change science information at the national level has been in response to the debate on energy policy. At issue was whether human-induced climate change could be so significant as to require immediate and steep reductions in fossil fuel emissions. The main constraint on any such reductions has been the desire to maintain modern living standards by preserving the ability to serve the energy needs of a growing economy with diverse economic sectors in the context of evolving societal values. Issues central to the debate have included distinguishing between natural climate variability and human-induced climate change; the adequacy of observations to determine climate variability and change; the reliability of climate modeling; and the prediction of the immediate costs and possible benefits of mitigation options.

The CCRI will initiate a process to identify policy decisions that should influence the focus of climate change research programs. It will be important to consider likely future policy decisions, because there can be lag time in the delivery of research results. This process will include meetings with current and past decisionmakers. The resulting articulation of potential policy questions will serve as a foundation for the subsequent decision support activities. One goal is to expand the range of decisions from an emphasis primarily on energy policy to a broader agenda that includes greenhouse gases and pollution other than carbon dioxide (CO₂), emissions that result from land use (particularly deforestation and the cultivation of certain crops), and the management of other resources and decisions at a regional level. Examples of other broad policy arenas that require science-based climate information are agriculture, water resources, air quality, forestry, wildfire management, public health, and foreign aid.

The importance of climate change and variability lies in its impacts on natural resources, the economy, human health, and ecosystem sustainability. Some regions, sectors, and assets will be more vulnerable and some more resilient to climate variability and change, and taking steps to seize opportunities or identify particularly vulnerable assets and enhance their resilience will help ensure economic productivity and the well being of citizens and the environment. Decisionmakers who operate in the resource management arena are confronted with an array of influences that impact their decisions, and these must be considered in work done under the CCRI. Climate variability and change, demographic change, land use, laws, and public values are only a few of the inputs into their decision processes. In addition, they are required to make decisions on a range of time scales from a day-to-day operational perspective to a longer-term planning perspective.

The climate science issues that have emerged over the last decade that have been raised by these decisionmakers include concerns about contradictions in, and the coarse spatial scale of, information on climate change from global climate models, and the lack of availability of useful and effective climate observations and products for use in their decision processes. Regional- and local-scale analyses of potential climate impacts are limited by the fact that currently available model projections are not reliable at the smaller scales that are required for these analyses. However, regional- and sectoral- scale climate diagnostics and analyses, in cases where they prove to be accurate, can be and have been used effectively in regional decisionmaking contexts, creating an important demand for the provision of useful observational products and data.

One goal of the decision-support efforts of the CCRI is to identify national-level decisions and to use that list to develop decision support activities as well as to help prioritize climate change research. A second goal is to articulate and expand upon our understanding of the role of climate in human affairs such that science-based information can be synthesized, analyzed, and incorporated meaningfully into policy analysis and operational resource management.

Research projects that contribute to decision support will be supported under CCSP. These research projects benefit from the results of the US Global Change Research Program (USGCRP) research efforts discussed in Chapters 5-11. Links will also be made to the reporting and outreach activities (Chapter 13) and to international research cooperation (Chapter 14). The CCRI will provide a means for synthesizing, analyzing, and evaluating scientific results that will provide supporting information for policymaking and operational resource management processes.

The type of issues requiring decisions at the national level for which information about long-term global climate change is relevant has evolved considerably in recent years. The CCRI will attempt to establish mechanisms to foster a new class of working relationships to ensure that relevant issues are identified, articulated, and communicated to the research community. This task is understood to be a particularly challenging one, where decisions for which science-based information will be useful will be a subset of a broader range of decisions. Accomplishing a productive and effective relationship among researchers, federal research managers, and policy specialists will require new working arrangements. The CCRI will devote attention to the type of institutional changes necessary to forge effective interaction between research processes and policy development.

For policy development related to mitigation, it will be difficult to generate a true representation of salient decisions. Over the last several years there has been an interest in issues as diverse as estimating the costs and impacts of concentration paths over time; costs and benefits of various stabilized atmospheric concentrations; priorities for technology R&D; evaluating regulatory instruments; analyzing uncertainties; analyzing the role of the United States with respect to the rest of the world; analyzing which gases to control and how to trade off certain greenhouse gases versus others; the connection of greenhouse gas emissions to other pollutants, such as aerosols; assessing impacts from possible climate change at a local level; high-consequence but low-probability events; and others.

Stakeholder interaction will be essential to the task of identifying decision issues at the national level, but managing this interaction will be a different type of experience than it has been at the regional level, where researchers have spent the last several years learning how to interact with resource managers and local planners. Certain sectors, such as energy, technology development, or international disaster management, are obvious candidates for exploring how to build improved stakeholder relationships. Many of the decision alternatives in these particular areas will be amenable to the "If..., then..." paradigm that uses the scenarios described in the next section.

The general approach for accelerating and enhancing decision support for regional resource management will be based on the following framework:

Resource managers are challenged every day by the need to make decisions despite the existence of scientific uncertainties and the inability of scientists to begin to make absolute predictions about future outcomes. Through the USGCRP, a sustained relationship between investigators and decisionmakers has been nurtured to create the basis for developing a shared understanding of the general potential for and nature of risk and benefit, and extracting from scientific findings the information to begin to support decisionmaking within a context of managing risk. Through regional and sector-specific research, investigators will continue to work closely with decisionmakers and resource managers to identify the level of certainty required for different decision contexts, and mechanisms for best communicating the uncertainties, which may include acknowledging that it may not be possible to provide meaningful information at the required level of certainty.

A major value of the regional resource management component is in deriving insights from "lessons learned" about how science can be integrated effectively into the operational decisionmaking process and, to the extent possible, into policy analysis and development. This activity involves the analysis of information from multiple disciplines -- including the social and economic areas -- to address the specific questions being asked by resource managers and other stakeholders. It also includes an analysis of adaptation options to improve society's ability to respond effectively to risks and opportunities as they emerge. Based on the regional and sector-specific research that has been conducted over the last decade, preliminary target areas for accelerated research that will be considered include air quality; water availability and quality; forest and wildfire management; drought; and public health.

"Decision support" refers to the provision of timely and useful information that addresses specific questions being asked by a decisionmaker. It could be a question that is pertinent to any of a full range of issues related to climate change, including adaptation, the management of resources in the face of scientific uncertainty, mitigation, or technology development. For example, a national-scale question addressing emissions might be framed as, "What are the economic consequences -- costs and benefits -- associated with the adoption of an emissions goal framed in terms of percentage reductions against a specified base year emissions level?" Alternatively, it might be framed on the regional or local scale to address adaptation questions, such as: "How could water resources be managed if winter snow melt shifts to an earlier time of year?"

Techniques that serve to articulate research findings in ways that resonate with decisionmakers and that incorporate parameters important from their perspective are a key part of the CCRI commitment to build and sustain productive, appropriate interaction between research and action. A variety of resources and approaches are being used to explore the possible range of consequences of climate change, including historical records; integrated assessment models; synthesis, analysis, and presentation of scientific conclusions for incorporation into existing decisionmaking frameworks; communication and outreach processes to policymakers; and sensitivity and "If..., then..." analyses. Although all of these contain sometimes profound uncertainties, their use can provide existing information for decisionmakers, resource managers, and other stakeholders.

A variety of methods are available for illustrating and analyzing how fluctuations in climate influence social, economic, and ecological systems, including:

For many decision alternatives, an "If..., then..." analysis enabled by scenarios can be performed that provides information to a decisionmaker. Assuming a particular action is taken, the analysis predicts the consequences of that action. Scenarios play a key role in the decisionmaking process by providing the opportunity to explore options against a variety of alternative possible backgrounds. The term "scenario," as used here, refers to any description of the world as it might evolve or be made to evolve in response to decisions. The goal of the CCRI scenarios activity is to develop, maintain, and enhance the capability to answer "If..., then..." questions relevant to the full range of climate change decisionmaking, from the management of resources to the formation of national and international policy. The activity will seek to ensure that a balanced approach is taken that maintains objectivity and avoids focusing on "worst-case analysis" alone.

Scenarios provide a vehicle for posing and analyzing questions, for example, "What if the United States adopts an emissions goal?" The question as framed above, however, is insufficiently specified. It lacks detail. For example, no mechanism by which the goal might be attained is specified. Further, there is no description of areas of concern, such as the effectiveness of the limitations in environmental terms; the impact on jobs, Gross Domestic Product, the economic health of important economic sectors and regions of the country, and international trade; the implications for energy and national security; and the effects on ecosystem goods and services. Decisionmakers and stakeholders, through interactions with researchers, can provide the necessary level of specificity and may together create a better list than either could separately generate. All scenarios start with information originating outside the system in question, contain some description of the system of interest, and provide a mechanism for evaluating a variety of approaches that may be employed.

Scenario development techniques abound, and range from qualitative approaches to formal computer models. Models link statements about key external factors, such as population growth and migration; the abundance and availability of resources; market structure; energy cost and use; international trade; and technology deployment, through algorithms that attempt to capture their relationships. Some scenario development techniques may combine both qualitative and modeling approaches, similar to gaming exercises that provide computer models for role-playing. The Intergovernmental Panel on Climate Change (IPCC) has made extensive use of scenarios to drive climate models, although the model outputs have seen limited use in studying the impacts of climate change. Other qualitative and quantitative scenarios have been used extensively in controversial assessments of the potential consequences of climate change for particular sectors and regions in the United States. The development of scenarios also makes possible potentially fruitful communications with other important policy realms such as the National Climate Change Technology Initiative (NCCTI).

Research is essential to every part of the scenario process. Scenarios will require the acquisition and synthesis of knowledge about factors that lie both within and outside of the processes in question, including economic growth; energy supply and demand; land use; agricultural practices; ecosystem characterization; and the characterization of the cryosphere, hydrosphere, ocean, and atmosphere. Models of such processes can be extremely detailed, with some requiring extensive time (weeks) on the fastest available computers. It is important to realize that the nature of the question being asked by the decisionmaker, as well as the level of scientific certainty required, influence the construction of the scenario and the type of modeling undertaken.

CCRI scenario development will go beyond past scenario activities such as those of the IPCC. Decisionmakers, resource managers, and other stakeholders will be engaged to help identify the types of scenarios that could be used to provide them with timely and useful information. The CCRI will develop logical and internally consistent scenarios with input from the full range of relevant stakeholders, which potentially include environmental non-governmental organizations (NGOs), industry representatives, natural resource managers, government agencies, and research scientists. It will undertake independent analysis to extract up-to-date information on projections for key variables (e.g., demography; technology characteristics and costs; and economic growth and characteristics) and the relationship of key driving forces to environmental change (e.g., land use and land cover) and adaptive capacity. The CCRI will coordinate its scenario development plans with the new IPCC scenario efforts. The IPCC may be interested in adopting some of the CCRI scenarios or combining CCRI and IPCC efforts.

Climate models have been a central part of the US climate program since the 1970's. Models are an essential tool for synthesizing observations, theory, and experimental results to investigate how the Earth system works and how it is affected by human activities. Such models can be used in both a retrospective sense, to test the accuracy of modeled changes in Earth system forcing and response by comparing model results with observations of past change, and in a prognostic sense, for calculating the response of the Earth system to projected future forcing. For the CCSP, we need to consider a subset of the broad domain of climate modeling, in particular those specific tasks that can provide near-term information products to inform management and policy decisions involving climate. This is the area of Applied Climate Modeling. It provides the means for translating the scenarios described in the preceding section into the decision support resources.

There are a number of obstacles facing the application of the best of US capability in climate science to these critical applied modeling issues. The NRC (2001b) found that when comparing US and European high-end modeling, the United States is still lagging in its ability to rapidly produce accurate high-resolution model runs. In addition, there is a need to increase confidence in model results and expand their immediate utility for decision support. These considerations prompt several priority directions for Applied Climate Modeling.

Climate sensitivity is a measure of the climate's response to a unit change in radiative forcing due, for example, to changing atmospheric concentrations of greenhouse gases. It accounts for a major part of the uncertainties in climate projections. The current crop of world-class climate models exhibits an unacceptably large range in climate sensitivity. The major US models that have been used for IPCC scenario assessments -- the Community Climate System Model (CCSM), operated at the National Center for Atmospheric Research, and the model developed at the Geophysical Fluid Dynamics Laboratory (GFDL) -- lie close to the opposite ends of this range, making them ideal resources for investigating the processes and assumptions responsible for uncertainty in sensitivity.

All current climate models fail to adequately simulate several climate system processes and their feedbacks. One example of such a process is ocean mixing, which to a large degree controls the rate of projected global warming. Atmospheric convection, hydrologic processes, and representation of clouds, all of which strongly influence the magnitude and geographical distributions of global warming, are also poorly simulated. These deficiencies are thought to be related to the large range of climate sensitivity and contribute significantly to model uncertainties. High-priority research will focus on representations in models of the relevant physical feedback processes, using available observational data and, where required, new field observations. This work will enable focused model comparisons to understand the reasons for differences in climate sensitivities. Products will include new knowledge about important climate feedback processes and their improved representation in climate models, potentially leading to a significant reduction in known uncertainties in climate projections. Particular attention will be devoted to cloud/water vapor processes, as described in Chapter 2 (see also Chapter 6).

It will be important to identify the one or two largest sources of uncertainty in feedback processes currently represented in climate models, determine the causes of the uncertainty, and improve the physical representation of those processes in the models. Comparing model simulations and observations indicates that the major problems are generic, affecting all climate models. Climate Process Teams (CPT), a new approach to focused research designed to more rapidly reduce known uncertainties in climate model projections, will conduct the research. The teams of climate process researchers, observing system specialists, and modelers will work in partnership with multiple modeling centers (see also Chapter 6).

In moving towards the development of a more operational applied climate modeling capability, it is necessary that models be put through a more rigorous program of testing than has been the case to date. For weather prediction, such testing is straightforward: information on the accuracy of the forecast is immediately available, and statistics can be generated. For applied climate modeling, such immediate feedback is impossible. It is necessary, as climate modeling moves beyond the research domain, that models be formally tested against specific observational data sets. This needs to be done with sufficient care and fidelity to detect small differences in future climate trajectories. The observations must have tight tolerances for accuracy, sampling protocols, data availability, and cost, and must meet the criteria for long-term stable climate records, as described in Chapter 3. Lastly, there must be a formally reviewed assessment of models' performance, both against these specialized data sets and against each other. The testing program would have four particular components:

As a near-term product, a critical comparison of the model sensitivity of major US models will be undertaken by the major modeling centers (1-1.5 years), followed by publication of a reviewed interim report (3 years). Considerable progress has been made already, as the modeling and diagnostics communities are developing scientific and protocol plans for examining differences between models, as well as differences between models and observations.

One of the highest priority applications of climate modeling is the development of new, state of the art projections of the impact on global climate resulting from different scenarios of greenhouse gas emissions. As described in the previous section, well-developed scenarios are essential vehicles for asking the central "If..., then..." questions. These scenarios must consider potential economic changes, possible changes in energy sources, and suites of potential new technologies, along with possible environmental changes which may themselves act as agents of climate change. Analysis of uncertainties will be included as part of the scenario exercise.

Several recent NRC reports have documented the need to strengthen US modeling capability. In response, a number of steps will be taken to enhance the US climate modeling capability:

Decisionmaking associated with climate change and variability can be viewed as a subset of a larger class of problems that involve decisionmaking under uncertainty. Decisions are made and public policy is developed in many areas other than climate change that involve uncertainties, such as terrorism and genetic engineering. Although each of these issues is associated with its own unique set of factors, they all involve the need to understand longer-term risks for systems where there are many variables, each of which interacts with the others in complex, often nonlinear ways. Fruitful lines of inquiry include many different approaches, such as game theory, preference elicitation, and decision sequencing.

Advancement of theory, approaches, and resources to improve decisionmaking associated with climate change and variability will take a variety of forms. New paradigms will be needed to better integrate the variable spatial, temporal, and organizational scales at which interconnected natural and human systems function. New approaches are needed to conceptualize problems and to obtain and analyze relevant data from a diverse set of sources. New resources need to be created that combine improved operational capabilities with more effective user interfaces, thereby making them more readily useful to decisionmakers and other stakeholders. These resources will require integration of the latest advances in information systems technology with statistical advances, such as visualization and stochastic modeling. Also needed are the development and deployment of more effective forms of communication to facilitate broader dissemination and implementation of scientific insights and information to a broad range of end users.

An accelerated fundamental research program will be put in place to develop applications of existing capabilities to the issues of uncertainty in the climate change decisionmaking context as well as to the robust analysis of risk and vulnerability of natural resource systems. Additional research programs will focus on the development of new resources for addressing scientific uncertainty in decisionmaking.

NRC, 2001b. Panel on Improving the Effectiveness of U.S. Climate Modeling, Board on Atmospheric Sciences and Climate, National Research Council, Improving the Effectiveness of U.S. Climate Modeling (Washington, DC: National Academy Press).