Assessing science policies for climate research: New options for organizing research in support of decision making under uncertainty

Since the 1970s, public funding of climate science has been justified in the United States for its potential to provide information to underpin important societal decisions. Both impacts from climate variability (such as an extremely cold winter in 1977 and droughts in the early 1970s) and concerns over the human-caused rise of carbon dioxide in the atmosphere were used in the late 1970s to justify a national climate research program. Both types of societal concerns remain foundational to the U.S. climate program today.

Scientific research is one tool that the federal government has adopted in order to enhance decision making capabilities. The primary logic expressed is that by creating new knowledge, society gains some extra advantage or reduction of the inherent uncertainty in each decision, and therefore may make a “better” decision than in the absence of that extra knowledge. For example, in 1977, the National Academy of Sciences released a report Energy and Climate which expressed this position with respect to one aspect of climate: “We …now summarize what will be needed to improve our understanding of the phenomena involved in the carbon dioxide problem and indicate some of the elements that should … close our gaps in knowledge, so that future decisions … can be made on as sound a basis as possible.” Such a statement could easily have been made (and has been) about developing science in order to make better decisions about food production, protecting life and property in the face of natural disasters, allocating water resources, and so on.

In order for scientific knowledge to contribute to improved decision-making capabilities in the face of an uncertain future, it must in fact be useful to decision makers. For many decades, scientific research has been pursued under the assumption that eventually, what is useful to society will be picked up and transformed into application, without much direct contact between scientists and societal decision makers themselves. Scientists and science-supporting agencies therefore largely govern themselves, making decisions about scientific priorities and direction according to internal professional norms. A collision of the need to solve specific societal issues (such as aspects of the climate problem) with this tendency of the scientific enterprise to demand insulation from societal and political pressures has resulted in a curious situation where it is possible for science to be justified as relevant to societal concerns, but remain free of the need to demonstrate such relevance. When well-meaning attempts are made to promote the use of scientific information in specific applications, such as water management, results can be disappointing, as experience shows that scientific information created without awareness of the context for decision- is often not useful, or may be misused.

Increasingly, after three decades of research and more than $25 billion in funding, climate science is now under pressure to demonstrate that its results are useful, and provide “decision support”. The latest revision of the climate program goals of the U.S. places high priority on “those elements that can best support improved public debate and decision-making in the near term,” rather than the decadal time scale originally expressed in the 1970s. The ability of climate research to effectively and efficiently support decision making depends directly on the science policies used to guide the research program itself.

Science policy decisions shape the conduct and output of climate research by guiding resource allocations, disciplinary and interdisciplinary priorities and methods, institutional design, human resources, and standards of evaluation. Science policy decisions are themselves made in the face of uncertainty about the outcomes and utility of research, and the potential needs of decision makers in the future.

After approximately 30 years of a nationally-organized federal research program in climate, it is timely to examine the assumptions of science policies used to justify and conduct climate research, and evaluate how well such research has served societal decision making. Each day, in the face of deep uncertainty, millions of decisions are made that respond to and influence the behavior of climate. How does the nation's multi-billion dollar investment in climate research affect those decisions? How can the societal value of this scientific investment be enhanced? These are the core organizing questions for our project, Science Policy Assessment and Research on Climate (SPARC). SPARC conducts research and assessments, outreach, and education aimed at helping climate science policies better support climate-related decision making in the face of fundamental and often irreducible uncertainties.

Society's strategy for responding to and preparing for climate change in the face of ongoing uncertainty hinges upon the relationship between science policy decisions and climate policy decisions. This relationship, however, has not been systematically examined. SPARC will help fill this gap. The SPARC research agenda focuses on two interconnected areas of science policy decision making where uncertainty strongly influences how knowledge is made available to society for responding to climate change, namely: 1) how climate research agendas are developed and implemented (Project on Reconciling Supply and Demand); and 2) how specific issues are prioritized given the multiple causes of global environmental change (Project on Sensitivity Analyses).

Here we summarize our research and findings under these two themes. We are learning from the Regional Integrated Sciences and Assessment and Carbon Cycle Science programs applying the method of Reconciling Supply and Demand (RSD). RSD can identify where there is a good match of information needs and supply, and where there is a "missed opportunity," or a chance to perhaps better connect the supply of scientific information to societal need. In our Sensitivity Analysis (SA) research, we examine the relative sensitivity of impacts to various causal factors, with a focus on ecosystems, in order to enhance the bases of effective decision making. We submit that these methods may prove useful in identifying science policies that can better support robust decision making under uncertainty. Previous experience suggests that such a systematic approach is necessary to realize the potential usefulness of research, avoid misuse or nonuse, and, indeed, meet the broader goals of the climate program to contribute useful information to decision makers.