Coping with global change, providing clean water for growing populations, and disposing of nuclear waste are some of the most difficult public policy challenges of our time. Unknowns in the physical sciences are one source of the difficulty. Addressing each challenge depends on researchers applying still-emerging principles from hydrology and other geophysical and biological sciences to multiple and coupled complex systems. Each challenge requires analyses at multiple spatial scales, and each is marked by significant scientific and technical uncertainties, magnified by long timescales. Observational data are often sparse or uneven in coverage and quality.
 In fact, real difficulties in meeting these challenges also arise in the behavioral sciences. What distinguishes these challenges from, say, putting a man on the moon, is the complex sociology and psychology of decision-making bodies, stakeholder groups, and the interested public, who receive and process the scientific information intended to inform their choices. These entities often have divergent interests in the status quo and conflicting views about uncertainty and ambiguity of scientific information. Thus the framing and presentation of information can strongly influence the policy debate and ultimately policy choices.
 Over the last two decades, legions of hydrologists and other geophysical and biological scientists have made monumental efforts to understand the physical phenomena and processes underlying these environmental challenges, and then inform the relevant policy-making bodies. The three integrated assessments of the Intergovernmental Panel on Climate Change  and the Department of Energy's more than $4 billion spent on Yucca Mountain research [U.S. Department of Energy, 2002] are but two examples of these kinds of efforts. This type of interaction has tended to be one way and premised on a “predict-then-act” model of decision making [Lempert et al., 2004; Sarewitz et al., 2000].
 While this work has had an impact, it is reasonable to ask whether scientists could be doing a greater service. Failed or marginally effective national and international policies thus far on climate change and nuclear waste, and in many regions of the world, water management, certainly suggest that new kinds of interactions with decision makers may be worth trying.
 A potentially rich vein of transdisciplinary research is to integrate the psychology of decision making, known as “judgment and decision making,” or JDM, with the development of technical information and decision support tools for complex, long-term environmental problems. Practitioners of JDM, including psychologists, economists, organizational researchers, and other types of decision analysts, conduct research on how individuals and groups respond to uncertainty and ambiguity, hedge against risks, anchor decisions to the status quo, compare relative risks and rewards of alternative strategies, and cope with other classes of decisions. JDM is empirically based, and relies largely on controlled experiments with different types of decision makers and decision situations. Practitioners use a variety of stimuli, chance devices, hypothetical and real choices involving small stakes, scenarios, and questionnaires, to measure (directly and indirectly) preferences under varying conditions.
 These kinds of experiments can help guide choices about the level of complexity required for different types of decision-making processes, the value of new data collection efforts, and the ways in which uncertainty in model outcomes can be cast to minimize decision-making paralysis. Most important, results from these experiments provide a scientific basis for interacting with decision makers throughout the model development process. This interaction includes designing better ways of eliciting and combining opinions and then communicating information relevant to public policy issues, with the goal being to improve the value of the scientific contribution to the social decision.
 For example, under a National Science Foundation grant, colleagues and I are applying JDM research to modeling tools for robust decision making on climate change under deep uncertainty, using the specific cases of long-term water resources planning in California and observational network design to detect abrupt climate change [Lempert et al., 2004; Keller et al., 2004] (the project's Web site may be viewed at http://www.rand.org/ise/projects/improvingdecisions/). In each case, insights from controlled JDM experiments will help us frame elicitation sessions with real decision makers to understand how they perceive uncertainty [Karelitz and Budescu, 2004], how they may pursue hedging strategies to deal with uncertainties, how they may value new information, and how much they may be willing to pay to improve confidence in their decisions. These findings will feed back to design of decision support tools, metrics, and graphics that are more responsive to decision makers' information needs.
 The RAND climate change-related collaboration between JDM and hydrologists could serve as a model for other environmental policy issues. The hope is that the fruits of this kind of synthesis will improve scientists' success rate in influencing societal choices about the major environmental challenges of our time.