UNCERTAINTY AND POLICY-RELEVANT SCIENCE

Scientists always have served as advisors to bodies that decide on policies to achieve goals ranging across a spectrum of human activity. For a great many practical problems, scientific and technological input is fairly straightforward. For example, delivery of potable water to a modest-sized community requires identification of a source of water, analysis of the quality of the resource, and the design of the water delivery, storage, and treatment structures. These types of activities fit neatly into the model whereby problems can be decomposed into two distinct parts: (1) the scientific and technical portion (dealing with "facts"); and (2) the political portion (where a decision on what to do and how to do it is made). This model assumes that a problem (e.g., development of a water supply) is set by a political unit, scientific and technical questions are answered by scientists and engineers, and a decision on implementing a policy is made by the political unit in light of the technical answers. The simple model wherein problems are neatly compartmentalized into a "science" part and a "policy" part may work for local problems (like water supply), but it does not describe science-policy interactions for today's more complex problems. Many problems facing us today have roots that are not “local.” For example, consider: long-range transport of atmospheric pollutants; large-scale changes in climate; and large-scale modifications to regional flows (e.g., Florida Everglades, North China Plain). The goals of policies related to such complex problems are not simple, the science related to them involves many disciplines, scientific investigations needed to inform decisions are not independent of the social context, and significant uncertainty surrounds the calculations necessary to illuminate the potential effects of different policy scenarios. Challenges for the geosciences community are to determine how to define science needs related to policy and to develop and refine the structure for dealing with uncertainties in scenario projections.