QUANTIFYING EXPERT UNCERTAINTY: A CONFLICT IN THE MAKING FOR HYDROGEOLOGY

The burden of proof in applied and academic branches of hydrogeology has different standards in the practice of hydrogeology. A high burden of proof in academic hydrogeology comes from rigorous publication standards and the grant review processes. In contrast, a much lower burden of proof is required for applied hydrogeology which often provides the means to resolve legal disputes and arrive at regulatory political consensus on environmental issues. Two examples from my own files illustrate my concern. First, scientifically robust geochemical data from a study of wetland-groundwater interaction were recently rejected by the funding regulatory agency because of minor compromises in sampling protocols and uncertainty in analytical precision. Second, in a landfill siting, consultants spent millions of dollars in field studies to determine the plausibility that a confined sand and gravel aquifer could be contaminated, but inexpensive isotopic analyses, which would have resolved the issue, were not done because regulatory protocols did not demand them.

The differences between how hydrogeology is practiced in academia and the applied world can be most striking where the burden of proof consists of only “more likely than not.” Recently, the 2007 Intergovernmental Panel on Climate Change (IPCC) report “quantified” scientific expert opinion within a numerical scale. This quantification may have unanticipated consequences by forcing an overlap between the academic and applied hydrogeology communities in ways that can be difficult to resolve. Forcing a hydrogeologist to quantify his or her professional judgment is inappropriate. When quantification cannot be scientifically made, the integrity of the scientific method and proper decision making are both compromised.