AN EXPANDED SURVEY OF GROUNDWATER MODELING PRACTITIONERS ABOUT HOW THEY QUANTIFY UNCERTAINTY: WHICH TOOLS THEY USE, WHY, AND WHY NOT

Hydrogeology is among the most data-limited of the earth sciences, so that uncertainty affects every aspect of subsurface flow and transport modeling, from conceptual models to spatial discretization to parameter values and their evaluation. The literature and conference proceedings are replete with approaches, templates, paradigms and such for treatment of uncertainty. However, these tools remain poorly used, especially those of the stochastic analytic sort, leading recently to explicit inquiries about why this is the case, in response to which entire journal issues have been dedicated. We recently reported on an informal small-scale survey of hydrogeology practitioners including scientists, engineers, academics and consultants, as the "marketplace" for techniques to deal with uncertainty. Preliminary results implied significant differences in perceived importance of uncertainty and in methods used to deal with it. Here we report on results of an expanded web-based casting of essentially the same survey, targeting a much larger international population of hydrogeology practitioners. The survey reports on quantitative and other methods for uncertainty characterization and analysis, frequency and level of usage, and reasons behind the selection or avoidance of available methods. Results shed light on fruitful directions for future research in uncertainty quantification in hydrogeology.