You are here:
Statistical Accounting for Uncertainty in Modeling Transport in Environmental Systems
Citation:
WEAVER, J. W., J. M. Ferguson, M. SMALL, B. Mukherjee, AND F. Tillman. Statistical Accounting for Uncertainty in Modeling Transport in Environmental Systems. Chapter 7, Yue Rong (ed.), Practical Application of Environmental Statistics and Data Analysis. ILM Publications, SAINT ALBANS, HERTFORDSHIRE, Uk, , 163-194, (2011).
Impact/Purpose:
see description
Description:
Models frequently are used to predict the future extent of ground-water contamination, given estimates of their input parameters and forcing functions. Although models have a well established scientific basis for understanding the interactions between complex phenomena and for guiding data collection, their ability to predict future environmental conditions accurately is much more limited than commonly believed. After reviewing the basis of subsurface transport modeling, and the typical sources and limitations of input data, three examples are presented. As a result, an uncertainty analysis is needed for forward-looking, “predictive”, model application. These results represent an improvement over the input parameter range analysis, because extreme events occur with low probability, which can be factored into a decision. Further, by presenting results in terms of possibilities, ranges, and probabilities, the underlying science of the models is honored which leads to two benefits identified by Oreskes (2003): better decisions for the public and enhanced credibility of scientific modeling.
