Project Manager, Weavertown Environmental Group, Canonsburg, PA
Pages: 18 Published: Jan 1996
It is impossible to know with certainty that application of a ground-water modeling code to a site-specific problem is producing correct predictions. This paper describes procedures in ASTM standards that can reduce uncertainty associated with a model's predictive capability for use in decision-making. Modeling projects should begin with a clear statement of the objective. The modeler should consider restating the problem in terms of bounded predictions rather than specific numerical predictions to allow conservatism in assumptions. The modeler should develop a conceptual model of applicable hydrogeologic framework and physical processes. The uniqueness problem should be addressed by using measured hydraulic conductivities, calibrating to ground-water flow rates as well as hydraulic heads, and/or calibrating to data collected during multiple distinct hydrologic conditions. Over-calibration can be avoided by performing application verification. A sensitivity analysis should be performed to identify model inputs that meet the calibration criteria but produce fundamentally different predictions. A field audit can be performed to directly test the model's prediction. Sound hydrologic judgement is necessary to perform these procedures and interpret modeling results.
ground-water modeling, calibration, sensitivity, verification, uniqueness
Paper ID: STP38377S