Predictive Models are the foundation that environmental management decisions are made. Models provide predictions of where and how quickly contaminates can be expected to spread through the subsurface and ground water, providing critical risk assessment information. Management decisions are and will be based on expected contaminant behavior in the subsurface, and on the potential of each contaminant to pose adverse health risks. However, the data available for predictive models is problematic.
A major challenge facing subsurface research is the development of accurate predictive models describing non-ideal transport of contaminants undergoing coupled processes. Ideal transport, in this sense, refers to transport deviating from behavior predicted under a given set of assumptions. The most commonly applied paradigm underlying environmental management decisions are based on assumptions of homogeneous porous media, and linear-instantaneous phase transfer transformation reactions. Predictions are difficult to make due to uncertainty and a number of unexpected factors such as: early arrival of contaminants, changes in behavioral patterns, and difficulty in removing the contaminants from the subsurface.
In truth, the only valid evaluation of a model’s performance is to attempt to predict measured data with data sets that are independent of those used to calibrate the model.
- Contacts:
- Mark Ankeny, (208) 526-5748, Send E-mail