Groundwater contamination risk assessment considering parameter uncertainty.

Hydrogeological parameters uncertainty is a major source of error in groundwater modeling.It is essential to analyze the effects of stochastic parameters uncertainty on predictinghydraulic heads and contaminant concentration. In this paper, a series of numericalexperiments were performed based on a groundwater numerical simulation to discuss theinfluence of parameters uncertainty on the prediction results. Hydrogeological parameterswith high sensitivity are identified by using local sensitivity analysis and taken as randomparameters. In order to analyze parameters uncertainty, the Markov chain Monte Carlomethod is used to select the groups of parameters which converge to the Bayesianposterior distribution. Then, the simulation results come out under different groups ofparameters which can show the distribution of hydraulic heads and contaminantconcentration, and the groundwater pollution risk at every point. It can providereference to groundwater resources evaluation and groundwater pollution control. [ABSTRACT FROM AUTHOR]