Stochastic modeling of exposure and risk in a contaminated heterogeneous aquifer. 1 : Monte Carlo uncertainty analysis.
This paper describes a stochastic methodology for modeling contaminant transport to determine the risk of potential human exposure to toxic chemicals in a heterogenous contaminated aquifer.
The Sequential Gaussian Simulation (SGS) method is used to generate random hydraulic conductivity fields.
Statistical sampling of hydraulic conductivity fields is based on Monte Carlo simple random sampling.
An optimum number of Monte Carlo runs is calculated for a single uncertain groundwater parameter, hydraulic conductivity, using several approaches.
The magnitude of human exposure via the ingestion of contaminated groundwater from a well is calculated considering the probabilistic distribution of the hydraulic conductivity results derived from a numerically modeled contaminant concentration profile.
This research shows that in exposure assessment, choosing the appropriate number of Monte Carlo simulations can be very critical.
The number of simulations should be well justified and should guarantee convergence toward a stable statistical distribution of the output.
Mots-clés Pascal : Pollution eau, Eau souterraine, Nappe eau, Devenir polluant, Contamination, Modélisation, Simulation numérique, Analyse risque, Modèle stochastique, Incertitude, Méthode Monte Carlo
Mots-clés Pascal anglais : Water pollution, Ground water, Aquifers, Pollutant behavior, Contamination, Modeling, Numerical simulation, Risk analysis, Stochastic model, Uncertainty, Monte Carlo method
Notice produite par :
Inist-CNRS - Institut de l'Information Scientifique et Technique
Cote : 99-0474492
Code Inist : 001D16A04B. Création : 22/03/2000.