Large scale nonlinear deterministic and stochastic optimization: Formulations involving simulation of subsurface contamination (Q912762)

The paper treats the important problem of groundwater resources modelling and management. From the methodological point of view, the principle difficulty lies in a sound combination of descriptive model blocks (characterization of goundwater flow and contaminant transport) with optimization concepts (selection of best well locations and optimal fluid withdrawel/injection rates, in order to assure the required water supplies and/or to maintain water quality standards). The resulting large-scale nonlinear and stochastic models are solved approximately, applying tools of model decomposition, finite element solution of PDE systems, scenario analysis, stochastic simulation, deterministic approximations of chance constraints, nonlinear (convex) programming. The approach is illustrated by (summarized) numerical examples.