Optimal stochastic algorithms for multidimensional sensitivity analysis of large ecological model

Abstract

Sensitivity analysis is a procedure for studying how sensitive are the output results of large-scale mathematical models to some uncertainties of the input data. The computational efficiency (in terms of relative error and computational time) of several stochastic algorithms for multidimensional numerical integration has been studied to analyze the sensitivity of Unified Danish Eulerian model output to variation of input emissions of the anthropogenic pollutants and of rates of several chemical reactions. The algorithms will be applied to compute global Sobol sensitivity measures corresponding to the influence of several input parameters on the concentrations of important air pollutants. The study will be done for the areas of several European cities with different geographical locations. We show that the optimal algorithms under consideration are very efficient for the multidimensional integrals under consideration and especially for computing small by value sensitivity indices.