Sensitivity Analysis for Environmental Models and Monitoring Networks

Statistical sensitivity analysis is shown to be a useful technique for assessing both multivariate environmental computer models and environmental statistical spatio-temporal models in the perspective of risk assessment. Methods are reviewed and extended to cover with two applications which are reported as case studies. The first, related to waste water biofilters for heavy metals, is aimed at assessing the input influence on both environmental and economical variables. The second, related to spatio-temporal models for air quality monitoring networks, is intended to study the influence of each station to the model performance.