A quantitative model-independent method for global sensitivity analysis ofmodel output

A new method for sensitivity analysis (SA) of model output is introduced. I t is based on the Fourier amplitude sensitivity test (FAST) and allows the computation of the total contribution of each input factor to the output's variance. The term "total" here means that the factor's main effect, as wel l as all the interaction terms involving that factor, are included. Althoug h computationally different, the very same measure of sensitivity is offere d by the indices of Sobol'. The main advantages of the extended FAST are it s robustness, especially at low sample size, and its computational efficien cy. The computational aspects of the extended FAST are described. These inc lude (1) the definition of new sets of parametric equations for the search- curve exploring the input space, (2) the selection of frequencies for the p arametric equations, and (3) the procedure adopted to estimate the total co ntributions. We also address the limitations of other global SA methods and suggest that the total-effect indices are ideally suited to perform a glob al, quantitative, model-independent sensitivity analysis.