Nonparametric estimation of aggregated Sobol’ indices: Application to a depth averaged snow avalanche model

Avalanche models are increasingly employed for elaborating land-use maps and designing defense structures, but they rely on poorly known parameters. Careful uncertainty assessment is thus required difficulty arises from the nature of outputs these models, which commonly both functional scalar. Hence, so far in avalanche field, few sensitivity analyses have been performed. In this work, we propose to determine most influential inputs an model by estimating aggregated first-order Sobol’ indices. We a nonparametric estimation procedure based Nadaraya–Watson kernel smoother, allows estimate indices given random sample small moderate size. Due limited size sample, biased. Therefore, bootstrap bias correction before selecting bandwidth cross-validation. To indices, reduce dimension output using principal components analysis. After different test cases showing efficiency our approach, it applied real case. Results show that friction parameters snow depth release zone determining characteristics.