Temporal Scale‐Dependent Sensitivity Analysis for Hydrological Model Parameters Using the Discrete Wavelet Transform and Active Subspaces

Global sensitivity analysis is an important step in the process of developing and analyzing hydrological models. Measured data different variables are used to identify number sensitive model parameters better constrain output. However, scarcity a common issue hydrology. Since hydrology we dealing with multi-scale time dependent problems, want overcome that by exploiting potential using decomposed wavelet temporal scales discharge signal for identification parameters. In proposed methodology, coupled discrete transform technique parameter dimension reduction, is, active subspace method. We apply methodology LuKARS model, lumped karst aquifer Kerschbaum spring Waidhofen/Ybbs (Austria). Our results demonstrate scale dependency processes affects structure subspaces. The reveal dimensionality increases increasing hydrologic affecting scale. As consequence, on scales. Finally, show total identified at larger than obtained complete signal. Hence, instead multiple series determine parameters, can also obtain more information about sensitivities from one single, series.