Comparison of different techniques for streamflow-related extremes estimation in ungauged catchments

Two different techniques are chosen here for estimating streamflow related variables: the top-kriging and the multivariate analysis. Top-kriging is chosen because it is directly connected to the hydrographic network structure and geometric organization, while the Multivariate analysis, based on natural logarithms of seven geomorphoclimatic characteristics, is able to take into account the catchment properties. These methods are applied over the geographical space of Tuscany Region, in Central Italy. The results are validated using a cross-validation procedure, and are compared even with classical interpolation approaches (e.g. the ordinary kriging). With the aim to define the most suitable procedure for streamflow extremes estimation, the results are compared through different error measurement methods (mean square error, mean relative error, etc.).