Data analysis and model building for understanding catchment
processes: the case study of the Thur catchment
Abstract. The development of semidistributed hydrological models that reflect the dominant processes controlling streamflow spatial variability is a challenging task. This study addresses this problem by investigating the case of the Thur catchment (Switzerland), an alpine and pre–alpine catchment that, while having a moderate (1702 km2) extension, presents a large spatial variability in terms of climate, landscape, and streamflow (measured at 10 subcatchments). The methodology for model development consists of a two–stages approach. In a first stage, we use correlation and regression analysis to identify the main influencing factors on the spatial variability of streamflow signatures. Results of this analysis show that precipitation (rainfall or snow) controls signatures of seasonality and water balance, while landscape characteristics (especially geology) control signatures of hydrograph shape (e.g. baseflow index and flashiness index). In a second stage, we use the results of the previous analysis to develop a semidistributed hydrological model that is consistent with the data. Model experiments confirm that only hydrological models that account for the heterogeneity of precipitation and geology produce hydrographs that have signatures similar to the observed ones. These models provide consistent results in space–time validation, which is promising for prediction in ungauged conditions. The presented methodology can be transferred to other case studies, since the data used in this work (meteorological variables, streamflow, morphology and geology maps) is available in many regions around the globe.