Regionalization for Ungauged Catchments — Lessons Learned From a Comparative Large?Sample Study

Model parameter values for ungauged catchments can be regionalized from hydrologically similar gauged catchments. Achieving reliable and robust predictions in by regionalization, however, is still a major challenge. Here, we conduct comparative assessment of 19 regionalization approaches based on previously published literature to contribute new insights into their performance different geographic regions. The use geographical information, physical catchment attributes, hydrological signatures, or combination thereof select donor subsequently transfer entire sets the receiver catchment. Each approach was tested leave-one-out cross-validation with bucket-type model (the HBV model) using 671 United States diverse hydroclimatology. We then evaluated several hydrograph aspects, compared it against calibration benchmarks, linked descriptors. results this large-sample study summarized three lessons: (a) Catchments benefit well-chosen independent region how well they modeled at best. (b) Almost perfect donors exist most an excellent relative reached current approaches. This implies that there considerable potential improvement prediction (c) ranking depends predicted hydrographs are evaluated. These findings indicate multi-criteria evaluation essential performance.