Physically consistent conceptual rainfall–runoff model for urbanized catchments

Hydrological models should be tested and evaluated for a wide variety of levels urbanization before they are used to predict the impact on catchment behavior. In this study, we illustrate top–down approach modifying step by an hourly conceptual model structure (GR4H) account features. Modifying original included accounting explicitly runoff from impervious surfaces bypassing soil moisture reservoir varying partitioning between quick flow slow flow. These adaptations were chosen based reported specificities urbanized catchments, namely, decreasing infiltration, increasing runoff, fast dynamics. Using split-sample test, relevance each modification with regard reproduction response (i.e., observed streamflow) was assessed large sample 273 located in France United States, which mean total area (TIA) varied 0.05 0.59. Six continuous three event-based criteria used, two statistical tests applied assess significance improvements. Results showed following: (i) Tested modifications improved ability reproduce response, especially high flows streamflow amid dry conditions. (ii) Event-based evaluation using more than 45,000 events improvement predicting event peak volume, whereas no significant improvements obtained timing (iii) Newly added parameters moderately highly correlated TIA, calibrated proportion surfaces, is promising as hydrological validation estimated measures land cover. The both representation processes streamflow, yielding simple credible future scenarios response.