Abstract: Identifying sensitive parameters and reducing model uncertainty is a key step to improve urban flood simulation accuracy. This paper constructs an urban flood simulation model for the Jincun District of Jincheng City, identifies sensitivity parameters using a local method and a global method separately, and examines parameter sensitivity for different land surface conditions. Based on sensitivity analysis, flooding and waterlogging characteristics under different land surface conditions are simulated for different rainfall scenarios and urbanization processes. The results show that the puddle storage in impervious areas is more sensitive. The sensitivity of maximum infiltration rate is increased locally and decreased globally as the rainfall return period increases. Parameters related to impervious surfaces are more sensitive in the areas with more hardened ground surfaces. Flooding and waterlogging in different areas are more severe with a longer return period. Under the same return period, waterlogging is more concentrated under the condition of urbanization. This study reveals parameter uncertainties and flooding and waterlogging characteristic responses for application of the urban distributed rainfall-flood model and urban resilience enhancement.

Key words: urban flooding, sensitivity, uncertainty, distributed model, urban resilience, storm water management model