Study on the sensitivity of urban inundation and watershed flood simulation to rainfall data spatial resolution

Under the background of climate change and urbanization, the localized characteristics of rainfall are becoming increasingly pronounced. The spatial resolution of rainfall data often fails to meet the requirements for accurately describing spatial distribution. Conversely, using fine spatial resolution rainfall data result in resource wastage. Based on observed and designed rainfall data, this work progressively increases grid size through spatial interpolation algorithm to achieve lower spatial resolution. By integrating spatial parameters such as rainfall center and regional characteristics, a rainfall dataset with varying spatial resolutions is constructed. This rainfall dataset is used to drive the hydrological-hydrodynamic coupled model, analyzing the influence of changes in rainfall spatial resolution on urban inundation and flood, and identifying the optimal spatial resolution of rainfall data. The results show that for urban inundation simulation, as the spatial resolution of rainfall decreases, the inundation water volume, inundation area, and inundation water depth all increase, revealing a tendency to overestimate inundation risk. For watershed flood simulation, as the spatial resolution decreases, the peak flow decreases, and the discharge process show a more flattened trend. The optimal spatial resolution of rainfall data for urban inundation simulation is 0.5 km, while for watershed flood simulation, it is 4 km.