Analyzing spatial variance of urban waterlogging disaster at multiple scales based on a hydrological and hydrodynamic model

The frequency of urban waterlogging is increasing significantly under the combined influence of natural factors (precipitation and terrain) and anthropogenic factors (drainage system and urbanization). Previous studies had explored the effect of landscape pattern and topography on waterlogging based on historical waterlogging events records. However, the research on current waterlogging issues based on historical records had limitations since the impact factors of waterlogging are inconsistent due to the changes of surface and meteorological conditions. This paper applied a hydrological and hydrodynamic model named InfoWorks ICM, to simulate the urban waterlogging depth (UWD). Under the consistent surface and meteorological conditions, UWD were selected as the dependent variable to analyze the influence of landscape pattern and topography on waterlogging at multiple scales. Pearson correlation analysis and stepwise regression models were used to discover the relationship between these indices. According to the results, in terms of landscape composition, the percentages of built-up area and urban green space have the most significant influence on waterlogging. In addition, organizing average built-up area patch sizes and integrating green spaces with complex shape and high connectivity can improve the state of urban waterlogging. Besides, the rational allocation of topographic gradient is an effective measure at small scale. The adjusted R 2 of regression model were 0.723 at 400 m analysis scale, 0.323 at 600 m analysis scale, and 0.193 at 800 m analysis scale, indicating that attention should be paid to scale effect in similar research. This research can provide a reference for mitigating urban waterlogging disasters.