Urban Flood Disaster Analysis Based on the GWR Model

Driven by population growth, economic development, and climate change, frequent urban flooding has significantly impeded socioeconomic progress worldwide. Understanding the mechanisms underlying urban flood disasters is crucial for sustainable urban development. This study employs the Geographically Weighted Regression (GWR) model to investigate the factors influencing urban flooding in the Taocheng District of Hengshui City, China. The results demonstrate that the Multi-scale Geographically Weighted Regression (MGWR) model outperforms the global regression model, revealing pronounced spatial heterogeneity. Notably, maximum rainfall intensity and manhole density are identified as significant explanatory variables affecting floodwater depth, with coefficients of 0.177 and −0.096, respectively. Maximum rainfall intensity exacerbates floodwater depth, whereas manhole density exhibits a negative effect. The GWR model effectively captures the spatial variability of these influences, providing valuable insights for urban flood risk management.