Analyzing the spatial scale effects of urban elements on urban flooding based on multiscale geographically weighted regression

• Published in: Journal of hydrology (Amsterdam) 2024-12, Vol.645, p.132178, Article 132178
• Main Authors: Wu, Meimei, Wei, Xuan, Ge, Wei, Chen, Guixiang, Zheng, Deqian, Zhao, Yang, Chen, Min, Xin, Yushan
• Format: Article
• Language: English
• Subjects: Flooding; MGWR; Spatial scale effects; Urban elements
• ISSN: 0022-1694
• DOI: 10.1016/j.jhydrol.2024.132178

[Display omitted] •There are spatial scale effects in the influence of urban elements on flooding.•MGWR can be used to quantify the scale of influence of different urban elements on urban flooding.•Elevation, NDO, DR, GDP, PCL, PRL are localized influence elements.•RRP, Slope, PIL, PPSL, PRA, PGL, ILMFCP, PD, MI, ILER are global influence elements.•Apart from rainfall, social elements are the most critical elements affecting urban flood damage. In the context of rapid urbanization and the frequent occurrence of extreme rainfall in cities, the risk of flooding in the future will further increase, and the problem of urban flooding cannot be ignored. Urban elements exhibit significant spatial heterogeneity, which largely determines the spatial distribution differences in urban flooding. Therefore, it is important to clarify the scale of influence of different urban elements and explore their scale effects on urban flooding to accurately assess the risk of urban flooding. Taking Zhengzhou City, China, as the study area, this study analyzed the urban elements associated with urban flooding, quantified the scale of the influence of urban elements on flooding using multiscale geographically weighted regression (MGWR), and further explored the spatial scale effects of urban elements on urban flooding. The results showed that MGWR can better fit the spatially non-uniform distribution of urban flooding and that the scale of the influence of urban elements on urban flooding can be reflected by the bandwidth of MGWR. The results of MGWR indicated that the bandwidths of elevation, number of drainage outfalls (NDO11Abbreviations: RRP, Rainfall return period; DDR, Distance to river; NDO, Number of drainage outfalls; ILMFCP, Input level of materials for flood control projects; PRL, Proportion of residential land; PCL, Proportion of commercial land; PIL, Proportion of industrial land; PPSL, Proportion of public service land; PRA, Proportion of road area; PGL, Proportion of green land; PD, Population density; MI, Manpower input; GDP, Gross domestic product; ILER, Investment level in education and research.), distance to river (DR), Gross domestic product (GDP), proportion of residential land (PRL), and proportion of commercial land (PCL) were small, and their influence scales were localized. In contrast, the influence of rainfall return period (RRP), slope, proportion of industrial land (PIL), proportion of public service land (PPSL), proportion of road area (PRA), proporti