Scholarlink

<p>Focusing on the influence of the road network structure metrics on flooding, this study systematically investigated flood depth and velocity distributions across seven representative road network forms using a two-dimensional hydrodynamic simulation framework. Ten structure metrics of road networks encompassing topological connectivity, topographic drivers, geometry, and local structure, were proposed and examined through Spearman correlation analysis and Machine Learning methods, with further validation using data from Jiangning District (Nanjing). The results show that the regular grid form exhibits the most effective drainage performance, with peak depth 21.67% lower than the average maximum of the other forms, while cul-de-sac-type networks are more flood-prone. Statistical analysis identifies Elevation (EV, r up to -0.537) and Depression Degree (DD) as the most influential drivers for flooding, while Edge Betweenness Centrality (EBC) and Depth Value (DV, r up to -0.421) also strongly affect runoff pathways. Furthermore, SHAP-based interpretability analysis demonstrates pronounced spatial heterogeneity in the influence of key metrics (with local SHAP_EV ranging from − 0.072 to 0.675) across different geographic units, which is based on the Geographically Weighted Random Forest (GWRF) model. These findings underscore the critical role of road network structure in regulating urban flooding, and provide a novel quantitative framework that enhances flood prediction and supports effective flood risk reduction and mitigation strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analyzing Pluvial Flooding Influenced by Urban Road Network Metrics Based on Hydrodynamic Simulation and SHAP Values

  • 期刊论文

  • 发表日期：

    2025年12月26日

查看全文
  • Yanfen Geng,
  • Peng Liu,
  • Xiao Huang

摘要

Focusing on the influence of the road network structure metrics on flooding, this study systematically investigated flood depth and velocity distributions across seven representative road network forms using a two-dimensional hydrodynamic simulation framework. Ten structure metrics of road networks encompassing topological connectivity, topographic drivers, geometry, and local structure, were proposed and examined through Spearman correlation analysis and Machine Learning methods, with further validation using data from Jiangning District (Nanjing). The results show that the regular grid form exhibits the most effective drainage performance, with peak depth 21.67% lower than the average maximum of the other forms, while cul-de-sac-type networks are more flood-prone. Statistical analysis identifies Elevation (EV, r up to -0.537) and Depression Degree (DD) as the most influential drivers for flooding, while Edge Betweenness Centrality (EBC) and Depth Value (DV, r up to -0.421) also strongly affect runoff pathways. Furthermore, SHAP-based interpretability analysis demonstrates pronounced spatial heterogeneity in the influence of key metrics (with local SHAP_EV ranging from − 0.072 to 0.675) across different geographic units, which is based on the Geographically Weighted Random Forest (GWRF) model. These findings underscore the critical role of road network structure in regulating urban flooding, and provide a novel quantitative framework that enhances flood prediction and supports effective flood risk reduction and mitigation strategies.