Much of the research regarding traffic crashes considers various geometric roadway features; however, ever-evolving urban watersheds and climate change increasingly impact roadway conditions. Little research has focused on the relationship between high-resolution drainage characteristics and the spatial distribution of crashes. This study incorporated local environmental and drainage risk factors in assessing network safety performance using spatial analysis techniques. Kernel density surfaces and the Local Getis Ord Gi* statistics were used to identify and visualize locations prone to experiencing crashes during wet conditions. Spatial regression modelling was used to link crashes to environmental and traffic risk factors across a citywide network. Proof-of-concept for this framework is demonstrated in the City of Philadelphia, Pennsylvania using publicly available spatial data. The results of this study show a relationship between local drainage, environmental characteristics, and wet crash distribution, providing novel insight into roadway safety during wet conditions.

许多关于交通事故的研究都考虑了各种几何道路特征；然而，不断变化的城市流域和气候变化对道路状况的影响越来越大。很少有研究关注高分辨率排水特征与碰撞空间分布之间的关系。本研究将当地环境和排水风险因素纳入使用空间分析技术评估网络安全性能。核密度表面和 Local Getis Ord Gi* 统计数据用于识别和可视化在潮湿条件下容易发生碰撞的位置。空间回归模型用于将碰撞与整个城市网络中的环境和交通风险因素联系起来。该框架的概念验证在费城进行了演示，宾夕法尼亚州使用公开可用的空间数据。这项研究的结果显示了当地排水系统、环境特征和潮湿碰撞分布之间的关系，为潮湿条件下的道路安全提供了新的见解。