Objective: The paper aims to explore the possibility of using traffic violations as indicators for spatial-temporal risk of traffic safety within road network constraints, identify key types of traffic violations that indicate spatial-temporal risks in road traffic safety, and investigate their distribution patterns at the road section level.
Methods: Firstly, we employ the Ripley's K function with network constraints and utilize rigorous statistical inference to thoroughly examine the spatial-temporal correlation between various types of traffic violations and crashes, identifying key types that exhibit significant correlation with crashes. Secondly, we combine Ripley's K function with network constraints, Network Kernel Density Estimation, and Local Moran's Index, to identify high-incidence road sections of these violations. Building upon this foundation, we introduce the concept of Influence Intensity for Land Use Type, which leverages Point of Interest information to analyze the land use characteristics at the road section level, revealing the distribution patterns of these key traffic violations.
Results: Analysis of actual data from Shenzhen, China reveals a total of 17 key traffic violations significantly correlated with crashes of varying severity across different time scenarios in the spatial ranges of 2.1-3.8 kilometers. These include types that are typically considered to have a relatively low likelihood of directly causing crashes that deserve more attention. These key traffic violations tend to aggregate in road sections categorized as "Business & Finance" and "Public Transport Infrastructure." Furthermore, in contrast to weekdays, weekends witness a higher number of key traffic violation types with more pronounced spatial aggregation characteristics, and the land use type of aggregation areas shifts from "Public Administration & Services" to "Public Green Spaces & Attractions" and "Residence & Living."
Conclusions: This study demonstrates that particular traffic violations can serve as signals for road traffic safety risk within specific space-time scopes, and the spatial-temporal aggregation patterns of these key traffic violations are closely linked to the urban land use. This finding can offer theoretical support for utilizing key traffic violations in real-time monitoring and early warning of road traffic crashes, while also providing inspiration for exploring the causes of these traffic violations from a land use perspective.
Keywords: Ripley’s K function; Spatial-temporal risk of traffic safety; land use; road network constraints; traffic violation.