Differentiating high-frequency and high-severity hotspots: A robust risk-evolution-volume (REV) framework

Area of Science:

• Transportation Engineering
• Traffic Safety Research
• Spatial Data Analysis

Background:

• Traditional crash hotspot identification methods may overemphasize high-volume roads.
• Limited insight into whether hotspots relate to crash frequency or severity.
• Need for a more nuanced approach to road safety management.

Purpose of the Study:

• To propose a generalizable risk-evolution-volume (REV) framework for differentiating high-frequency and high-severity crash hotspots.
• To apply the framework using crash data from Jiaozhou City, China (2022-2024).
• To inform targeted countermeasures and resource allocation in road safety.

Main Methods:

• Severity-weighted network kernel density estimation (SW-NKDE).
• Threshold screening using Hurdle Gamma model cumulative distribution.
• Spatiotemporal evolution analysis.
• Two-dimensional relative risk assessment (critical crash rate [CCR] and severity-weighted critical crash rate [SWCCR]).

Main Results:

• A clear divergence exists between crash frequency and severity patterns.
• Urban signalized intersections show higher frequency; suburban segments/intersections show higher severity.
• High crash density does not always equate to abnormal risk when traffic exposure is considered.

Conclusions:

• The proposed REV framework effectively distinguishes between crash frequency and severity hotspots.
• The framework can be adopted by traffic authorities for network-level screening and priority setting.
• Results inform targeted interventions and efficient resource allocation for road safety improvement.