Traffic Collisions in Dubai: Temporal Patterns and Severity-Weighted Hotspot Analysis

Abstract

Understanding traffic collision patterns is important for road safety planning in fast-growing urban environments. This study examines temporal and spatial patterns of traffic collisions in Dubai, United Arab Emirates, with emphasis on collision severity. Traffic incident records from November 5, 2024, to June 2, 2025, were obtained from the Dubai Police open dataset. After spatial validation, 33,480 collision records were analyzed, including 1,365 pedestrian-related collisions, representing 4.07% of the dataset. Hourly, daily, and month-to-month variations in collision frequency and severity were examined within the available study period. Temporal associations with severity were evaluated using chi-square tests and Cramér’s V, while spatial patterns were analyzed using severity-weighted Getis–Ord Gi* hotspot analysis, complemented by inverse distance weighting interpolation. The results show observable temporal variation in collision frequency and severity proportions, with higher collision frequencies during evening and nighttime periods. For overall collisions, severity is significantly associated with time of day, day of week, and month, although the effect sizes are small or negligible. The observed proportion of high-severity outcomes is 44% higher during nighttime than during the afternoon. Pedestrian-related collisions show higher occurrence during late-evening hours, but pedestrian severity is not significantly associated with the temporal variables considered. Spatial analysis identifies severity-weighted collision concentrations in northern and northwestern Dubai and along the Al Ain–Dubai Highway, while pedestrian hotspots are concentrated near the industrial area in the southwestern region. The findings provide a severity-focused diagnostic basis for identifying locations that may warrant further investigation and targeted safety assessment in Dubai.