Spatiotemporal patterns and driving mechanisms of provincial road traffic accidents in China, 2016-2024

Road traffic accidents have become a major public health issue of global concern.

Drawing on panel data from 31 provincial-level regions in China from 2016 to 2024, this study develops an evaluation framework comprising ten indicators related to transport infrastructure, socioeconomic development, and the natural environment. It employs the imbalance index, geographical concentration index, standard deviational ellipse, spatial autocorrelation analysis, and the geographical detector model to examine the spatiotemporal evolution and multidimensional driving mechanisms of provincial traffic accidents in China.

The results show substantial interprovincial imbalance but weak overall spatial adjacency. Global Moran's I does not indicate significant spatial correlation, while the distribution center shifted generally southwestward and the standard deviational ellipse expanded and then contracted. The geographical detector results indicate that the number of motor vehicle drivers, resident population density, gross regional product, and civilian vehicle ownership are the main factors explaining spatial differentiation. Temperature, precipitation, and total road mileage have moderate individual explanatory power, while higher educational-attainment strata are associated with lower mean accident counts. Most factor pairs exhibit bivariate or nonlinear enhancement.

These findings suggest that traffic accidents are shaped by the multidimensional coupling of "people-vehicles-roads-environment" and provide evidence for regionally differentiated traffic-safety policies.