A dataset on the spatio-temporal distributions of traffic accident in China

Road traffic injuries and fatalities are critical issues in global public health and urban governance; consequently, related research necessitates the support of accident data with high spatiotemporal resolution. This study constructs and publicly releases the first dataset detailing the spatiotemporal distribution of traffic accidents in China, comprising approximately 1.57 million records. These records provide precise geographical coordinates (longitude and latitude), timestamps, and fatality outcomes for the documented traffic accidents. The dataset is derived from publicly accessible national court judgment records in China. This research utilizes large language model techniques to extract road-level traffic incident records from unstructured judicial texts characterized by extensively long contexts. This dataset provides a higher-precision spatiotemporal infrastructure for interdisciplinary research across fields such as traffic safety, urban planning, public health, economics, and social governance. Furthermore, it makes a significant contribution toward supporting the realization of the World Health Organization's Sustainable Development Goal (SDG) Target 3.6, which aims to halve the number of global deaths and injuries from road traffic accidents by 2030.