강우 조건에서의 사망교통사고 공간분석 및 사고다발지점 비교 연구

초록

This study investigates the spatial characteristics of fatal traffic accidents under rainfall conditions and their relationship with conventional accident-prone areas. Using fatal accident data from the Traffic Accident Analysis System (TAAS) for 2012–2023 combined with precipitation data from the Korea Meteorological Administration’s Automated Weather Stations (AWS), rain-related fatal accidents were identified through spatial matching and rainfall thresholds. Kernel Density Estimation (KDE) was applied to detect high-density clusters, and QGIS-based overlay analysis examined overlaps with existing accident-prone zones. Results show that rain-related fatal accidents accounted for 1.8% (794 cases) of all fatal accidents, with 73% overlapping existing hotspots and 27% occurring independently, particularly on rural highways and national roads. KDE analysis revealed significant spatial heterogeneity, with a small proportion of cells exceeding 2 cases/km², indicating localized vulnerability. These findings suggest that rainfall introduces unique risk factors not fully captured by current hotspot identification methods. The study emphasizes the need for a complementary risk management framework that incorporates weather-specific hazards, recommending measures such as adaptive speed control, drainage improvements, and real-time hazard prediction systems linked to weather forecasts. By integrating meteorological factors into traffic safety strategies, this research provides empirical evidence to enhance Intelligent Transport Systems (ITS) and supports the development of autonomous vehicle safety under adverse weather conditions.

키워드