查看文章

Precipitation increases the risk of motor vehicle accidents through loss of friction on the road as well as reduced visibility. Given increasing utilization of navigation systems for driver assistance as well as self-driving vehicles, an important question is whether the potential exists to substantively inform navigation decisions with available real-time rainfall data and/or hydrologic models. To explore this potential, matched pair analysis was performed between 6-hourly radar rainfall data and traffic accident data from 2006 to 2017 for the Los Angeles Metropolitan Region. The relative accident rate (RAR) was plotted against average rainfall intensity. As each accident datapoint is also tagged with weather conditions (eg, clear, cloudy, rain) and road conditions (eg, dry, flooded), we tested whether 6-hourly radar rainfall data was predictive of either weather or road conditions as reported by police officers in traffic accident …