Injury Severity Analysis along Major Arterial Roads in Kentucky Using High-Resolution Weather Data

Abstract

Recognizing that relatively limited studies have investigated the influence of microscopic real-time weather conditions on crash injury severity along arterial roads, this study took the initiative and comprehensively investigated those factors influencing injury severity along major arterial roads in Kentucky using recent 5,259 crashes between 2019 and 2023. Furthermore, real-time weather information during 1 h of the crash (retrieved from the high-resolution rapid refresh “HRRR” model) was integrated with the crash data. Three modeling approaches (random parameter logit “RPL,” correlated RPL “CRP, ” and CRPL with heterogeneity in means “CRPLHM”) were employed to account for unobserved heterogeneity across the crash observations. The CRPLHM model was deemed the best-fit model compared with the other two models (i.e., RPL and CRPL). The CRPLHM model results showed that “air temperature 50–70°F, relative humidity ≥ 90%, and wind speed ≥ 10 mph” significantly increased severe injury (KA) outcome by 1.19%, 4.12%, and 1.83%, respectively, and minor injury (BC) outcome by 10.60%, 22.10%, and 12.28%, respectively. Additionally, wider medians and right shoulders were associated with reduced likelihood of severe outcomes along arterial roads. The study results helped to propose proactive countermeasures to reduce severe outcomes along major arterial roads, for example displaying safety alert messages through dynamic message signs (DMS) along arterial roads when any of the following weather states is satisfied: “air temperature = 50–70°F,”“relative humidity ≥ 90%,” or “wind speed ≥ 10 mph.” This proactive approach ensures timely weather warnings to drivers, which will promote safer driving behavior in real-time.