Characteristics of Recent Vehicle-Related Fatalities during Active Precipitation in the United States

Abstract Data from the National Highway Traffic Safety Administration’s (NHTSA) Fatality Analysis Reporting System (FARS) database were used to identify vehicle-related fatalities that occurred during active precipitation from 2013 to 2017. Changes to FARS for 2013–present allow the identification of freezing rain, in addition to rain, snow, sleet, and precipitation mixtures as prevailing precrash atmospheric conditions. The characteristics of vehicle-related fatalities for each precipitation type identified in FARS were assessed in terms of total numbers, roadway surface conditions, location, and annual and diurnal variability. Vehicle-related fatalities were also matched to nearby Automated Surface Observing System (ASOS) and Automated Weather Observing System (AWOS) precipitation-type reports to determine their agreement with precipitation types documented in FARS. Of the vehicle-related fatalities examined, 8.6% occurred during precipitation, with these fatalities further divided by precipitation type of approximately 81% rain, 14% snow, and 5% sleet, freezing rain, and mixtures of precipitation. Unexpected discrepancies between the numbers of sleet- versus freezing-rain-related fatalities reveal that caution should be taken when using FARS to identify these precipitation types. ASOS/AWOS precipitation-type reports have moderate agreement with FARS at 20 mi (32.2 km), and are capable of distinguishing precipitation and nonprecipitation indicated in FARS. Rain and snow have good agreement between the databases, whereas sleet, freezing rain, and precipitation mixtures have significantly reduced agreement due to a combination of ASOS/AWOS limitations and suspected FARS limitations. To provide a more accurate account of precipitation types for fatal crashes, the use of crashes where FARS-identified precipitation types are confirmed by nearby ASOS/AWOS reports is suggested.