Adverse Weather and Motor Vehicle Fatalities

Introduction: Precipitation results in hazardous road conditions that are associated with motor vehicle fatalities (MVF). We comprehensively assesses the risk of MVF associated with precipitation for the entire United States, highlighting the seasonal, diurnal, and spatial variation.

Methods: We used radar-based precipitation data from the National Oceanic and Atmospheric Administration, available at 1 square km resolution at 5-minute intervals. The geographic coordinates and time of crash for every MVF in the US were obtained for 2006-2011 from the Fatality Analysis Reporting System (FARS). We linked the precipitation values for an hour preceding each crash based on crash time and coordinates. We hypothesized that if the frequency of MVF preceded by precipitation was greater than the 10-year average frequency of precipitation in a certain region, it would indicate a non-random association between MVF and precipitation. We used binomial tests to check this for different amounts of precipitation, across different seasons and US climate regions. We used logistic regression to check if known risk factors (like alcohol use, speeding, improper restraint etc.) were more likely to occur with MVF associated with precipitation.

Results: For seven percent of 193,163 MVF analyzed, some precipitation was observed before the crash. All climate regions showed a statistically significant increase in frequency of MVF associated with any precipitation preceding a crash than would be expected based on local climatology of precipitation. MVF associated with precipitation was higher during the day and in the winter months (except Southwest). Highway driving was the only known risk factor for MVF more likely to happen when MVF was associated with precipitation.

Conclusion: The type of adverse weather and the rate of associated fatality vary geographically. This is a public health concern as climate models project precipitation patterns to intensify and vary more across the country.