EXAMINING PUBLIC RESPONSE TO IMPACT-BASED ICE STORM FORECASTS

Abstract

Ice storms are potentially devastating weather events in which at least ¼ of an inch of ice accumulates on surfaces due to freezing rain. The impacts of these storms include power outages, travel issues, road shutdowns, personal injury, and cold exposure. A growing body of literature in the risk communication space has focused on the use of impact-based warnings and forecast products in which the forecast includes a description of what impacts are possible. Across many types of disasters and weather events, such as hurricanes, tornadoes, earthquakes, and nuclear plant emergencies, impact-based warnings have been proven to be effective at increasing threat perceptions and encouraging recipients to take protective action. However, significantly less research on impact-based alerts has focused on winter weather, particularly ice storms. In this study, I use content analysis of NWS Facebook posts to determine that current forecast practices for ice storms are inconsistent across offices regarding what impact information is included. Additionally, I develop the Impacts of Ice Storms Survey to measure risk perception about ice storms and likelihood of taking protective action. This includes an experimental component in which I directly test different types of forecast products to determine whether impact-based messaging leads to changes in behavior or perception of the severity of ice storms. Through this study, I do not find sufficient evidence that a causal relationship between forecast format and protective action or perceived severity exists.