Reliability of imprecise information of mobile location data and the influence of social demographics on evacuation traffic

Hurricane evacuations are difficult to reconstruct when endeavoring to explore and analyze the movements of evacuees, making it difficult to validate model predictions with actual evacuation traffic.Mobile location data offers an opportunity to reconstruct evacuation traffic and associate mobile evacuees with place-based social demographics such as income, race/ethnicity, and family size.These data offer a unique insight into the evacuation travel behavior of different social groups but is imprecise in several aspects.The objective of this study is twofold: to analyze the reliability of mobile location data to capture evacuation traffic and to explore the influence of social demographic variables on travel behavior.Specifically, the study analyzes the timing of evacuation during Hurricane Michael, which affected Florida, USA, in 2018.The limitations of mobile location data are reviewed through a sensitivity analysis comparing mobile location data with actual traffic counts.Results comparing travel behavior between demographic groups show that the high-income group tend to increase trips, while low-income and Black groups show a tendency to decrease trips.This presentation contributes to evacuation modeling theory by exploring the social influences of evacuation and evacuation modeling practice and demonstrating the applicability and limitations of imprecise information of mobile location data to reconstruct evacuation behavior.