Evacuation patterns and socioeconomic stratification in the context of wildfires

Wildfires are becoming more frequent and intense, leading to increased evacuation events that disrupt mobility and socioeconomic structures, impacting access to resources, employment, and housing. Understanding the interplay between these factors is crucial for developing effective mitigation and adaptation strategies. We analyse evacuation patterns during the wildfires that occurred in Valparaíso, Chile, on February 2-3, 2024, using high-definition mobile phone records. Applying a causal inference approach combining regression discontinuity and difference-in-differences, we focus on socioeconomic stratification to isolate the wildfire impact on different groups. We find that many people spent nights away from home, with the lowest socioeconomic group staying away the longest. Overall, people reduced their mean and median night-to-night travel distances during the evacuation. Movements initially became irregular but later concentrated in areas of similar socioeconomic status. Finally, we demonstrate a comparability potential of the mobile phone records to the Facebook Disaster Maps, although the latter have a coarse time resolution and are generated only after the wildfire onset. Our results highlight the role of socioeconomic differences in evacuation dynamics, offering valuable insights for response planning.