Democratizing Safety During an Earthquake: Multi-Disciplinary Spatial Optimization for Defining Evacuation Centers with Fair Access in L.A. County

Abstract:

Data-driven evacuation modeling for disaster response and planning is particularly important for populous cities with high earthquake risk. Weighted overlay analysis is a common approach to resolve subsurface data on earthquakes and human-driven factors to plan for relief efforts. In this study, a multidisciplinary GIS workflow is presented to define earthquake evacuation centers for Los Angeles (L.A.) County considering geologic and socio-economic drivers. Environmental covariates pertinent to earthquakes in L.A. County such as distance to faults and soil type are integrated within a multi-criteria optimization function to model evacuation areas. Proposed evacuation areas are updated with a spatial allocation algorithm to ensure fair access regarding socio-economic drivers and population density. A suitability score is designed with respect to access to shelter locations. Service areas of proposed shelters are defined as regions with a similar population using numerical regionalization. Results indicate spatially extensive evacuation areas in the northeast part of the county with smaller areas in the south are appropriate for shelter locations. Numerical analysis shows that only 30% of the earthquake shelters designated by the American Red Cross are in regions with high suitability scores. The study highlights the impact of socioeconomic drivers on defining effective earthquake shelters and the need for revisiting L.A. County’s earthquake evacuation plans.