Investigating Social Vulnerability in Flood Prone Areas

Abstract

Flooding is a deadly and expensive natural disaster. In the United States, the Federal Emergency Management Agency’s (FEMA) National Flood Insurance Program (NFIP) provides federal flood insurance, on a property-by-property basis, to residents to reduce the burden of addressing damage and property losses after a flood event. Despite this mission, in many regions of the country flood risk mapping is sparse, outdated, or undercounts the number of properties exposed to the 100-year floodplain. In addition, housing type and other sociodemographic factors can impact the ability to prepare for and adapt to natural hazard events. This study uses a high-resolution flood risk data set, FEMA undercounts and bivariate Local Indicators of Spatial Association (LISA) models, to identify areas where high flood risk and a high proportion of properties are unaccounted for by FEMA. Then regression modeling is used within four high flood risk subregions (the Pacific Northwest, Central Appalachia, the Gulf Coast, and the Southeast Atlantic Coast), to identify socially vulnerable communities in flood-prone areas. The results illustrate that undercounted regions by FEMA coincide with high flood risk and that in large swaths of the four regions socially vulnerable communities will face unprecedented challenges. In all four regions, some housing types, especially mobile homes and multifamily units correlate with high flood risk. In Appalachia, a region where FEMA systematically undercounts properties, poverty and lack of vehicle access correlate with high flood risk. These findings align with previous flood vulnerability studies yet provide a more detailed analysis of regional differences in communities vulnerable to flooding. We also identify pathways in these subregions to reduce the procedural injustice associated with the NFIP.