Social Vulnerability and Flood Exposure during Hurricane Sandy

Abstract

Background: Hurricane Sandy caused widespread health and economic impacts in the greater New York City (NYC) area. While these impacts may vary across different communities, analyses understanding and identifying community-level vulnerability are lacking. We compared resilience measures and Hurricane Sandy flooding data to quantify the extent of community-level disparities. Methods: The 2010 Social Vulnerability Index (SVI) was downloaded from the Centers for Disease Control and Prevention to provide a resilience measure at the census tract level across NYC, Long Island and Westchester Counties. The overall SVI captures four dimensions: Socioeconomic, Household Composition, Minority Status/Language, and Housing/Transportation, resulting in a summed percentile ranking for each census tract, where a higher score indicates a greater social vulnerability. Flooding for the same census tracts were obtained from the Federal Emergency Management Agency Modeling Task Force Hurricane Sandy Impact Analysis. Results: 771 census tracts (26.3%) experienced flooding due to Hurricane Sandy, and 533 (69%) of these tract were located in NYC. The overall SVI index (mean±SD) was higher (7.70 ± 1.94) in NYC tracts than non-NYC tracts (5.32 ± 1.80), and higher in tracts experiencing flooding in NYC (7.2 ± 2.10) compared to non-NYC tracts that experienced flooding (5.05 ± 1.74). In NYC, 23.0% of census tracts were >75th percentile of both overall SVI and mean flood height, while 30.6% of tracts met this criteria in non-NYC. Conclusion: The social vulnerability and resilience of communities affected by Hurricane Sandy differed across NYC and Long Island, with a clear division among NYC and non-NYC study areas, and again among flooded areas in NYC and flooded areas outside NYC. There exists an urban population with high SVI and flood exposure at the highest risk, and actions must be taken to reduce disparities in social vulnerability and bolster community-level resilience.