Date of Award

August 2021

Degree Type


Degree Name

Doctor of Philosophy



First Advisor

Zengwang Xu

Committee Members

Alison Donnelly, Ryan Holifield, Woonsup Choi, Margo Anderson, John Logan


hurricane, natural hazards, social vulnerability


Extreme weather events are expected to increase as a consequence of climate change, increasing the intensity and frequency of natural hazards. Their catastrophic impact is attributable to both the geophysical characteristics of a hazardous event itself and the socio-demographic characteristics of people who are at a greater risk of harm in the aftermath of natural hazards. Previous studies have largely used a place-based approach, measuring the relative level of social vulnerability between places using a social vulnerability index (SoVI), a prevalent spatially explicit method in geographic scholarship. As a composite index, SoVI, has been criticized by scholars due to its over-generalization; it cannot indicate the contribution of specific local social indicators to vulnerability, obscuring demographic heterogeneity and making it difficult to understand who is vulnerable. In contrast to the spatiality of vulnerability, the temporal dynamics of social vulnerability have been relatively understudied. This dissertation seeks to address these drawbacks of the SoVI approach and to assess hazard-specific vulnerability by incorporating geophysical characteristics of natural hazards and differential vulnerabilities of affected populations. There are four primary objectives of this study: (1) To investigate major patterns in the spatial and temporal dynamics of social vulnerability of U.S. counties from 1970 to 2010 using quantile standardization, sequence alignment analysis, and cluster analysis; (2) To identify the contributions of the components of SoVI and the local primary factors that contribute to social vulnerability using geographically-weighted principal component analysis (GWPCA) and explore how those factors have evolved over time using Greater Houston as a case study; (3) To estimate the spatial extent and intensity of storm surge inundation and wind damage caused by hurricanes along the Gulf and Atlantic Coasts in the United States from 1950 to 2018 using geospatial analysis; and (4) To understand differential vulnerabilities of distinctive demographics within hurricane at-risk areas using a spatial and temporal analysis. The results show that the U.S. counties have four major temporal trajectories, revealing areas of persistently low and high vulnerabilities and areas with dynamically changing vulnerabilities. The application of GWPCA reveals the most influential local social factors that constitute the SoVI index. Moreover, the spatial and temporal trends of the local factors can indicate what socioeconomic conditions are prevailing and consistently affect the vulnerability of a particular region. In terms of the vulnerability of people to hurricane hazards, this study also identifies generalized patterns of demographic changes that are within hurricane-risk zones and which population groups are increasingly or decreasingly exposed. The results in this study have significant implications for policymakers and national disaster management in surveilling vulnerable areas and establishing potential hazard mitigation plans. The findings reported here shed new light on social vulnerability assessment urging decision-makers to provide more resources to the hardest-hit groups living in the most exposed counties. This study is the first comprehensive investigation of hurricane-specific vulnerability encompassing the Atlantic and Gulf Coasts and at a national scale. The analytical framework suggested in this study can enrich the approach to vulnerability assessment of natural hazards by converging geographic and demographic perspectives.

Included in

Geography Commons