Assessing Coastal Erosion Vulnerability: A Case Study of Georgetown County, South Carolina
Advisor: Dr. Susan L. Cutter
For decades, the hazards community has researched the concept of vulnerability, and as the concepts change, so do the definitions and methods used to measure it. In recent years, as populations flock to the coast and the threat of sea level rise is looming, more studies have focused on the vulnerability of coastal regions. This research examines multiple physical, social, and built environment characteristics to determine which locales, people, and infrastructure resources are vulnerable to coastal erosion. The objective of the study is to determine where the most vulnerable places are and what factors contribute to the geographic variation in coastal erosion vulnerability. To reach this goal, the following questions will be answered:
- What geographic locations, populations, and structures are vulnerable to coastal erosion? Why?
- What is the spatial clustering of these vulnerabilities?
- What is the total vulnerability to coastal erosion at a given location? How does this vary depending on integration method?
- Do physical, social, and built environment characteristics contribute equally to the vulnerability or is one more dominant than another?
The research questions are primarily answered using GIS-based techniques, but all methods include spatial overlays, statistical reductions, spatial autocorrelations, sensitivity analyses, and regression analyses. The results from the individual vulnerability assessments found locations along the coast are more physically prone to coastal erosion, while those inland are less physically vulnerable. The social characteristics increasing vulnerability are largely race and class (poor/black populations) and large populations of service industry workers and immigrants, which were found near the industrial core of Georgetown and in southern rural areas. Low vulnerability was also characterized by race and class (wealthy/white populations), as well as small populations of service industry workers and immigrants. Areas where the built environment was most vulnerable were in structurally dense downtown areas, and it was least vulnerable in very rural areas. Each individual index had some positive spatial autocorrelation. The coastal erosion vulnerability index had the highest for the county, followed by the built environment and then the social vulnerability index. After incorporating the models to create a total vulnerability index, the results give a different outlook on vulnerability than the individual ones. A sensitivity analysis shows changing the weights of the indices produce different results, but changing the method of standardization has little impact on the overall results. A regression analysis shows that each set of characteristics do not contribute equally to the total vulnerability. Rather, the physical characteristics are most influential followed by the social and built environment characteristics.
This research serves as a practical tool for emergency managers to identify vulnerable areas along a number of dimensions. This will more accurately and effectively target mitigation measures to correspond with the drivers of place vulnerability. However, the model construction is time consuming and varies by location and scale of analysis. Further research should investigate methods of automating the index building process, as well as allowing variations in location and scale. Also, future research should address ways of keeping data used in the index up to date. Other aspects of vulnerability should be investigated and incorporated into the index for an even more complete understanding of coastal erosion vulnerability.