MULTIVARIATE STATISTICAL ANALYSIS OF PHYSICAL VULNERABILITY METHODS IN TROPICAL COASTAL ZONES

Reduce subjectivity, common in studies of physical vulnerability, was the aim of this research, which evaluated the methods Natural Vulnerability (NV), Environmental Vulnerability (EV) and the proposed Environmental Vulnerability to mean Sea Level Rise (EVSLR) through Principal Component Analysis (PCA) of the environmental variables and associated weighting. The coastal municipalities of Macau and Guamare, in Rio Grande do Norte state, northeastern Brazil, feature multiple conflicts of land use/land cover since the main economic activities consist in the oil and gas industries, salt and shrimp farming, and wind power installed in segments of great environmental susceptibility to climate change impacts. The method with more pessimistic outcome was the NV and the method with a more optimistic was the EVSLR, who presented hotspot areas to higher relative sea level rising. The municipality of Guamare presented upper physical vulnerability in all methods by displaying its coastal zone in flat areas with low heights. According to the PCA all variables were important in NV and EV methods, but with different intensity in each area. In EVSLR method the most important variables were geology, slope, and soil. These results may support the decision-making process for environmental managers and tested methods are replicable in both coastal as inner municipalities.