Citation

Sung CH, Liaw SC. Int. J. Disaster Risk Reduct. 2020; 46: e101531.

Copyright

(Copyright © 2020, Elsevier Publishing)

DOI

10.1016/j.ijdrr.2020.101531

PMID

unavailable

Abstract

Owing to the escalating environmental hazards caused by climate change, the mitigation of disaster becomes extremely important. The investigation of social vulnerability is a prerequisite for formulating a mitigation plan to environmental hazards. This research applies a GIS-based approach with the Social Vulnerability Index (SoVI) to investigate and quantify the social vulnerability to environmental hazards in Yilan County, Taiwan. In order to construct the SoVI, the literature review was conducted, and 12 variables were selected. Through Principal Component Analysis (PCA), the 12 variables were reduced into four principal components. In order to explore the spatial pattern of SoVI, the spatial autocorrelation analysis was applied. The result showed that there were 26.5% of communities in Yilan County with a high level of SoVI, and most of these communities were mainly located in mountain areas. The unfavorable topography features cause the distributions in mountain areas. On the other hand, there were 37.3% communities with a related low level of SoVI, and these communities were located in plain areas. The inaccessibility caused by topography creates an incapability, resource-lacking environment and lead to a high value of SoVI. In addition, this research applied Geographically Weighted Regression (GWR) to validate SoVI, and the result of the R2 value was 0.769. Also, the standardized residuals showed no spatial autocorrelation, meaning the SoVI had the adequate explanatory ability. This research provided a set of valid indicators to explore the social vulnerability for decision-makers to formulate the mitigation plan of environmental hazards. Besides, SoVI is a suitable tool for visualizing and quantifying the potential loss to environmental hazards.

Language: en

Keywords

Geographically weighted regression; Social vulnerability; SoVI; Spatial autocorrelation