Social vulnerability in the coastal region of Bangladesh: An investigation of social vulnerability index and scalar change effects

Abstract

Vulnerability to natural hazards not only depends on the magnitude of the hazards but also on the socio-economic conditions of people. Considering vulnerability as a social construct, the social vulnerability index (SoVI) was introduced to measure the vulnerability of people across space. Very little is known about the social vulnerability of the coastal region of Bangladesh despite their high exposure to natural hazards like cyclones and coastal flooding. Applying modified SoVI, we investigated the social vulnerability of this region at the union level. Using Principal Component Analysis (PCA), twenty-seven variables are reduced to nine components, which explain 80.79% of the variance in the data. We mapped the composite SoVI score, which was calculated by adding the nine principal components. Results show that 24.26% of the areas of the region are high to very highly vulnerable and most of these areas are either situated near the Meghna estuarine or in the Cox's Bazar district. Results from Local and Global Moran's I show that there are significant clusters of composite scores of social vulnerabilities. Furthermore, we investigated the effects of scalar changes (from the union to mouza) on the distribution of SoVI. Introducing a binary classification method to SoVI analysis, we found that scalar change reduces about 30% of the accuracy of vulnerability classification. The findings of this study will help policymakers and disaster managers to formulate informed policy and disaster management plans.