Cluster Analysis of Drought Damage to Local Governments Across the Country Using the National Drought Information Statistics

Abstract

Recently, Korea has regularly published statistics on the current state of annual droughts that have been occurring since 2018 in the National Drought Information Statistics. For a clear categorization of drought damage, it is necessary to first establish the major factors that affect the degree of damage. In this study, damage-related data from the National Drought Information Statistics Collection (2018-2020) were examined to obtain the regional characteristics of water use resulting in regional drought damage to 161 local governments in Korea. Cluster analysis was performed, and a hierarchical clustering algorithm was used for the regional classification of drought damage. The factor composition was analyzed by dividing it into four major methods to represent the drought damage caused by general, living, and agricultural droughts. The analysis confirmed that regional clustering appeared according to the characteristics of water use and past damage history. If such clustering-based regional damage classification information and factors affecting drought damage by water are considered, it is expected that the damage sensitivity of local governments according to the drought situation and regional water use characteristics can be presented quantitatively.