Identification and clustering of drought-prone areas based on geographical, climatic, and socio-economic indicators: Supporting sustainable environmental management policies

Abstract

Background: Drought is one of the most crucial environmental issues with widespread impacts across various regions of the world, including Indonesia, where changes in rainfall patterns and land use exacerbate the condition. This study aims to identify and cluster drought-prone areas in Central Sulawesi based on geographic, climatic, and socio-economic indicators, with 22 variables representing these three indicators. Methods: The analysis uses a quantitative approach based on data mining through the K-Means Clustering technique. Secondary data from 2019 to 2025 were integrated from multiple agencies, including BPS, BMKG, and BNPB, while the optimal number of clusters was determined using the Silhouette method executed via RStudio. Findings: The analysis results show three clusters with different levels of drought vulnerability, namely Cluster 1 (high drought), Cluster 2 (moderate drought), and Cluster 3 (low drought). Cluster 1 is characterized by high temperatures, low rainfall, and intensive mining activities. Cluster 2 has moderate rainfall and better environmental conditions. Cluster 3 shows relatively stable hydrological and socio-economic conditions. Climatic factors, particularly rainfall are the most influential indicators of drought vulnerability. Geographical factors such as irrigated areas and the extent of forests and water bodies also contribute, as do socioeconomic factors such as population density, poverty levels, and access to clean water. Conclusion: This analysis provides a spatial overview of the distribution of drought risk and serves as a scientific basis for policy formulation. The analysis then provides policy recommendations, including irrigation development and water conservation in moderately vulnerable areas, sustainable resource management in low-risk areas, and green economy development in safe areas to support sustainable environmental management. Novelty/Originality of this article: The novelty of this study lies in its integrated multidimensional approach, combining geographic, climatic, and socio-economic indicators through K-Means Clustering to map drought vulnerability in Central Sulawesi.