A multi-variable threshold approach to landslide prediction: integrating rainfall, soil moisture, and soil texture in Indonesia
Year: 2026
Type: article
Abstract: Landslides are one of the most devastating hazards globally, including in Indonesia. To reduce casualties, Indonesia has established the Blueprint of Indonesian Landslide Early Warning System (BILEWS). This system currently relies on rainfall-induced landslide thresholds for issuing the warning signals derived from historical landslide occurrences in Java, Sumatra, and Kalimantan Islands. However, landslide occurrence is strongly influenced not only by rainfall but also by antecedent soil moisture conditions and soil characteristics. Therefore, BILEWS still faces technical challenges in accurately predicting landslides, especially in Sulawesi and Papua Islands, where Java’s threshold is applied there. This study aims to enhance the existing rainfall thresholds and incorporate soil moisture into a multi-variable threshold framework. Specifically, rainfall thresholds, soil moisture thresholds, and the combination of rainfall-soil moisture thresholds are developed. In addition, the relationship between soil moisture and landslide occurrence is evaluated across different soil textures. The results indicate that the optimal thresholds vary across islands. The existing rainfall thresholds yield the highest performance in Java and the Sumatra Islands with Youden Index of 0.545 and 0.740, respectively. In Sulawesi Island, the best performed threshold is the newly developed rainfall threshold with a Youden Index of 0.477. The rainfall-soil moisture thresholds show optimum performance for Kalimantan (0.465) and Papua (0.458) Islands. Furthermore, soil moisture significantly affects landslide risks in sandy clay loam and loam soil textures. Developing thresholds based on significant soil textures improves the performance of landslide prediction.
Source: Natural Hazards
Language: English
Subfield: Management, Monitoring, Policy and Law
Field: Environmental Science
Domain: Physical Sciences
SDG: Life on land
Open Access status: green
OpenAlex ID: W7167609669
