Understanding and Zoning Rainfall-Induced Landslide Hazards in Indonesia: Insights from Observation to Forecasting

As one of the most destructive natural hazards, landslides pose persistent threats to human life, property, and critical infrastructure in Indonesia, where intense rainfall and steep, complex terrain strongly control landslide occurrence and impacts. Although landslides may be triggered by multiple factors, including earthquakes and prolonged rainfall, rainfall remains the only trigger that can be forecasted, making it central to operational landslide early warning. Between 2019 and 2024, based on Indonesian Disaster Information Database (DIBI–BNPB), more than 4,000 landslides were recorded across Indonesia, causing substantial loss of life and widespread damage to housing and public infrastructure.

At present, landslide early warning in Indonesia relies on a single nationwide rainfall threshold, which may limit forecast accuracy and reliability given the country’s strong spatial variability in rainfall patterns and geomorphological conditions. Developing rainfall thresholds at large spatial scales is therefore challenging. To address this limitation, this study adopts a zoning approach that prioritises areas with high landslide susceptibility and potentially severe impacts, providing a targeted basis for subsequent threshold development.

Landslide susceptibility maps are produced using the Analytical Hierarchy Process (AHP), chosen in preference to data-driven methods due to biases and incompleteness in the available landslide inventory, which tends to reflect population distribution rather than true landslide source areas. Two provinces, Central Java and South Sulawesi, are selected as initial case studies. According to the data from Local Indonesian Disaster Management (BPBD), more than 2,000 landslides were recorded in Central Java between 2016 and 2025, while over 500 events were documented in South Sulawesi between 2021 and 2025.

Population density, building distribution, landslide susceptibility, and landslide runout probability are integrated to identify zones with the highest potential impacts. These high-impact zones serve as priority areas for developing more representative rainfall thresholds, with the aim of improving landslide forecasting and risk reduction in Indonesia.