Threshold-Based Landslide Early Warning Method
Description
This method estimates landslide likelihood by comparing accumulated precipitation
against predefined rainfall thresholds. It integrates both observed and forecast
precipitation data to assess short-term triggering conditions.
Data Sources
- ECMWF Forecast Precipitation ( 10 KM Resolution )
- MSWEP (Multi-Source Weighted-Ensemble Precipitation)
Methodology
- Retrieve precipitation data from ECMWF (forecast) and MSWEP (observed).
- Compute cumulative rainfall using backward aggregation:
- 24-hour: Forecast rainfall
- 48-hour: Forecast + previous 24-hour observed rainfall
- 72-hour: Forecast + previous 48-hour observed rainfall
- 120-hour: Forecast + previous 96-hour observed rainfall
- 168-hour: Forecast + previous 144-hour observed rainfall
- 240-hour: Forecast + previous 216-hour observed rainfall
- Compare accumulated rainfall with predefined thresholds.
- Flag areas where thresholds are exceeded as potential landslide risk.
Output
Binary or categorical landslide likelihood (e.g., No Risk / Potential Risk)
Temporal Resolution
Daily (24h, 48h, 72h, 120h, 168h, 240h accumulation windows)
Spatial Resolution
Based on input precipitation datasets
Key Assumption
Landslides are primarily triggered when rainfall accumulation exceeds critical thresholds.