LANDSLIP knowledge product : landslide forecasting

A landslide forecast provides an estimation of the likelihood of landslides occurring within a specific location and time period in the future. Regional-scale landslide forecasting systems rely on landslide forecasting models. The forecasting models developed in LANDSLIP focus on forecasting rainfall-induced landslides at a regional scale.They build on historical information about landslide occurrence and their associated rainfall trigger conditions in the two pilot study areas. LANDSLIP’s prototype models run daily, and attempt to forecast the likelihood of landslides occurring in these study areas, given rainfall forecasts and information based on local landslide susceptibility. This document provides an overview of the landslide forecast models used or explored by LANDSLIP. It describes the various approaches used, their main assumptions and limitations, and highlights model calibration and evaluation phases. It concludes with a summary of recommendations for others interested in landslide risk management, data management, decisionmaking and communication of risk.