An Integrated Approach for Unsupervised Data-Driven Landslide Prediction System

Abstract

We have developed a landslide prediction system, based on the integration of geomorphological, geological and climatic information. The approach to developing the system was to make forecasts using slowly varying factors (geomorphological parameters) and factors with high seasonal variability (soil humidification parameters and rainfall quantities). In this sense, the system involves integration with various real-time data sources such as precipitation forecasting systems, rain gauges and SRS systems. Our objective is to estimate the landslide risk based on the parameters provided as input to the system according to the scale 1. Very low, 2. Low, 3. Medium, 4. High, 5. Very high