Risk assessment of rainfall-induced landslides considering regional characteristics with machine learning

Read full text Summarise with VIRAT
Authors: Tatsuya Ishikawa, Ren Ono, Tomoki Ueno, Kiyonobu Kasama
Cited by: 0
References: 5

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In Japan, the risk assessment of rainfall-induced landslide disasters is based on the risk base line of sediment disasters occurrence (CL: Critical Line) set for each 1 km mesh using two values: the 60-minute accumulated rainfall (RI: Rainfall Intensity), a short-term rainfall index, and the soil precipitation index (SWI: Soil water Index), a long-term rainfall index. However, the current Japanese early warning system is set without considering the ground and topographical characteristics that are the endogenous factors of slope failures, and it assumes debris flows and large-scale landslides that occur intensively and does not target small-scale landslides such as cut slope collapse and embankment collapse. This study aims to develop a wide-area slope disaster risk assessment method for small-scale landslides that occur locally along roads by proposing a method for setting slope disaster risk base line that can be applied to various regions with the machine learning based on the numerical data representing regional characteristics, including topography and geology, in addition to meteorological data such as temperature and precipitation.

Fields of Study & Subjects

Landslides and related hazards Geotechnical Engineering and Analysis Tree Root and Stability Studies

Publication details

Journal / source
Japanese Geotechnical Society Special Publication
Volume
12
Issue
2
Pages
9–18
Publication date
5 March 2026
Language
EN