Year: 2025
Type: article
Abstract: Landslides are the third most frequent form of natural disaster in Malaysia, following floods and storms. It can cause significant damage to anything in its path, depending on the size and velocity of its debris. Due to the danger that it poses, determining the susceptibility of an area to landslides is a crucial step in risk mitigation. Landslide occurrences are dependent on the numerous environmental variables, which can provide information on the level of susceptibility of other locations with similar variables. To quantify the significance of each variable to landslide occurrence, a supervised Machine Learning model – an Artificial Neural Network was developed for this study. Furthermore, landslide occurrences have been associated with the disturbance of natural slopes to accommodate development, which was the main reason behind the selection of Western Sarawak as the area of interest in this study. The model was developed to understand and make landslide susceptibility predictions based on aspect, curvature, elevation, lithology type, rainfall intensity, slope angle, soil type, and TWI. Evaluating the area under the curve score and recall for the model revealed that, based on the available inputs, the model performed well with a score of 1 and 0.99, respectively.
Authors: Nur Hisyam Ramli, Siti Noor Linda Taib, Norazzlina M. Sa'don, Dayangku Salma Awang Ismail, Raudhah Ahmadi, Imtiyaz Akbar Najar, Rosmina Ahmad Bustami, Tarmiji Masron, Nazeri Abdul Rahman
Institution: Universiti Malaysia Sarawak
Language: English
Subfield: Management, Monitoring, Policy and Law
Field: Environmental Science
Domain: Physical Sciences
Open Access status: diamond
Funder: Universiti Malaysia Sarawak
OpenAlex ID: W7103999696
FWCI:
Cites:
Cited by:
Related to: