Landslide Susceptibility Mapping Using Random Forest Algorithm and Its Correlation With Land Use In Batu City, Jawa Timur

I. M. Riestu, H. Hidayat*

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

Abstract

Landslides is one of the most detrimental natural disasters because their occurrences are often destructive to natural and artificial structures on earth and reduce the quality of the surrounding environment. Prediction of the level of vulnerability to landslides in an area can be used to reduce losses caused such as material, loss of life and other environmental damage. BPBD Batu City stated that landslides are the most frequent disasters in Batu City. This research is expected to determine the level of vulnerability to landslides in Batu City, East Java and discover the correlation with the land use. Random Forest (RF) is an algorithm that can be used to predict landslide disasters. The disaster occurance data is then linked to the parameters that cause landslides such as slope, rainfall, soil type, lithology, land use, and soil movement susceptibility zones. The dataset is then divided into 70% training data and 30% testing data. The performance test results show that the RF algorithm can be applied to predict landslide-prone areas in Batu City. This can be seen in the results of the accuracy test which obtained a value of 0.8981 and an AUC value of 0.9327. The result shows that the high prone areas are in settlements, industry & public facilities while the low prone areas are in forest.

Original languageEnglish
Article number012017
JournalIOP Conference Series: Earth and Environmental Science
Volume1127
Issue number1
DOIs
Publication statusPublished - 2023
Event7th Geomatics International Conference, GEOICON 2022 - Virtual, Online
Duration: 26 Jul 2022 → …

Fingerprint

Dive into the research topics of 'Landslide Susceptibility Mapping Using Random Forest Algorithm and Its Correlation With Land Use In Batu City, Jawa Timur'. Together they form a unique fingerprint.

Cite this