Landslide Vulnerability Assessment in East Java, Indonesia, using Fuzzy Analytical Hierarchy Process - Natural Breaks
Arna Fariza(a*), Arif Basofi(a), Abier Rahma Sofyantie(a)

a) Department of Informatics and Computer Engineering, Politeknik Elektronika Negeri Surabaya
Jl. Raya ITS Sukolilo Kampus PENS
*arna[at]pens.ac.id

Abstract

Landslides pose a significant threat in flood-prone regions like East Java, Indonesia, causing fatalities, disruption, economic losses, and environmental damage. Traditional assessment methods may not adequately capture the complexities and uncertainties inherent in landslide occurrence for early warning into people. This study explores the potential of the Fuzzy Analytical Hierarchy Process (FAHP) combined with Natural Breaks classification to map landslide vulnerability in East Java resulting in low, moderate, and high landslide vulnerability indices in 38 cities/districts using 6 criteria: hazard, social vulnerability, economic vulnerability, physical vulnerability, environmental vulnerability, and capacity index. The weights obtained from FAHP will be classified using the natural breaks algorithm. the FAHP method produced a GVF of 0.9800, with a lower Standard Deviation of Assessment Metrics (SDAM) of 0.0059, indicating greater consistency across districts. In contrast, the AHP method achieved a slightly higher GVF of 0.9922 but had a higher SDAM of 0.0155, reflecting increased variability in its assessments. The importance of utilizing robust methodologies of FAHP and AHP for effective landslide vulnerability assessment, ultimately contributing to better disaster preparedness and risk mitigation strategies in the region.

Keywords: Landslide- vulnerability- Fuzzy analytical hierarchy process- natural breaks- spatial mapping

Topic: Artificial Intelligence (AI)