Landslide vulnerability and risk assessment: a guideline for critical infrastructure in Malaysia.

Abstract

Landslide vulnerability is a crucial element that connects hazard and risk for a specific element-atrisk. Currently, landslide vulnerability study in Malaysia is limited and attention is given to susceptibility and hazard assessments. Ideally, vulnerability assessment should address various aspects of element-at-risk including physical, social, economic, and environmental. In 2018, a guideline for landslide vulnerability and risk assessment for critical infrastructure in Malaysia was developed for the Construction Research Institute of Malaysia (CREAM). The guideline aimed at developing large-scale landslide vulnerability and risk assessment methods for local authorities as a level of basic and supporting information for land-use plan, landslide mitigation purposes, and risk assessment for any development of the critical infrastructure (CI) i.e. road, dam, building and electricity pylon. The aim of this study is to develop a simple methodology to support more detailed on-site landslide vulnerability and risk assessment. Using a case study from the Cameron Highlands District in northern Malaysia remotely sensed and field data were combined to create a detailed landslide inventory and element-at-risk mapping. Due to the limited landslide damage records, a vulnerability model was developed using the qualitative indicator-based method (IBM). The indicators and the corresponding sub-indicators are divided into four clusters i.e. 1) the susceptibility of element-at-risk (C), 2) surrounding environment (E), 3) intensity of landslide hazard (I), and 4) affected community (P). Suitable indicators and sub-indicators were selected and proposed based on a thorough literature review and a series of focus group discussions (FGD) with agencies involved with landslide hazard management in Malaysia. The FGD sessions also focused on experts assigning scores for each indicator and sub-indicator based on their relationship to the likelihood of landslide vulnerability. The final scores were then converted to final weighting values and a landslide vulnerability map was generated by combining the individual vulnerability cluster maps i.e. C, E, I and P. The resulting landslide vulnerability index was classified into five classes; very high, high, medium, low, and very low with a clear definition of the potential damage to CI and the community. Using a qualitative risk-matrix approach a landslide risk map was generated by combining the landslide hazard and vulnerability maps and was then validated against past landslide event in the Bukit Antarabangsa, Selangor, Malaysia. The results confirm good agreement between the derived vulnerability and risk maps and actual landslide damage in the area. The methodology proposed here is however strongly dependent on several key elements including, the quality of landslide hazard map, the landslide inventory map and the experience of the experts.