Vol.:(0123456789) 1 3 Modeling Earth Systems and Environment https://doi.org/10.1007/s40808-019-00615-w ORIGINAL ARTICLE Landslide vulnerability assessment based on entropy method: a case study from Kegalle district, Sri Lanka E. N. C. Perera 1,2  · D. T. Jayawardana 3  · P. Jayasinghe 4  · Manjula Ranagalage 5,6 Received: 28 February 2019 / Accepted: 10 June 2019 © Springer Nature Switzerland AG 2019 Abstract The concept of landslide vulnerability to a given location is hard to quantify. Few studies have been carried to determine susceptibility using social and physical factors. This study is the first attempt in Sri Lanka to quantify level of vulnerability by integrating major physical and social indicators to map the spatial distribution of vulnerability. Considering the limitations of traditional weight evaluation method in calculation of the multiple indicators and ignorance of the associations among evaluating indicators, a new weight evaluation process, entropy method was introduced in this study. This improved method for determination of weight of the evaluating indicators was applied to estimate weight for the 14 selected indicators. The primary data were obtained from a comprehensive questioner survey (n = 402) of households or buildings (elements) with their coordinates based on a spatially balanced approach for ensuring spatial coverage of the entire landslide distribution. The spatial distribution of vulnerability was mapped using Kriging interpolation. According to the map, landslide vulnerabilities in the study area demonstrate notable regional specifications. Besides, the spatial distribution of vulnerability has shown a close relationship with rural and urban settlements. Results of spatial vulnerability reflect discrimination and inequalities in the development of the study area. According to landslide vulnerability analyses, 14.6% (247 km 2 ) of the entire area is found to be the highest vulnerable zone for a landslide and 39.8% (675 km 2 ) of area categorized under the lowest zone to vulnerability. Further, the study revealed a reasonable contribution by entropy method on analysis of social and physical indicators, which is useful for other vulnerability assessments. Keywords Landslides · Entropy method · Interpolation · Vulnerability Introduction Landslide is the most damaging and deadly natural haz- ards (Clague and Roberts 2012). During recent years, land- slides are responsible for over 26,000 deaths, homeless for 2.5 million and US$ 40 billion losses in total (CRED 2011). The effects of landslides can fluctuate according the magnitude of the hazard and the exposure of vulner- able community. The concentration of assets on exposed regions, socio-economic conditions, and land-use type all combine to make society vulnerable to landslide events (Papathoma-Köhle et al. 2007). There is a consensus that * E. N. C. Perera chinssu@gmail.com D. T. Jayawardana daham@sci.sjp.ac.lk P. Jayasinghe jpathmakr@gmail.com Manjula Ranagalage manjularanagalage@gmail.com 1 Institute of Human Resource Advancement (IHRA), University of Colombo, Colombo, Sri Lanka 2 Faculty of Graduate Studies, University of Sri Jayewardenepura, Nugegoda, Sri Lanka 3 Department of Forestry and Environmental Science, Faculty of Applied Science, University of Sri Jayewardenepura, Gangodawila, Nugegoda, Sri Lanka 4 Landslide Division, National Building Research Organization, Colombo, Sri Lanka 5 Faculty of Life and Environmental Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8572, Japan 6 Department of Environmental Management, Faculty of Social Sciences and Humanities, Rajarata University of Sri Lanka, Mihintale 50300, Sri Lanka