Landslide Hazard Zonation in Sri Lanka: An Assessment of Manual and GIS Based Automated Procedure in Preparation of Geology Weight Map

Abstract

The term landslides comprise almost all varieties of mass movements on slopes including rock falls, topples and debris flows that involve little or no true sliding [Varnes, 1984]. Landslides occur when the critical combinations of many internal and external causative factors are met with a triggering event such as intense rainfall, earthquake shaking, volcanic eruption, rapid snow melt, rapid change of water level, storm waves or rapid erosion that causes a quick increase in shear stress or decrease in shear strength of the slope material. In many countries, slopes which stood safe for centuries are now frequented by landslides and hence socioeconomic losses due to its impact are growing. This is mainly due to the expansion of human activities into more susceptible hill slopes under the pressure of increasing population and associated demands for housing and infrastructure facilities. Though this has reduced sustainable development efforts today more than ever before especially in developing countries, practice has shown that adequate hazard mitigation is possible. Landslide hazard zonation involves one of the most complex analysis of interrelated terrain factors such as

Application of Bagging, Boosting and Stacking Ensemble and EasyEnsemble Methods for Landslide Susceptibility Mapping in the Three Gorges Reservoir Area of China

“…With the continuous development of instruments and methods to obtain spatial data, the quality and quantity of spatial data have improved. Data-driven models, such as support vector machines (SVMs) [ 6 ], RF [ 7 ], artificial neural networks (ANNs) [ 8 ], and weight-of-evidence [ 9 , 10 ] models, have been used to produce regional LSM. In the data-driven model category, machine learning models provide a better prediction effect and higher accuracy than other approaches, such as expert-opinion-based methods and analytical methods [ 11 ].…”

Spatial Distribution of Landslide Susceptibility in New Road Construction Mengwitani-Singaraja, Bali-Indonesia: Based on Geospatial Data

“…The Baturiti and Sukasada sub-districts are areas that have a high susceptibility for landslides of 15.62%. Based on the occurrence of landslides in the field, Sukasada District experienced landslides more often [27][28][29], especially in Gitgit Village and Wanagiri Village, as shown in Fig 4a . Landslide events in the area One of the Wanagiri road sections occurred on March 25, 2021 (Fig 4b) and February 21, 2020, on the Gitgit Village road section (Fig 4b). This area is an area that has steep slopes (25-40%) to very steep (>40%) with high scores ranging from 4 to 5.…”

Application of bagging, boosting and stacking ensemble and EasyEnsemble methods to landslide susceptibility mapping in the Three Gorges Reservoir area of China

“…To study the mapping obtain spatial data, the quality and quantity of spatial data have also been improved. Data-driven models 85 have been used in regional LSMs, including support vector machine (SVM) (Yao et al 2008;Pradhan 2013), RF (Catani et al 2013;Youssef et al 2016), artificial neural network (ANN) Gorsevski et al 2016), and weight-of-evidence (Jayathissa et al 2019;Hussin et al 2016) models. In the data-driven model category, machine learning models have a better prediction effect and higher accuracy than other approaches, such as expert opinion-based methods and analytic methods (Chowdhuri et al 90 2021;Pham et al 2016).…”