University of Isfahan Geography and Environmental Planning 2008-5362 28 4 2018 02 20 Integration of Multi-criteria Decision Making Methods with Logistic Regression Approach to Assess Landslide-Prone Areas in Zilbir-Chai Basin Integration of Multi-criteria Decision Making Methods with Logistic Regression Approach to Assess Landslide-Prone Areas in Zilbir-Chai Basin 33 52 22451 10.22108/gep.2017.98297.0 FA Mohamad Reza Nikjo Assistant Professor of Geomorphology and RS Groups, University of Tabriz, Tabriz, Iran Mohamad Hosein Rezaei Moghadam 2 Professor of Geomorphology and RS Groups, University of Tabriz, Tabriz, Iran Soghra Andaryani Ph.D Candidate in Geomorphology, University of Tabriz and Master of RS & GIS, Tabriz, Iran Journal Article 2016 07 12 Identifying susceptible areas to massive movements, including landslides, through risk modeling with appropriate and efficient models, is one of the basic measures in reducing potential damage and managing risk. The purpose of this study is to combine multi-criteria decision making techniques with logistic regression method to investigate landslide-prone areas in Zilbir-Chai basin. One of the methods is Weighted Linear Combination (WLC) method. In this model, firstly, the variables that are effective in landslide occurrences are subdivided into sub-criteria and each specific criterion is given a certain weight. In order to prevent the great role of the experts’ taste, the landslides in the studied area were discarded with the sub-criteria of each variable, and the percentage of landslide occurrence in each of them was the criterion of weighting the sub-criteria. This way, Fuzzy standardization was done accordingly. In the WLC model, in addition to weighting the subcategories, the variables themselves are also generally weighted by the Analytical Hierarchy Process (AHP) method. In the present study, in addition to conventional weighting, this weight was measured by logistic regression (RL) statistical method and weighing criterion were obtained, normalized, and applied directly to the variables in the WLC model. To achieve this purpose, modeling was performed using nine independent parameters. Geological factors, land use type, altitude, slope gradient and aspect, distance from fault, distance from road and distance from drainage network were considered as environmental factors, and rainfall quantity (obtained from precipitation layer) was considered as a trigger factor. According to the results obtained from WLC model based on logistic regression weighting, high and very high risk zones constitute about 7.4% of the area with a validation of 94/0, but based on AHP weighting, these zones constitute about 4.7% of the area with a validation of 90/0. Identifying susceptible areas to massive movements, including landslides, through risk modeling with appropriate and efficient models, is one of the basic measures in reducing potential damage and managing risk. The purpose of this study is to combine multi-criteria decision making techniques with logistic regression method to investigate landslide-prone areas in Zilbir-Chai basin. One of the methods is Weighted Linear Combination (WLC) method. In this model, firstly, the variables that are effective in landslide occurrences are subdivided into sub-criteria and each specific criterion is given a certain weight. In order to prevent the great role of the experts’ taste, the landslides in the studied area were discarded with the sub-criteria of each variable, and the percentage of landslide occurrence in each of them was the criterion of weighting the sub-criteria. This way, Fuzzy standardization was done accordingly. In the WLC model, in addition to weighting the subcategories, the variables themselves are also generally weighted by the Analytical Hierarchy Process (AHP) method. In the present study, in addition to conventional weighting, this weight was measured by logistic regression (RL) statistical method and weighing criterion were obtained, normalized, and applied directly to the variables in the WLC model. To achieve this purpose, modeling was performed using nine independent parameters. Geological factors, land use type, altitude, slope gradient and aspect, distance from fault, distance from road and distance from drainage network were considered as environmental factors, and rainfall quantity (obtained from precipitation layer) was considered as a trigger factor. According to the results obtained from WLC model based on logistic regression weighting, high and very high risk zones constitute about 7.4% of the area with a validation of 94/0, but based on AHP weighting, these zones constitute about 4.7% of the area with a validation of 90/0. https://gep.ui.ac.ir/article_22451_715df34b7a67d7d53cf17b7dbeca35d6.pdf