Assessment of coastal landslide sensitivity with geographical information systems and logistic regression method: Between Ayancik-Gerze, (Sinop)

Landslide, which is one of the natural disaster types, causes a lot of material damage and death, especially in the mountainous areas of the world. Underestimating the extent of the damage caused by landslides, which have a destructive effect, makes it necessary to carry out detailed studies on landslides. In this study, it was aimed to evaluate the coastline between Ayancık-Gerze Districts (Sinop) in terms of coastal landslide susceptibility using multivariate statistical methods. For this purpose, based on the coastal landslides that occurred previously in the study area, it was aimed to determine the regions with the possibility of coastal landslides and it was aimed to prepare a coastal landslide susceptibility map by using Logistic Regression Analysis, which is one of the multivariate statistical methods. In order to create a coastal landslide susceptibility map, parameters were determined based on previous studies. These parameters are; slope, aspect, height, lithology, topographic wetness index (TWI), profile curvature, plan curvature and normalized difference vegetation index (NDVI). In the study, the logistic regression method was performed and the regression constant and the regression coefficients of the parameters were obtained. The accuracy of the coastal landslide susceptibility map created by logistic regression analysis was evaluated by ROC analysis and the AUC value was 79%. As a result of this study, it was seen that the model was quite successful and significant. The sensitivity map produced by logistic regression analysis can predict areas at risk for possible future landslides and guide the studies on disaster management planning. Keywords: Remote Sensing, Geographic Information Systems, Coastal Landslide, Landslide Susceptibility, Logistic Regression Analysis