LANDSLIDE SUSCEPTIBILITY ASSESSMENT BASED ON SPATIAL DATA ANALYSIS TECHNIQUES: A CASE STUDY FROM NORTHEAST OHIO

A quantitative method for preparation of landslide susceptibility map is assessed in a study area of Northeastern Ohio. This study uses susceptibility models in Geographic Information Systems (GIS) based on statistical relationships between the landslides and the controlling factors. The landslide locations in the area were identified from the aerial photographs, field checks, exiting literature, and a landslide inventory map was prepared for the region at a scale of 1:24,000. The occurrence of landslides in a given area generally depend upon the complex interaction of different dependent and independent factors like slope angle, slope aspect, soil type, erodable soil, depth to groundwater, landcover pattern, distance from the river, etc. These factors were imported as raster data layers in ArcGIS for the landslide susceptibility analyses of the study area. Each of the above-listed factors was classified and coded using a numerical scale corresponding to the physical conditions of the region. In order to investigate the role of each factor in controlling the spatial distribution of landslides, landslide susceptibility index model, and logistic regression model were generated using the digital dataset. Each model was superimposed on the landslide inventory map and was evaluated for its suitability. The logistic regression model was found to be the best model for predicting the landslide susceptibility of the area. The results indicate that the factors such as slope angle, distance from the river, and the erodable soil are statistically significant in controlling the slope movement, whereas slope aspect, soil type, landcover, and depth to watertable are not so important and thus excluded from the model. The data from this model was used in ArcGIS to produce a landslide susceptibility map. The landslide susceptibility was classified into three categories: low, moderate and high. The results of the study demonstrate that landslide susceptibility of a region can be effectively modeled using GIS technology and logistic regression analysis.