Journal of Watershed Management Research

     The landslides impose serious damages to the economy, environment and human throughout the world. Identification of areas susceptible to landslides is necessary to avoid risks. In This research for Landslide hazard zonation in sarkhoon karoon watershed have been used Shannon’s Entropy Index and Information Value Methods. For this purpose, at first, landslide locations were identified using satellite images and field surveys and then landslide inventory map was created for study area. In the next step, 10 Effective Factors in Landslide occurrence include altitude, slope, aspect, distance from road, distance from fault, distance from river, lithology, land use, stream power index,  topography wetness index, Plane Curvature and Profile Curvature were identified and mentioned maps will be digitized in GIS.  In order to determine the weight of factors used Shannon’s Entropy Index and to determine the weight of classes used Information Value. The final Zonation map in the five classes include potential risk of very low, low, moderate, high and very high were prepared. The ROC (Receiver operating characteristic) curves and area under the curves (AUC) for landslide susceptibility map were constructed and the areas under curves were assessed for validation purpose and its value ​​showed that the hybrid model has a higher efficiency (0.781) for landslide hazard zonation. Results showed that land use and distance to road factors have the greatest impact on landslides. According to the results of landslide maps 14.45% (11220.4 ha) of the area are ranked as very dangerous areas and 6.11% (4744.1 ha) as dangerous areas.The results of this research can help planners to choose favorable locations for development schemes, such as infrastructural, buildings, road constructions, and environmental protection.
 

One of the goals of geomorphologists in working with the models of different landforms is to obtain better relations in realizing the physical realities of environment. In this study, to evaluate the performance of geomorphometric parameters to increase accuracy of zoning landslide susceptibility map has been studied. As the first step by the application of nine initial conditioning factors including slope, aspect, elevation, land use, lithology, distance from roads, rivers and vegetation index (NDVI) the zoning map was provided. In the next step geomorphometric parameters influential on the occurrence of landslide including topographic location index (TPI), surface curvature, curved sections, slope length (LS), Topographic wetness index (TWI), stream flow power (SPI), surface area ration index (SAR), was added to the model and then the zoning map was obtained. In the final step, the zoning maps was evaluated by using ROC curve. To provide zoning maps a new mixed model was applied, so, for determination of criteria weights multivariate regression and to determine weight of the classes' frequency ratio method was utilized. The findings of this research indicated that geomorphometric factors have a considerable influence on the increase of identification of regions that are susceptible to the landslides and enhance the accuracy of zoning maps from 0.731 to 0.938. These factors have also increased the resolution of the slip classes. According to the results, topography position index, plan curvature and surface area ratio have the highest influence on the accuracy of zoning maps. Based on superior approach, 8.68% (6737 ha) of the region are at very high risk and 15.3% (11906 ha) have been identified as high risk areas. According to the high ability of geomorphologic parameters in the identification of susceptible areas to the landslide, the application of these parameters is recommended in landslide hazard zonation.