General Department of Natural Resources and Watershed Management, Mazandaran, Sari, Iran
Subject Areas : جنگلها و منابع طبیعی
1 - MSc in Water Engineering - Water Structures, General Directorate of Natural Resources and Watershed Management of Mazandaran, Sari, Iran
Keywords: Slope, Distance from road, Jackknife, Maximum Entropy, Validation ,
Abstract :
The aim of the present study is to investigate the most important factors affecting landslides in Behshahr-Galogah coastal watershed. For this purpose, by field survey in the study area, the locations of landslides were recorded with GPS. 7 factors including, distance from the road, distance from the fault, geology, distance from the river, the average of annual rainfall, and land use changes as the most important influencing factors and independent variables were prepared using Geographical Information System (GIS) and entered into the model. For modeling and calculations, the maximum entropy model was used in the Maxent software environment. The results of the Jackknife diagram showed that the three factors of slope, distance from the road and geology have the greatest effect on landslides in the study area. As, by increasing the slope (20-40 percent), landslides increase; also, landslides have an inverse relationship with the distance from the road. For the geological factor, there is the highest probability of landslides within PZq.d and Qra units. The validation of the model using the area under the ROC curve showed that this model is acceptable in the study area with a value of 0.77. On the other hand, the results of landslide risk classification in the studied area showed that 40% of the area has landslide potential. The findings of this study provide an important foundation to assist decision makers in disaster prevention and mitigation and also, it helps to better understand the dangers of landslides which is very useful for the development of risk management measures, so it is suggested to take action in different regions in order to identify high-risk areas.