Document Type : Research Article
MSc. Student, Department of Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, Noor, Iran
Assistant Professor, Department of Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, Noor, Iran
PhD in Tectonic-Geology, Geology Expert in TaHa Consulting Engineers
Providing effective solutions to prevent and reduce the damage caused by landslides is inevitable. Predicting and zoning landslides are among these solutions. Accordingly, the present study was conducted to compare and assess the accuracy of Analytical Hierarchy Process (AHP) and Multivariate Regression (MR) methods in landslide hazard zoning in Valiasr Watershed with an area of 198 km-2 located in western Ardebil Province. Six factors including aspect, slope, elevation, lithology, land use and distance from the river were known as the most effective factors in landslide occurrence in the study area. Landslide zones were then obtained in five classes using two AHP and MR methods. Finally, landslide zoning maps were compared and evaluated by Density ratio index (Dr) and Quality sum index (Qs) to assess the accuracy of two studied methods. The results showed that distance from river, slope, land use, lithology and height were weighting in 0.426, 0.173, 0.145, 0.134, 0.089, and 0.033 in the AHP methods and 0.531, 0.109, 0.344, 0.273, 0.123 and 0.061 in the MR method, respectively. The amount of two Dr and Qs indices were calculated to be 5.51 and 0.44 respectively in the AHP method and 6.45 and 0.72 respectively in MR method which indicated that MR method having 28% disagreement with reality was more accurate than AHP method having 56% disagreement with reality for landslide hazard in the study area.