Landslides are one of the greatest environmental hazards that cause economic, financial, and life losses, along with the destruction of facilities and rising costs. Iran is considered as a high-risk country due to the favorable geographical conditions and the lack of comprehensive management and failure to observe environmental thresholds. So that is one of the 10 most hazardous countries in the world and every year landslides in most of the provinces cause economic damage to roads, railways, communications lines, irrigation and water supply channels, mineral installations, extraction facilities, oil and gas refinery, vital networks of cities, factories and industrial centers, artificial and natural dams and lakes, forests and pastures and natural resources, farms and residential areas and villages or threatening them. Therefore, identification of slopes instability and movements and landslides and their causing factors, and classification of villages in terms of vulnerability to landslide and determination of safe points for relief is essential. In order to reduce the damage caused by landslide, accurate understanding and full identification, the study of the aggravating factors, as well as the identification of susceptible areas and its prediction is necessary.
2. Materials and methods
Rudbar basin with an area of 1550 square kilometers is located in the southwest of Gilan province. The basin is one of Sefidrud sub-basin. In this area, due to the type of formations, the high depth of sediments in the slopes and the average rainfall, a lot of landslides occur. In this research, using the network analysis method, the area is zoned in terms of landslide occurrence and then the rural areas are zoned in terms of vulnerability to landslide hazard. In terms of type, this research is of descriptive-analytical study, and in terms of methodology is a descriptive-analytical study. The method is based on the analysis of criteria in the Decision Super software, and then overlapping the information layers in the ARCGIS software environment and integration of weighing models such as analytic network process and overlapping index. In this research, the factors affecting the occurrence of landslide were first identified and the network structure was created, then the weight of each factor was obtained and the final map of landslide risk zonation was obtained. Finally, the rural settlements were zoned in terms of vulnerability to landslide hazard.
3. Discussion and results
In this research, in order to zoning the risk of landslide, 14 effective factors including elevation, slope, slope direction, distance from the road, distance from fault, distance from the river, geology, soil type, climate, land use, rainfall, the topographic wetness index (TWI), length- slope index (LS), and stream power index (SPI) were used. In this research, in order to zoning the area in terms of the landslide hazard occurrence using the network analysis process, first, the weight of each factor was obtained in the Super decision software. Then the obtained weights were applied to the Arc GIS software and the final landslide zonation map was obtained. In the analytic network process, the incompatibility rate is calculated and presented by the Super Decisions software for each pairwise comparison matrix, that if it exceeds 0.1, then judgment is inconsistent and judgment should be reviewed. In this research, the rate of inconsistency was 0.06486 which is acceptable. In this study, geology, land use, distance from fault, distance from the waterways and slope were respectively the most important factors.
The present study was carried out in order to zoning the risk of landslide and also the zoning of vulnerability of villages against the landslide risk by analytic network process (ANP) in one of the Sefidroud sub-basins in Rudbar county. For this purpose, first, effective factors for landslide occurrence were identified, and the weight of each factor was determined in the Super decision software. Then the obtained weights in Arc GIS software were applied e and the final landslide zonation map was obtained. The evaluation of results showed that more than 50% of the study area has a medium to high risk. Of this amount, 14.57 percent are located in high and very high risk zones. Most of these areas are located in the southern parts of the basin and this is due to the existence of landslide-sensitive formations. A numerical value (0.914) obtained for the Roc curve also showed that the landslides in the study area has a strong relationship with the probability values derived from the analytic network process. According to the results, of total 188 villages, 49 villages (25.53%) are located in high and very high risk areas and this is a serious threat for the inhabitants of these areas.