Landslide risk zoning in the upstream basin of Yamchi Dam in Ardabil province, using multi-criteria decision making methods MARCOS and CODAS

Among the biggest disasters and natural disasters, landslides have taken the seventh rank in terms of casualties. Considering that, landslides are one of the most destructive natural disasters and cause severe changes in landscape morphology and damage to natural and artificial structures on earth (Tanias and Lombaro, 2019). Identifying landslide-prone areas and producing accurate maps of landslide susceptibility zoning are important issues for risk management studies and are vital to reduce landslide disasters (Kalksen et al., 2016:54; Rabi et al. colleagues, 1:2022).The upstream basin of Yamchi dam, with an area of 698 square kilometers, is located between the Sablan volcanic massif in the north and Bozgosh mountains in the southwest. The highest point of the basin is the peak of the Heram Mountain with a maximum height of 4505 meters above sea level and a minimum height of 1506 meters at the Yamchi Dam site. The climate of the region is semi-arid and very cold. In terms of tectonics, the studied area is located in the tectonic zone of Western Alborz-Azerbaijan. The different lithological units of the region are spread in volcanic and sedimentary forms.

The current research is of an applied type and its research method is an analysis based on the integration of data analysis, geographic information system, and the use of multi-criteria analysis techniques. the ENVI, Ecognition, Arc GIS, Idrisi, and Excel software were used for image processing and data analysis. To assess the risk of landslides, first, the effective factors (including slope, aspect, dem, lithology, soil, land use, rainfall, distance from communication road, distance from the river, and distance from fault), according to natural and human conditions The area was identified. In the next step, information layers related to each of the factors were prepared in the geographic information system environment. The weighting of the investigated factors was done according to the CRITIC method and the final analysis, using CODAS and MARCOS multi-criteria methods. After preparing the landslide sensitivity map, the accuracy of the models has been checked using the ROC curve.

According to the obtained results, respectively; Slope factors with a weight of 0.14, land use with a weight of 0.13, and lithology with a weight of 0.12 assigned the greatest role in the occurrence of landslides in the basin. Examining the high-risk and high-risk points introduced by the reviewed algorithms shows; In terms of the slope criteria, according to the output of the CODAS algorithm, the areas with high and very high-risk probability are located between 10-65% slopes. Examining the output of MARCOS and comparing it with the slope map also shows that the slope values of high-risk and very high-risk points are between 15-50%. According to the criteria of land use, according to the results of both methods, agricultural use, pastures, and man-made areas have the highest level of areas with a very high probability of danger. About the lithology map, it can be said that according to the zoning map resulting from the application of CODAS and MARCOS methods, very high-risk and high-risk classes, mainly; They are seen in formations with very low and medium resistance and limited form, in areas with resistant lithology.The results of the overlap of the output from the examined models, with the distribution of sliding points; showed that according to the CODAS multi-criteria decision-making algorithm, 22.58 and 54.84 percent of the slip points are in the high-risk and high-risk category, respectively, and according to the results of the MARCOS method, 51.62 and 29.03 % of the slip points are in the very dangerous category.

According to the output of the CODAS method, respectively; 139.46 and 58.17 square kilometers of the area of the basin and according to the results of applying the MARCOS method, 114.01 and 54.07 square kilometers of the area are in the high-risk and very high-risk categories. The investigations carried out in this analysis show that in the upstream basin of Yamchi Dam, due to its mountainous nature, cultivation in sloping lands, excessive grazing of livestock in pastures, presence of deep soils on Steep slopes, and water infiltration into the lower layers of the soil have created a suitable ground for the formation of the landslide phenomenon and will cause a lot of damage. The results of the ROC curve, showed that the accuracy of the method CODAS, with an area under the curve of 0.72, is very good, and the accuracy of the MARCOS method, with an area under the curve of 0.81, is excellent.