EXTENDED ABSTRACT

Introduction

Landslides are one of the most frequent natural hazards in the world, and knowing the assessment and risk zoning of landslides can be effective in managing, controlling, controlling and reducing the amount of damage. The main purpose of this research is to investigate and evaluate the risk of landslides in the Khanian watershed (located in the west of Mazandaran province and southeast of Tankabon city).In order to conduct this research, a multi-criteria decision making method of the network analysis process (ANP) was used. The information layers included lithology, distance from the fault, slope, direction of slope, height level, land use, distance from the road, precipitation, distance from the river and density of springs as effective factors in the occurrence of landslides.
The study area of Khanian is one of the three thousand sub-basins of Tonkabon, which is located in the west of Mazandaran province and the southeast of Tonkabon city (Figure 1). Its area is about 2918.44 hectares. The studied area of Khanian watershed is located in longitude 50 degrees 55 minutes 11 seconds to 50 degrees 56 minutes 16 seconds and latitude 36 degrees 33 minutes 12 seconds to 36 degrees 41 seconds. In terms of political divisions, this basin is located in the south of Tankabon city and is considered one of the sub-basins of Sehazar River. Almost the entire access route to its central and southern parts is located in the mountainous and high area, and in this sense, it is considered among the very high watersheds. In addition, the general direction of the Khanian watershed is south to north, and in the south of the basin, they form the Sehazar River in the course of its intersection with each other. Also, the highest elevations of the basin are located in its southern part. Its maximum height is 4794 meters, and its minimum height is 319 meters. The average annual rainfall of the region is about 800 mm and its climate type according to the Dumarten method is a cold humid climate. In terms of vegetation, it is part of the semi-dense forest areas and the species present in it include beech, comfrey and alder.

Materials and methods
In this research, in order to assess the risk of landslides in the studied watershed, an integrated approach based on field studies, expert opinions and spatial modeling was used. 10 key factors used including lithology, fault, slope, slope direction, topographic level, land use, road, precipitation, river and springs were considered. Digital maps of height with a resolution of 10 meters, slope, direction of slope, geology with a scale of 1:100000, land use, fault, river and slip points were prepared. Validation of these layers was done using field visits and satellite images. Field studies were conducted in two stages. In the first stage, landslide-prone points were identified by considering geological evidence, morphology, land use and visual evidence such as cracks, minor landslides and weak vegetation. In the second stage, at each suspicious point, more detailed information was collected, including rock type, geological structure, current land use, slope and slope direction, distance from the river, the presence of faults, the amount of weathering, sliding blocks, surface materials, and the distance from the watershed. . In the next step, the accuracy and validity of the collected information layers were carefully checked and evaluated.
Analytical Network Method (ANP) as an extension of Analytical Hierarchy Method (AHP), is able to model correlations and feedbacks between elements in a decision making process. Compared to AHP, this method takes into account all the internal effects of the effective components in decision-making and enters into the calculations. In the AHP model, only criteria are related to their superior criteria. This means that the criteria are dependent on the goals, the sub-criteria are dependent on the criteria and the options are dependent on the sub-criteria. But the dependence between the criteria itself is not considered. In contrast, the ANP model, in addition to the hierarchical dependencies of AHP, also includes side effects and feedback between elements. In this model, not only the clusters affect the elements and the elements affect the options, but the options affect the elements and even the elements themselves affect each other and other clusters.
Results and discussion
The results of the model showed that the ANP method is very accurate, the order of prioritization of the effective factors in it is reasonable and acceptable, in the meantime, according to the effective factors of landslides in the ANP method, the slope parameters with a weight of 0.29 Land use with a weight of 0.24 and lithology with a weight of 0.20 have the most importance in the development of landslides, because it provides the possibility of comparing sub-criteria.
Conclusion

Field studies and investigation of the effective factors in landslides show that the existence of land slope and geological formations and the change of land use and the construction of villas on the slopes due to the touristic nature of the studied area strongly cause the occurrence of landslides. It is possible, the destruction of pastures and loss of vegetation is effective in intensifying landslides in the researched basin. Statistical analysis shows that the highest density of landslides occurred on the side of the roads, especially in the distance class of 0 to 100 meters of the road. This issue shows the direct relationship between the construction of the road and the occurrence of landslides in this area. In the northern watersheds of the country, in order to exploit the forest, build a network of access roads for logging operations, and also build roads for the residents, many roads have been built.
The results of the model showed that the ANP method is very accurate, the order of prioritization of the effective factors in it is reasonable and acceptable, in the meantime, according to the effective factors of the landslide of the ANP method, the land slope parameters with a weight of 0.29, land use with a weight 0.24 and lithology with a weight of 0.20 have been assigned the most importance in the development of landslides, because it allows the comparison of sub-criteria. According to the field visits, it can be acknowledged that the use of the ANP method has a relatively high accuracy for the study of landslides.

.