پديد آورندگان :
اميراحمدي، ابوالقاسم نويسنده دانشگاه تربيت معلم سبزوار Amir-Ahmadi, A. , شكاري بادي، علي نويسنده دانشجوي كارشناسي ارشد هيدروژيومورفولوژي در برنامه‌ريزي محيطي، دانشگاه حكيم سبزواري , , معتمدي راد، محمد نويسنده , , بينقي، مريم نويسنده كارشناس ارشد ژيومورفولوژي در برنامه‌ريزي محيطي، دانشگاه حكيم سبزواري ,
چكيده لاتين :
Introduction
Landslide is one of the natural phenomena that imposes many human and financial losses to human society every year. This phenomenon is really dangerous and is among the world’s seven natural disasters. There are several natural and human factors that is effective in the occurrence of this phenomenon and geomorphology as a practical science can investigate the causes and the factors causing this phenomenon in different regions. Identifying the susceptible areas for landslides and the mass movements are among the essential work in natural resource management and development planning. So that by recognizing these areas, we can contribute with human communities in natural resource management and development and construction planning. In this research, using Analytical network process (ANP), we practice on the landslide hazard zonation in the basin of Pivejan with an area about 31.2 square kilometers in the southern slopes of Binaloud, which is composed of limestone and neo-genes formations and the fluvial processes and flow are among the most important processes in this region that all in all these factors causes that this region being a susceptible zone.
Methodology
Due to the lack of access to the topographic maps of 1:25000, we used the 1:50000 maps that is related to Geographical Organization of Armed Forces and we extracted the accurate topographic information such as altitude, topographic slope, slope orientation, and distance form roads and river in this area. The next step was extracting the geological information including geological basin type and distance from the fault that for this purpose we used the 1:100000 maps that is again related to the Geographical Organization of Armed Forces. In addition, we used the land use and vegetation map with the scale of 1:50000 and also rainfall data from meteorological stations to provide information layers. For providing information relating to landslides, questionnaires that were prepared by the Office of Research and Evaluation of Watershed in Jihad Department in Khorasan Razavi province, and particularly the group of landslide investigation was used. In this study, at first we practice on ten effective parameters including altitude, rainfall, slope, slope orientation, geology, land use, vegetation density, distance from the river, faults, and roads and then using ANP, we tried to prepare a landslide hazard zonation map in the study area. The ANP is include four phases that are: 1. Creating the model and its configuration, 2. Pairwise comparisons of priority vectors, 3. Forming the super matrix and 4. Selecting the best option.
Results and discussion
Using the ten criteria considered in this study, the model structure was established in which elements of each cluster have interdependence with the elements of clusters and also have and external dependency. In the second stage, we performed a pairwise comparison and estimate the relative weight of layers and the highest weight of layer was awarded to that one which has the maximum impact on determining the goal and the numerical values from 1 to 9 for the degree of importance of criteria were considered. After this pairwise comparison, using the Super Dictation software, compliance rates of results were evaluated. In this study the amount of C.R was equal to 0.07181 that represent the compliance of study layers. Then, for calculating the final factor and coefficient, three super matrices were calculated: 1. Non weight super matrix, 2. Weight super matrix, 3. Comparative super matrix. At first, the weights that obtained by the early super matrix were entered that illustrates the interrelationships between elements of the system. In the next step, a super matrix was created that the total numbers of its columns was equal to one. In another step, the square amount of values was calculated that is for converging values in weight super matrix until all the elements become the same. And finally, the final weight of each criteria and sub-criteria was calculated and then the ten layers’ factors in occurrence of landslides in this study were prepared in GIS. In the next step, based on the weight of each layer that was resulted from the paired comparison, all data in GIS were converted to layers with raster format. At the end, the landslide zoning map with the raster format were produced through the operation of reconciliation of layers and to deliver the better results, the entire area was divided into five classes from very low to very high using the Natural Break approach.
Conclusion
The results of using this model in this study show that factors such as distance from river and road reach the highest weighted score (0.228, 0.204) and slope and slope orientation reach the minimum weighted score in this basin, respectively. So that, among the 31 landslide that were occurred in the study area, 17 landslides were occurred in the distance of 0-50 meters and 14 landslides in distance of 50-150 meters from the rivers that indicate the direct and important effect of erosion and river laundering. In addition, human activities such as road construction in the study area caused the increase of possibility of landslide several times so that 22 landslides were occurred in the distance of 0 to 500 meters from the roads and of course attacked the roads. On the other hands, factors such as slope and slope orientation had the less effect of landslide in the study basin. The results of this study and adaptation of landslide data and field observation (64.5 percent of slip points in very high or high risk zones, equal to 54.06 percent of basin area) also show the high efficiency of ANP model as a convenient and efficient model for predicting landslide.
