پهنه بندی خطر رانش زمین با استفاده از روش های همپوشانی وزنی و تحلیل شبکه مطالعه موردی: منطقه نصرآباداستان گلستان

یکی از راهکارهای مهم برای کاهش خسارت ‌های ناشی از وقوع زمین ‌لغزش ‌ها، شناسایی این مناطق به منظور رعایت فاصله مجاز از آنها است. به‌ این منظور باید نقشه پهنه ‌بندی خطر زمین ‌لغزش دقیقی برای این مناطق تهیه شود.  برای انجام این پژوهش، به ‌وسیله نقشه‌ های زمین ‌شناسی موجود و بررسی‌ های میدانی، نقشه پراکنش زمین ‌لغزش ‌ها در حوضه نصرآباد استان گلستان تهیه شد.  در تحقیق حاضر از ترکیب روش تحلیل شبکه‌ ای (anp)   و روش وزنی، پهنه ‌بندی مناطق مستعد زمین ‌لغزش و شناسایی عوامل موثر بر آن انجام گرفت.  بدین منظور از معیارهای اصلی شامل شیب، جهت شیب، انحنای زمین، ارتفاع، سنگ شناسی، خاک ‌شناسی، کاربری زمین، تراکم پوشش گیاهی، فاصله از جاده، فاصله از سکونتگاه ‌ها، تراکم رودخانه و آبراهه، بارش برای تعیین مناطق دارای پتانسیل لغزش استفاده گردید.  با مقایسه زوجی میان خوشه ‌ها و عناصر آن، وزن هریک از عناصر محاسبه شد.  سپس با اعمال وزن و ضرایب هر عنصر در لایه مربوط به خودش و همپوشانی لایه ‌ها، نقشه نهایی پتانسیل زمین لغزش تهیه گردید.  نتایج نشان داد که عامل توپوگرافی  (با وزن  0/662) بیشترین تاثیر، و عامل هیدروگرافی  (با وزن  0/009)  کمترین تاثیر را در میان سایر عوامل دارند.  نتایج مدل نشان داد حدود  50 درصد از منطقه مورد مطالعه دارای پتانسیل بالای زمین ‌لغزه است و بیشتر زمین ‌لغزش‌ ها در مناطق با دامنه ‌های پرشیب  یا واحدهای سنگی نفوذ‌ناپذیر نزدیک به سطح زمین رخداده است.  همچنین به کارگیری روش همپوشانی وزنی به همراه تحلیل شبکه توانسته است به عنوان ابزاری کارآمد در پهنه ‌بندی خطر زمین ‌لغزه مورد استفاده قرار گیرد و به خوبی قابلیت و توانایی مدل تحلیلی  پژوهش را به اثبات برساند.

Landslide hazard zoning using weighted overlay and Analytic Network Process Methods Case study: Nasrabad region in Golestan Province

Extended Abstract Introduction As a type of mass movement involving slow or rapid movement of soil, rock material or both on the lower hillsides, landslide is under the effect of gravity.Landslide is recognized as one of the most common geological disasters causing worldwide damages and casualties.Landslide susceptibility maps provide important and valuable information,including time scale of possible future landslides, which are usedfor predicting landslide hazards. Since predicting the time of landslide occurrence is beyond the capability of science and knowledge, identifying areas susceptible to landslide and ranking them can extensively restrict the damages caused by landslide. Therefore, it is essential to zone landslide risk and identify factors affecting it. Analytic Network Process(ANP) is aGISbased MultiCriteria Decision Analysis(GIS MCDA) method successfully applied to many decisionmaking systems. The present study seeks to evaluate landslide risk and achieve a zoning map for the subbasin under study using ANP and Weighted Overlaymethods.   Materials and Methods Based on the literature and using different experts’ viewpoint, criteria affecting landslide risk were identified and five major criteria including topography, land use and land cover, geology, hydrometry and infrastructure were selected. The selected criteria include the following subcriteria: slope, slope direction, curvature, elevation, lithology, soil type, land use, vegetation density, distance from roads, distance from habitat, river and drainage density and precipitation. The effective factor layers were standardized and a specific scale was defined for their units.Then, each layer was assigned a weight based on its role and importanceusing Analytic Network Process.Proposed to modify Analytic Hierarchical Process(AHP), this method (ANP) relies on the analyses of the human brain for complex and fuzzy problems.Network Analysis Process generally includes the following steps: determining indicators, criteria and options;classifying identified criteria into clusters and elements; determining the relationship between clusters, elements and options; performing pairwise comparisons between clusters, elements and options, and finally calculating the final weight of elements and options. UsingWeightedOverlaymethod, these elements were then integrated with their related coefficients and the final landslide risk map was obtained. Results and discussion Each criteria and subcriteria were weighted using Analytic Network Processmethod.Topographic and land cover criteria had the most and hydrographic criteria had the least impact on the landslide occurrence. According to the final map, most landslides have occurred in eastern and southern slopes at an altitude of 500 to 2,200 meters. Moreover, 17/31% of the study area was located in the very highrisk class and 33% in the high risk class (about half of the area has high potential of landslide). Previous landslide data were used to assess the landslide zoning map results. Results indicate that most landslides have occurred in the high risk class (about 35% of landslides) and only about 4% of landslides have occurred in the very low risk class.   Conclusion Landslide is one of the natural hazards causing serious harms and problems for human life. Identifying the factors affecting landslide and zoning its hazard is especially important for the identification of risky and susceptible areas.So, landslides were selected as one of the main topics of the study with the aim of controlling and managing its hazards.The ANP network analysis method was used to model and predict landslide risk in this research.Each criteria and subcriteria were weighted and overlapped to producethe map of relative landsliderisk.The lowest risk was observed in the northern parts of the region, and the highest landslide risk was observed in the northern hillsides with higher humidity.WeightedOverlaymethod and network analysis model were effective in predicting landslide susceptibility and producing landslide zoning map.