مقایسه روش های مختلف تصمیم گیری چند معیاره در اولویت بندی سیل خیزی زیرحوضه های آبخیز کشف رود (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
سیل یکی از رایج ترین خطرات طبیعی است که هم زندگی و هم دارایی های مردم را تحت تأثیر قرار می دهد. تغییرات آب و هوایی اخیر باعث افزایش فراوانی و شدت سیل شده است. در چنین شرایطی شناسایی مناطق بالقوه خطر سیل برای کاهش خسارات ناشی از سیل بسیار مهم است. ازاین رو، این مطالعه با استفاده از شاخص های مورفومتری و روش های تصمیم گیری چند معیاره به دنبال شناسایی مناطق آسیب پذیر سیلاب در زیرحوضه های حوضه آبخیز کشف رود است. بدین منظور 16 شاخص مورفومتری از DEM منطقه استخراج شد و سیل خیزی زیرحوضه های مختلف با استفاده از 5 روش تصمیم گیری چندمعیاره شامل AHP، ANP، VIKOR، TOPSIS و ELECTRE تعیین گردید و در پنج طبقه سیل خیزی خیلی زیاد، زیاد، متوسط، کم و خیلی کم طبقه بندی شد. نتایج روش های مختلف تصمیم گیری چندمعیاره با استفاده از همبستگی اسپیرمن و بررسی درصد تغییرات مورد ارزیابی قرار گرفت. درنهایت مشخص گردید روش ANP در تهیه نقشه سیل خیزی حوضه دارای دقت بیشتری است و اختلاف معنی داری بین روش ANP،AHP و VIKOR با نقشه سیل خیزی حوضه بر اساس داده های مشاهداتی وجود ندارد. در بررسی درصد تغییرات نیز مشخص شد کمترین درصد تغییرات مربوط به روش ANP است. بر اساس نتایج این روش 3/24 درصد حوضه در طبقه سیل خیزی خیلی زیاد و 7/25 درصد در طبقه زیاد قرار دارند و درمجموع بیش از 50 درصد حوضه در طبقه سیل خیزی زیاد و خیلی زیاد قرار می گیرند.Comparison of different multi-criteria decision-making methods in flood prioritization of Kashafrood sub-basins
Extended Abstract
Introduction
Flooding is one of the most destructive environmental phenomena, causing significant human and socioeconomic losses worldwide every year. Over the past two decades, the frequency of floods worldwide has increased by more than 40%. Between 1995 and 2015, approximately 109 million people were affected by floods annually, with damages amounting to 58 billion Euros (75 billion USD) annually (Khosravi et al., 2019). . In Iran, floods are the most dangerous natural phenomenon, causing significant human and financial losses in various parts of the country annually. Historically, floods have been recognized as the most frequent, deadly, and costly natural hazard (Hejazi, Andariani, Almaspour & Mokhtari Asl, 2015).
Multi-Criteria Decision-Making (MCDM) methods have been implemented in diverse areas of water management, including evaluating flood management options, optimizing resource allocation, and assessing flood risk potential. Compared to models based on physical watershed characteristics, MCDM methods require less detailed information and generally provide high prediction accuracy (Dou et al., 2015). Therefore, this study aims to assess the flood risk of the Kashafrood River sub-basins using morphometric indices and various MCDM methods.
Material and Methods
The study area is located between 9°E longitude and 35°38´ to 37°N latitude, covering an area of 15,650 square kilometers. This applied research integrates data analysis, geographic information systems (GIS), morphometric parameters, and multi-criteria decision analysis techniques. Data analysis was conducted using ArcGIS and Excel software.
Morphometric data were extracted and analyzed to enable pairwise comparisons for hierarchical and network analysis methods, calculating the relative importance and validity of parameters.
The TOPSIS method, introduced by Yun in 1980 and further developed by Yun and Wang in 1981, is widely used in multi-indicator decision-making. The VIKOR method, first introduced by Aperivik in 1979, focuses on selecting and ranking alternatives based on conflicting criteria (Mardani, Zavadskas, Govindan, Amat Senin & Jusoh, 2016). The ELECTRE method, developed by Bernard Roy in the 1960s, uses dominance relationships to compare alternatives.
Validation of the models was performed using percentage change analysis and recorded flood points in the region.
Results and Discussion
The criteria for this research were the morphometric characteristics of the 20 studied sub-basins, extracted using a Digital Elevation Model (DEM). These characteristics were categorized into three groups: linear, surface, and relief parameters.
Linear parameters included watercourse length and branching ratio. Basins with longer watercourses typically produce higher runoff and faster peak flow during rainstorms compared to basins with shorter watercourses (Batt, 2019). Surface parameters included basin area (A), basin length (Lb), basin perimeter (P), shape factor (Ff), drainage density (Dd), roundness ratio (Rc), elongation ratio (Re), texture ratio (Tr), waterway density (Fs), shape index (Bs), and holding constant (C). Relief parameters included basin elevation (Bh), elevation ratio (Rr), and elevation number (Rn). Basin elevation (Bh) indicates slope, waterway gradient, and runoff discharge.
The prominence value of the studied sub-basins ranged from 2247 to 1049 meters, with the highest prominence observed in sub-basins K04, K09, and K13, and the lowest in sub-basins K05, K12, K14, and K17. Among the criteria, basin area (weight: 0.188) was the most significant, while basin prominence (weight: 0.008) was the least significant. The compatibility ratio of the parameters was 0.05, indicating valid comparisons since the value is below the threshold of 0.1.
The percentage change analysis revealed that the ANP method exhibited the lowest variability (25%), while the AHP and TOPSIS methods showed the highest variability. A comparison of observed flood points with model results indicated that the ANP method had the highest correlation with flood points, followed by the AHP method, while the ELECTRE method exhibited the lowest correlation.
Conclusion
To prepare a flood zoning map for the studied sub-basins, multi-criteria decision-making (MCDM) methods such as AHP, ANP, TOPSIS, VIKOR, and ELECTRE were employed. Morphometric criteria were utilized to evaluate the performance of these methods.
The sub-basins were classified into five flood risk classes by each method. According to the AHP method, 53.6% of the area was classified as high and very high flood risk. The ANP method classified 50% of the area as high and very high flood risk. The TOPSIS method classified 47.6% of the area as high and very high flood risk. The VIKOR method identified 61.13% of the area as high and very high flood risk, while the ELECTRE method classified 50.75% of the area in this category.
Validation using flood points showed that the ANP method had the highest accuracy. The percentage change analysis also confirmed that the ANP method exhibited the lowest variability.
These findings suggest that morphometric indices and MCDM methods are effective tools for assessing flood risk in watersheds. However, flood risk mapping is an iterative process requiring periodic updates due to changes in land use, weather patterns, and other factors influencing flood dynamics. Regular reassessment and updating of flood risk maps are essential for accurate and effective flood management.
تبلیغات
