انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Assessment and Comparison of Random Forest and Support Vector Machine Models in Landslide Hazard Mapping: A Case Study of the Firoozabad Watershed ارزیابی و مقایسه مدل‌های جنگل تصادفی و ماشین‌بردار پشتیبان در پهنه‌بندی خطر وقوع زمین‌لغزش (مطالعه موردی: حوضه آبخیز فیروزآباد) 232801 10.22034/gmpj.2025.543572.1578 FA موسی عابدینی استاد، گروه جغرافیای طبیعی(گرایش ژئومورفولوژی) دانشگاه محقق اردبیلی 0000-0002-4243-4670 امیر حسام پاسبان دانشجوی دکترای ژئومورفولوژی، گروه جغرافیای طبیعی، دانشکده علوم اجتماعی، دانشگاه محقق اردبیلی، اردبیل، ایران Journal Article 2025 08 27 Introduction Landslides are among the most destructive geomorphological hazards, causing extensive damage to human lives, infrastructure, and natural environments, particularly in mountainous regions. Triggered by factors such as intense rainfall, seismic activity, geological conditions, and human interventions, landslides have become increasingly frequent due to urbanization, deforestation, and climate change. In Iran, where mountainous terrains and tectonic activity dominate, landslides pose a persistent challenge, inflicting significant economic and social losses annually. Mapping landslide susceptibility is therefore essential for risk reduction, sustainable land-use planning, and hazard management. Recently, machine learning algorithms such as Random Forest (RF) and Support Vector Machine (SVM) have been widely adopted for landslide susceptibility mapping (LSM), owing to their ability to capture nonlinear and complex relationships among conditioning factors. Additionally, the use of Google Earth Engine (GEE) provides an efficient cloud-based platform for processing multi-source geospatial data, enhancing model scalability and accuracy. The Firouzabad watershed is located in the south of Ardabil province, covering an area of 1599 square kilometers of the province. This basin is located at a geographical location of 48 degrees 14 minutes to 48 degrees 32 minutes east longitude and 37 degrees 25 minutes to 37 degrees 54 minutes north latitude (Figure 1). The Firouzabad watershed is bounded by the Gharesu watershed from the north, the Qarnqu River watershed from the west, the Talesh Mountains from the east, and the foothills of the Qezel Ozan River from the south (Abadi et al., 2013). The Firuzabad watershed in Khalkhal County, as one of the landslide-prone areas in Ardabil Province, has been facing an increase in the occurrence of this phenomenon in recent years, and the identification and zoning of high-risk areas in this area is of great importance. Given the complexity of the factors affecting landslides and the nonlinear interactions between them, the use of classical methods alone in identifying high-risk areas cannot meet today's needs. In this regard, machine learning algorithms such as random forest (RF) and support vector machine (SVM) have attracted the attention of researchers as new and efficient tools in landslide hazard modeling due to their high accuracy and ability to analyze large and diverse data. Also, the use of Google Earth Engine (GEE) as a cloud computing platform for analyzing remote sensing data provides the basis for fast, accurate, and spatially based analyses and enables better modeling at different scales. Methodology This study was conducted in the Firoozabad watershed, Ardabil Province, northwestern Iran, which is highly prone to landslides due to its rugged topography, lithological diversity, heavy precipitation, and human-induced land-use changes. A landslide inventory containing 300 confirmed events was compiled from field surveys, Google Earth image interpretation, and official reports. Twelve conditioning factors were selected: slope, aspect, elevation, lithology, land use/land cover, NDVI, topographic wetness index (TWI), curvature, distance to roads, distance to rivers, distance to faults, and precipitation. These were derived from DEMs, remote sensing products, and climatic datasets. The dataset was divided into training (70%) and validation (30%) subsets. The RF and SVM algorithms were implemented using GEE, supplemented with statistical analyses in R. Both models generated susceptibility maps, which were classified into five categories (very low, low, moderate, high, and very high susceptibility). Model performance was assessed using the area under the ROC curve (AUC) and classification accuracy. Results and Discussion Both RF and SVM successfully delineated landslide-prone zones, though their predictive accuracies differed. The RF model classified approximately 1059.72 km² (66.31%) as very high risk and 132.32 km² (8.28%) as high risk. The SVM model similarly identified 1059.89 km² (66.32%) as very high and 133.53 km² (8.35%) as high risk. While spatial patterns were comparable, the models differed in factor importance. RF identified slope (14%) and elevation (11%) as the most influential variables, whereas SVM emphasized slope (34%) and precipitation (10%). Validation metrics confirmed the superiority of RF. The AUC value for RF was 0.895, compared to 0.798 for SVM, indicating higher predictive reliability. This aligns with prior research suggesting RF’s robustness in handling nonlinear relationships, multicollinearity, and overfitting, making it more effective than SVM in landslide hazard modeling. Conclusion The study highlights the effectiveness of machine learning algorithms in landslide susceptibility mapping, with RF outperforming SVM in predictive accuracy for the Firoozabad watershed. Scientifically, these findings confirm that advanced algorithms can capture the complex interactions among environmental factors influencing landslide occurrence. Practically, the susceptibility maps produced can guide policymakers and land managers in prioritizing preventive measures, including restricting construction in high-risk zones, slope stabilization, improved drainage systems, and vegetation reinforcement. The study also acknowledges limitations, including reliance on static datasets for precipitation and land use, as well as uncertainties introduced by IDW interpolation of rainfall. Future studies should integrate high-resolution, time-series datasets and explore ensemble or deep learning methods to enhance dynamic susceptibility modeling. Overall, this research demonstrates that combining multi-source geospatial data, machine learning, and cloud-based platforms provides a powerful framework for landslide risk assessment, supporting sustainable watershed management and disaster risk reduction in Iran and similar regions worldwide. پژوهش حاضر با هدف پهنه‌بندی خطر وقوع زمین‌لغزش و مقایسه کارایی دو الگوریتم قدرتمند یادگیری ماشین، در حوضه آبخیز فیروزآباد استان اردبیل انجام شد. بدین منظور، دوازده عامل مؤثر در رخداد زمین‌لغزش منطقه مورد مطالعه شامل شیب، جهت شیب، ارتفاع، زمین‌شناسی، کاربری اراضی، شاخص پوشش گیاهی نرمال‌شده (NDVI)، شاخص رطوبت توپوگرافی (TWI)، انحنای دامنه، فاصله از جاده، فاصله از آبراهه، فاصله از گسل و بارش انتخاب و نقشه‌های مربوطه با استفاده از داده‌های سنجش‌ازدور، مدل رقومی ارتفاعی (DEM) و اطلاعات زمین‌شناسی و اقلیمی تهیه گردید. نقشه پراکنش 300 رخداد زمین‌لغزش موجود نیز به‌عنوان متغیر وابسته با بهره‌گیری از داده‌های میدانی، تصاویرGoogle Earth و گزارش‌های منابع طبیعی استخراج و به دو بخش آموزشی (70 درصد) و اعتبارسنجی (30 درصد) تقسیم شد. مدل‌سازی با استفاده از سامانه گوگل ارث انجین و تحلیل آماری در نرم‌افزار R انجام گرفت. نتایج حاصل نشان داد که در نقشه پهنه‌بندی مدلRF، حدود 72/1059 کیلومتر مربع (66 درصد) از مساحت حوضه در کلاس خطر بسیار زیاد و 32/132 کیلومتر مربع (8 درصد) در کلاس زیاد قرار دارد، در حالی که مدل SVM به‌ترتیب 89/1059 کیلومتر مربع (66 درصد) و 53/133 کیلومتر مربع (8 درصد) را در همین طبقات نشان داد. بررسی اهمیت متغیرها بیانگر آن بود که در مدل RF عامل شیب (14 درصد) و ارتفاع (11 درصد) بیش‌ترین نقش را در وقوع لغزش دارند، در حالی‌که در مدل SVMشیب (34 درصد) و بارش (10 درصد) اهمیت بیش‌تری داشته‌اند. ارزیابی نهایی با شاخص ROC و مقدار AUC نشان داد که مدل RF با مقدار 895/0 نسبت به مدل SVM با مقدار 798/0 از دقت بالاتری در پیش‌بینی مناطق مستعد لغزش برخوردار است. در مجموع، یافته‌ها نشان‌دهنده عملکرد مطلوب هر دو مدل در پهنه‌بندی خطر زمین‌لغزش است و تأیید می‌کند که به‌کارگیری الگوریتم‌های یادگیری ماشین در محیط سامانه GEE، ابزار مؤثری برای مدیریت ریسک، می باشد.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Evaluating the accuracy of machine learning algorithms in landslide hazard zoning in the Samian watershed ارزیابی دقت الگوریتم‌های یادگیری ماشین در پهنه‌بندی خطر زمین‌لغزش حوضه آبخیز سامیان ‏ 234416 10.22034/gmpj.2025.551301.1583 FA صیاد اصغری استاد ژئومورفولوژی، گروه جغرافیای طبیعی، دانشکده علوم اجتماعی، دانشگاه محقق اردبیلی، ایران 0000-0002-5015-904X مهدی فروتن دانشجوی دکتری آب و هواشناسی، گروه جغرافیای طبیعی، دانشکده علوم اجتماعی، دانشگاه محقق اردبیلی، اردبیل، ‏ایران 0009-0002-9979-6944 Journal Article 2025 10 05 Introduction Landslides are important natural hazards in mountainous areas that cause extensive human, financial and environmental losses by suddenly changing the land surface. This phenomenon destroys vegetation, agricultural lands and infrastructure and increases secondary destruction by intensifying erosion. Between 1995 and 2014, more than 3,876 landslides occurred in the world, resulting in thousands of injuries and more than 163,000 deaths. Identifying prone areas and zoning landslide susceptibility are effective solutions in reducing damage and managing the risk of this phenomenon. In this context, statistical methods and machine learning algorithms such as random forest and support vector machine can predict the occurrence of landslides with high accuracy. The aim of the present study is to zone landslide risk in the Samian watershed using machine learning algorithms and compare their performance in order to provide an accurate and reliable model for planning and reducing potential damages. Methodology This study aimed to map landslide risk in the Samian watershed using machine learning algorithms in the Google Earth Engine environment. First, topographic data (elevation, slope, direction, and curvature), geology (rock type), distance from river, road, and fault, and climatic and environmental variables (rainfall, soil moisture, NDVI, and land use) were collected and processed in a GIS environment. Field data included 55 landslide points and 55 non-slip points that were used to train and test the models. The layers were standardized using the fuzzification method to numerically represent the relative impact of each factor on slope instability. Three machine learning algorithms, including random forest (RF), support vector machine (SVM), and nearest neighbor (KNN), were implemented to generate a landslide probability map. Finally, the output of the models was converted into a hazard zoning map and categorized into five classes from very low to very high. The accuracy of the models was evaluated with AUC and comparison with real landslide data to determine the most accurate algorithm for the Samian basin. Results and Discussion The study of the distribution map of landslide and non-slip points in the Samian watershed showed that out of a total of 110 field-collected points, half were related to real landslides and were mainly concentrated in the western half of the basin, especially the slopes of the Sabalan heights and Nir county. This spatial distribution confirms the direct role of slope and high altitude in the occurrence of landslides. The analysis of the significance of the variables also showed that the slope factor with a share of 16.10% has the greatest impact on the occurrence of landslides, followed by altitude and rock type, which indicates the importance of morphometric features in controlling slope stability. At a spatial scale, the steep and high areas of the northwest of the basin (Sabalan heights) with slopes between 15 and 68 degrees and the southeastern part of the basin (Hire) have the highest potential for landslides due to the concentration of faults, high soil moisture, and abundant rainfall. In the southern part of the basin, especially in the Kuzeh Topraghi area, the northern slopes are more humid due to being in the shade of the sun, and long periods of frost and repeated cycles of freezing and thawing reduce the cohesion of the materials and increase instability. In the southwest of the basin (Nir) and along the Balikhlochay and Imamchay rivers, human activities such as road excavation and land use change have weakened the natural balance of the slopes and increased the risk of instability. In the northeast of the basin (Namin), despite dense vegetation, steep slopes and soil saturation during the rainy season have caused surface landslides. In contrast, the central areas of the basin with a tendency to the east (from Ardabil city to Abibiglou) are the least risky areas, which have higher stability due to their gentle slope, lower altitude, and less precipitation, and are in the low to very low risk class in all models. Conclusion A comparison of the performance of three machine learning algorithms (RF, SVM, and KNN) showed that all three are capable of predicting landslide-prone areas, but they have differences in sensitivity and accuracy. The accuracy index (AUC) for SVM, RF, and KNN was calculated to be 0.906, 0.864, and 0.848, respectively, indicating that SVM is more accurate in distinguishing landslide-prone and stable areas. Zoning results also showed that the total area of high and very high risk areas for RF, SVM, and KNN was 2222.1, 1680.79, and 1890.39 square kilometers, respectively, which is equivalent to 53, 40, and 45 percent of the total area. These values indicate that the RF algorithm takes a more conservative view and identifies more areas as high-risk, while the SVM provides a more balanced approach but with higher accuracy. پهنه‌بندی خطر زمین‌لغزش از ابزارهای کلیدی در مدیریت ریسک و برنامه‌ریزی پایدار منابع طبیعی به‌شمار می‌رود. این پژوهش با هدف ارزیابی دقت الگوریتم‌های یادگیری ماشین در پهنه‌بندی خطر زمین‌لغزش حوضه آبخیز سامیان، در محیط گوگل ارث انجین انجام شد. در گام نخست، مجموعه‌ای از متغیرهای مؤثر بر پایداری دامنه‌ها شامل عوامل توپوگرافی (ارتفاع، شیب، جهت و انحنای زمین)، ویژگی‌های زمین‌شناسی (جنس سنگ)، فاصله از رودخانه، جاده و گسل، و نیز پارامترهای اقلیمی و زیست‌محیطی (بارش، رطوبت خاک، شاخص پوشش گیاهی NDVI و کاربری اراضی) گردآوری و در محیط GIS فازی‌سازی شدند. برای آموزش و ارزیابی مدل‌ها، از ۱۱۰ نقطه نمونه شامل مناطق لغزشی و غیرلغزشی استفاده شد که ۷۰ درصد آن‌ها به آموزش و ۳۰ درصد به آزمون مدل‌ها اختصاص یافت. سه الگوریتم جنگل تصادفی، ماشین بردار پشتیبان و نزدیک‌ترین همسایگان به‌منظور تولید نقشه احتمال وقوع لغزش به‌کار گرفته شدند. نقشه‌های حاصل با داده‌های واقعی مقایسه و شاخص دقت مساحت زیر منحنی ROC(AUC) برای هر مدل محاسبه شد. نتایج نشان داد که عامل شیب با سهم تأثیر ۱۶٫۱۰ درصد، مهم‌ترین متغیر در بروز زمین‌لغزش است. از نظر مکانی، نواحی مرکزی با گرایش شرقی کمترین خطر لغزش و نواحی پیرامونی حوضه بیشترین خطر را نشان دادند. تمرکز بالای خطر در این نواحی عمدتاً ناشی از ارتفاع و شیب زیاد، رطوبت بالای خاک، بارش فراوان، هم‌پوشانی مسیر گسل‌ها با جاده‌ها و رودخانه‌ها و گسترش اراضی کشاورزی و بایر در دامنه‌های سبلان است. بر اساس نتایج پهنه‌بندی، به‌ترتیب ۵۳، ۴۰ و ۴۵ درصد از کل منطقه توسط مدل‌های SVM، RF و KNN در طبقات خطر زیاد و بسیار زیاد قرار گرفتند. مقدار شاخص AUC برای مدل‌های مذکور به‌ترتیب ۰٫۹۰۶، ۰٫۸۶۴ و ۰٫۸۴۸ به‌دست آمد که نشان‌دهنده دقت بالاتر مدل SVM در تفکیک نواحی لغزشی و پایدار است.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Investigating the Diversity of Landforms in the Geomorphological Units of Northeast Eqlid County with an Emphasis on Spatial Neighborhood بررسی تنوع لندفرم‌ها در واحدهای ژئومورفولوژی شمال شرقی شهرستان اقلید با تأکید بر همسایگی مکانی 252511 10.22034/gmpj.2026.583396.1613 FA سعید نگهبان گروه جغرافیا، دانشکده اقتصاد، مدیریت و علوم اجتماعی، دانشگاه شیراز، شیراز، ایران. مرضیه مکرم گروه جغرافیا، دانشکده اقتصاد، مدیریت و علوم اجتماعی، دانشگاه شیراز، شیراز، ایران. حجت اله مشایخ گروه سنجش از دور و GIS، دانشکده سنجش از دور و لیزر، دانشگاه شهید بهشتی، تهران، ایران. سارا قاسمی گروه جغرافیا، دانشکده اقتصاد، مدیریت و علوم اجتماعی، دانشگاه شیراز، شیراز، ایران. Journal Article 2026 05 26 Extended Abstract Introduction Landforms, as fundamental units of geomorphology, result from complex interactions between tectonic, climatic, erosional, and hydrological processes. Their identification and classification are crucial for understanding environmental changes, soil science, land use planning, and natural resource management. Traditional landform mapping methods, based on visual interpretation of topographic maps and expert judgment, are time-consuming, expensive, and lack reproducibility. Consequently, automated and semi-automated methods based on Digital Elevation Models (DEM) have gained prominence. Among various morphometric indices, the Topographic Position Index (TPI) has received special attention due to its conceptual simplicity and high efficiency in distinguishing low features (valleys) from high features (ridges and peaks). However, TPI performance is highly dependent on two key parameters: scale (moving window size) and neighborhood type (neighborhood geometry) . Scale determines whether a landform is identified at a local or regional level, while neighborhood type determines which surrounding cells influence the mean elevation calculation. Despite their importance, many previous Iranian studies have used default window sizes and neighborhood types without optimization. Therefore, the primary objective of this study is to classify landforms in northeastern Eqlid using TPI and determine the optimal neighborhood type (circular, rectangular, annular, and triangular) and optimal scale to produce the most accurate landform map. The study area, covering 51.27 km², is located in Fars Province, between longitudes 52°37' to 52°44' E and latitudes 30°45' to 30°51' N. Methodology This study utilized an ASTER DEM with 30-meter spatial resolution (2012). All analyses were performed in ArcGIS software. The Topographic Position Index (TPI) was calculated as \(TPI = Z_0 - \frac{1}{n}\sum_{i=1}^{n} Z_i\), where \(Z_0\) is the elevation of the central point, \(Z_i\) is the elevation of neighboring points, and \(n\) is the number of neighbors. Positive TPI values indicate that the central point is higher than its surroundings (ridges, peaks), negative values indicate lower positions (valleys, streams), and near-zero values indicate flat areas or constant slopes. To determine the optimal scale, square moving windows ranging from 3×3 to 45×45 pixels were evaluated. The optimization criterion was the minimum Root Mean Square Error (RMSE) derived from polynomial distribution fitting. Subsequently, four neighborhood geometries were compared: circular (all cells within a specified radius), annular (less weight to distant cells), triangular (directional pattern), and rectangular (full and symmetric coverage). After calculating TPI and classifying landforms according to the standard Weiss (2001) classification table, the extracted landforms were assigned to geomorphological units: mountain, pediment, and playa. These units were defined based on slope: slopes >25° (mountain), slopes between 1° and 25° (pediment), and slopes <1° (playa). Results and discussion The results are significant in three aspects. First, the strong scale dependency of TPI was confirmed. The TPI range at the 45×45 scale (range ≈452 units) was more than double that at the 3×3 scale (range ≈197 units), meaning a single point could be classified as a "local peak" at a small scale but as part of a "mountain slope" at a larger scale. This aligns with theoretical geomorphometry and studies by Weiss (2001) and Wilson & Gallant (2000). Second, and most importantly, the rectangular neighborhood was decisively superior. This superiority stems from its full and symmetric coverage of surrounding cells, enabling more accurate modeling of elevation changes in complex, heterogeneous topography. In contrast, annular and circular neighborhoods tend to smooth out intermediate and small landforms, while the triangular neighborhood's inherent directionality makes it unsuitable for regions with random, multi-directional topography. Third, the challenging finding of mountain peaks within the pediment unitrequires careful interpretation. Although seemingly contradictory, given that geomorphological units were classified solely based on slope with a 25° threshold, it is possible that some weathered, broad peaks with slope angles below 25° (due to long-term denudation or footslope deposition) were classified as pediment. This highlights that slope alone is insufficient for distinguishing large geomorphological units; complementary criteria such as absolute elevation, ruggedness index, or larger-scale TPI should be used. The main limitation of this study is the lack of field validation and dependence on 30-meter DEM resolution, which may not capture small-scale local landforms (<30 m). Conclusion This study performed automated landform classification in northeastern Eqlid using TPI, with a focus on optimizing scale and neighborhood type. Results conclusively demonstrate that the rectangular neighborhood significantly outperforms circular, annular, and triangular neighborhoods, being the only type capable of discriminating all nine landforms in the region. Optimal scales for TPI calculation were the 3×3 window (for local landforms) and 45×45 window (for regional, structural landforms). Stream channels occupy the largest area, indicating the dominant role of erosional and drainage processes in shaping the landscape. In geomorphological unit separation, no playa unit was present; mountain peaks dominate the mountain unit, while stream channels dominate the pediment unit. However, the presence of mountain peaks within the pediment unit indicates that slope-based classification alone is insufficient for geomorphological unit separation, and complementary criteria are recommended. The approach presented here—scale optimization using RMSE and systematic comparison of neighborhood types—represents a methodological advancement over previous Iranian studies that used fixed windows and default neighborhoods. Automated TPI-based landform classification, with careful, area-specific selection of scale and neighborhood type, can be an efficient tool for environmental studies, land-use planning, and natural resource management. Nevertheless, field validation and higher-resolution satellite imagery (≤10 m) are strongly recommended for future research طبقه‌بندی و شناسایی نواحی مختلف بر اساس ویژگی‌های مورفومتری آن‌ها که بازتاب‌دهنده فرایندهای گوناگون مؤثر بر سطح زمین است، همواره از اهمیت ویژه‌ای برخوردار بوده است. هدف پژوهش حاضر، طبقه‌بندی اشکال مختلف سطح زمین (لندفرم‌ها) در شمال شرقی شهرستان اقلید می‌باشد. در این مطالعه، از شاخص موقعیت توپوگرافی (TPI) به منظور شناسایی و طبقه‌بندی لندفرم‌های منطقه مورد استفاده قرار گرفت. داده‌های ورودی به این مدل شامل شیب، انحنای عرضی، انحنای حداقل و انحنای حداکثر است. همچنین، به منظور دستیابی به بهترین درون‌یابی و تفکیک لندفرم‌های منطقه، انواع مختلف نقاط همسایگی (دایره‌ای، مستطیلی، حلقه حلقوی و مثلثی) مورد بررسی و مقایسه قرار گرفتند. یافته‌های پژوهش نشان داد که همسایگی مستطیلی، بهترین نوع همسایگی برای تعیین لندفرم‌های منطقه از روی مدل رقومی ارتفاع (DEM) می‌باشد. افزون بر این، نتایج حاصل از طبقه‌بندی مورفولوژیکی منطقه مطالعاتی حاکی از آن است که منطقه مورد مطالعه شامل ۹ نوع لندفرم (آبراهه، دره‌های آبراهه‌های میانی، زهکش‌های مرتفع/بالارود، دره‌های U شکل، شیب‌های باز، شیب‌های بالایی، یال‌های مرتفع، یال‌های شیب میانی و قله کوه) می‌باشد. بیشترین مساحت در منطقه مورد مطالعه به لندفرم آبراهه با ۷۷/۲۳ کیلومتر مربع تعلق دارد. همچنین، نتایج نشان داد که به ترتیب در واحد کوهستان و دشت سر، لندفرم‌های قله کوه و آبراهه بیشترین گستره را دارند.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Landslide Susceptibility Modeling and Assessment Using Logistic Regression (Case Study: Goujebel Basin, Ahar) مدل‌سازی و ارزیابی حساسیت وقوع زمین‌لغزش با استفاده از رگرسیون لجستیک (مطالعه موردی: حوضه گویجه بئل، اهر) 250405 10.22034/gmpj.2026.577152.1602 FA مسعود رحیمی استادیار ژئومورفولوژی، گروه جغرافیا، دانشکده ادبیات و علوم انسانی، دانشگاه فردوسی مشهد، مشهد، ایران 0009-0004-9418-8194 Journal Article 2026 02 22 Extended Abstract Introduction Landslides are a common geological hazard in hilly and mountainous terrains worldwide. They affect society and the livelihoods of the impacted communities. Landslides can lead to loss of life and cause considerable damage to infrastructure facilities, agricultural land, and public and private assets (Sujatha & Sridhar, 2021). Due to the interaction with other hazards and the spatially distributed nature of landslide events, it is essential to map susceptibility, particularly in areas that contain elements at risk (Bednarik et al. 2010). Landslide susceptibility can be assessed by applying a statistical model to records of past landslide occurrences along with a set of explanatory variables (Atkinson and Massari 2011). This study evaluated landslide susceptibility in the approximately 75 km² Goujebel Basin, located in the Ahar region of northwestern Iran. The basin is of particular importance for geomorphological hazard assessment because of the high frequency of landslides, the presence of landslide-susceptible geological formations, complex topographic conditions, and the major Tabriz–Ahar transportation corridor. The main objective of the study was to identify the key landslide conditioning factors and model landslide susceptibility using logistic regression within a GIS environment. Methodology In this study, logistic regression was employed as a multivariate statistical method to model landslide susceptibility. This method is widely used in landslide susceptibility studies because it can simultaneously analyze quantitative and qualitative variables, does not require normally distributed data, and effectively models binary outcomes. Eleven landslide conditioning factors, including elevation, slope, aspect, slope length, surface curvature, lithology, distance to faults, distance to streams, Topographic Wetness Index (TWI), Normalized Difference Vegetation Index (NDVI), and land use, were incorporated as independent variables. The landslide inventory layer was also defined as a binary dependent variable (presence or absence of landslides). The datasets used in this study included a geological map, an ASTER Digital Elevation Model (DEM), Sentinel-2 satellite imagery, Google Earth imagery, and field surveys. For model training and evaluation, the landslide inventory was randomly divided into training (70%) and testing (30%) datasets. Data processing and model implementation were performed in a GIS environment using TerrSet software. The model output was presented as a landslide susceptibility map classified into different susceptibility classes, and model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Results and discussion The results indicated that landslides are among the most significant geomorphological processes in the Goujebel Basin. More than 40 landslides of varying sizes were identified across the study area, some of which remain active and have the potential for reactivation. Landslides are primarily concentrated in the middle reaches of the basin, highlighting the combined influence of topographic, geological, and hydrological factors on slope instability. From a topographic perspective, most landslides occurred at elevations between 1,550 and 1,750 m and on slopes ranging from 10 to 20%. These findings indicate that moderate slopes provide favorable conditions for landslides because of the greater thickness of weathered materials, the presence of water sources, and human activities. In addition, north and northwest facing slopes exhibited higher landslide susceptibility because of greater soil moisture and lower solar radiation. Analysis of surface curvature further revealed that hill slopes and transitional zones between plains and mountains exhibit the highest susceptibility to landslides. From a hydrological perspective, proximity to streams and moderate to high values of the Topographic Wetness Index (TWI) played important roles in increasing slope instability. Water infiltration and the resulting increase in pore-water pressure were identified as the primary triggering mechanisms for landslides in these areas. Among the geological factors, lithology was identified as the dominant conditioning factor controlling landslide occurrence. The PlQ-c formation, consisting of unconsolidated conglomerates with marl interbeds, exhibited the highest landslide susceptibility, with about 88% of the recorded landslides occurring within this unit. This finding highlights the dominant role of the geotechnical properties of geological materials in controlling slope instability. Distance to faults also played a moderate role in landslide occurrence and may act as an aggravating factor under seismic conditions. From a land-cover perspective, land use and vegetation cover played relatively limited roles in controlling landslide occurrence, probably because of the relative homogeneity of vegetation cover across the basin and the dominant influence of geological and topographic factors. The logistic regression results indicated that lithology (2.82), slope (1.89), and elevation (1.17) were the most influential landslide conditioning factors. The model achieved a high predictive performance, with an area under the receiver operating characteristic curve (AUC) of 0.93. The resulting landslide susceptibility map showed that approximately 35% of the basin falls within the high and very high susceptibility classes, which are mainly distributed in the middle and downstream parts of the basin and overlap with settlements, agricultural lands, and transportation infrastructure. Conclusion The results indicate that landslide susceptibility in the Goujebel Basin is controlled by a combination of topographic, geological, and hydrological factors, among which lithology, slope, and elevation are the most influential factors controlling slope instability. The concentration of landslides in mid-elevation zones, on moderate slopes, and within unconsolidated geological formations highlights the importance of geomorphological conditions and material properties in landslide occurrence. The results also indicate that a considerable portion of the basin falls within the high and very high landslide susceptibility classes, overlapping with settlements and transportation infrastructure and thereby increasing the area's vulnerability. The logistic regression model implemented within a GIS environment successfully identified landslide-prone areas with high predictive accuracy and provides a scientific framework for hazard management, land-use planning, and risk mitigation. The findings of this study can support management decision-making, sustainable regional development, and the reduction of geomorphological hazards. این تحقیق با هدف ارزیابی حساسیت زمین‌لغزش در حوضه آبخیز گویجه‌بئل‌ (واقع در اهر) انجام شد. برای این منظور، از مدل آماری رگرسیون لجستیک به عنوان یک روش قدرتمند و پرکاربرد در تحلیل احتمال وقوع پدیده‌های دووجهی استفاده گردید. در این مدل‌سازی، یازده لایه عاملی مؤثر شامل ارتفاع، شیب، جهت شیب، طول دامنه، تحدب سطح زمین، لیتولوژی، فاصله از گسل، فاصله از آبراهه، شاخص رطوبت توپوگرافی (TWI)، شاخص پوشش گیاهی (NDVI) و کاربری اراضی به عنوان متغیرهای مستقل و لایه پراکنش زمین‌لغزش به عنوان متغیر وابسته (باینری) به کار گرفته شدند. نتایج تحلیل رگرسیون لجستیک نشان داد که سه عامل لیتولوژی (با ضریب 82/2)، شیب (با ضریب 89/1) و ارتفاع (با ضریب 17/1) به‌ترتیب مهم‌ترین عوامل کنترل‌کننده وقوع زمین‌لغزش در مقیاس حوضه هستند. الگوی مکانی زمین‌لغزش‌ها عمدتاً در بازه ارتفاعی 1550 تا 1750 متر، شیب‌های 10 تا 20 درصد و بر روی سازند حساس PlQ-c (کنگلومرای تحکیم‌نیافته با میان‌لایه‌های مارنی) متمرکز است. عواملی مانند فاصله از آبراهه، جهت شیب، تحدب سطح و شاخص رطوبت توپوگرافی در رده تأثیرگذاری متوسط و با نقش تشدیدکننده محلی قرار گرفتند، در حالی که فاصله از گسل، کاربری اراضی، پوشش گیاهی و طول دامنه کمترین تأثیر را در مدل نشان دادند. نقشه نهایی حساسیت زمین‌لغزش با دقت بالا (93/0 = AUC) تهیه شد. بر اساس این نقشه، حدود 35 درصد مساحت حوضه (شامل 5/14 درصد در کلاس خطر بسیار زیاد و 6/20 درصد در کلاس خطر زیاد) در معرض خطر بالای زمین‌لغزش قرار دارد. این پهنه‌های پرخطر عمدتاً در بخش‌های میانی و پایین‌دست حوضه واقع شده‌اند و با موقعیت سکونتگاه‌ها، اراضی کشاورزی و محور ارتباطی مهم تبریز-اهر همپوشانی قابل‌توجهی دارند.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Spatial assessment of relative land-subsidence risk by integrating ascending and descending Sentinel-1 LiCSBAS time series, infrastructure exposure, and physical–hydrogeological susceptibility: Marvdasht Plain, Iran ارزیابی مکانی ریسک نسبی فرونشست زمین با تلفیق سری‌های زمانی LiCSBAS مدارهای صعودی و نزولی، مواجهه زیرساختی و حساسیت فیزیکی–هیدروژئولوژیک: دشت مرودشت 252149 10.22034/gmpj.2026.595704.1625 FA ابوالقاسم گورابی دانشکده جغرافیای - دانشگاه تهران 0000-0002-2787-8687 ابراهیم مقیمی گروه جغرافیای طبیعی، دانشکده جغرافیا، دانشگاه تهران 0009-0002-5456-8622 منصور جعفر بگلو گروه جغرافیای طبیعی، دانشکده جغرافیا، دانشگاه تهران 0000-0001-9745-2318 موسی حسینی گروه جغرافیای طبیعی، دانشکده جغرافیا، دانشگاه تهران 0000-0001-7161-8711 رقیه نژادحسینی گروه جغرافیای طبیعی ، دانشکده جغرافیا ، دانشگاه تهران Journal Article 2026 08 06 Extended Abstract Introduction Land subsidence in alluvial plains cannot be interpreted for risk management solely from deformation rates because spatial priority also depends on exposed people and assets and on physical–hydrogeological susceptibility. The Marvdasht Plain in southwestern Iran is an agriculturally important alluvial plain that depends heavily on groundwater and contains rural settlements, transportation and energy infrastructure, and major archaeological sites. Previous studies have documented subsidence and hydrogeological stress in the region, but an integrated framework combining deformation hazard, exposure, susceptibility, component discrimination, and uncertainty has remained limited. This study developed a reproducible relative land-subsidence risk index (RI) by integrating ascending and descending Sentinel-1 LiCSBAS time series with spatial exposure and physical–hydrogeological susceptibility. The RI is intended for screening and monitoring prioritization, not for estimating structural-failure probability, economic loss, or absolute disaster risk. Methodology Sentinel-1 ascending and descending LiCSAR time-series products were processed using LiCSBAS. Annual deformation rates were estimated for valid pixels, and the series were divided at 2019.5 into P1 and P2. Vertical-equivalent rates were calculated under the assumption of predominantly vertical motion, and inter-period change was defined as Δv = vP2 − vP1. Agreement between ascending and descending products was assessed using Pearson correlation, bias, and RMSE as an inter-geometry consistency test rather than independent validation. The model comprised subsidence hazard (H), exposure (E), and physical–hydrogeological susceptibility (V). Hazard was derived from aligned deformation-rate products and normalized to 0–1. Exposure was constructed from 19 layers grouped into settlements, transportation, lifeline infrastructure, and productive land uses, using distance-decay functions and robust scaling. Susceptibility combined lithology, a groundwater-abstraction pressure proxy based on smoothed well density, and fault proximity. Because complete engineering and social vulnerability data were unavailable, V was treated as environmental susceptibility. Upper-level weights were obtained using a role-based virtual AHP framework. Consensus weights were 0.548551 for H, 0.293941 for E, and 0.157508 for V, giving RI = 0.548551H + 0.293941E + 0.157508V. The index was classified into five equal-width classes. Robustness was evaluated through deterministic weighting scenarios, leave-one-component-out tests, and 10,000 Monte Carlo simulations of AHP weight uncertainty. Villages, population, households, infrastructure, productive land uses, and selected heritage sites were screened against the final classes. Results and Discussion The final model covered 897.131 km², or 99.223% of the study boundary. Very-low, low, moderate, high, and very-high classes occupied 269.105, 223.920, 273.288, 128.352, and 2.465 km², respectively. Thus, 130.817 km², or 14.582% of the valid area, was classified as high or very high. These areas were concentrated mainly in the central and central-eastern plain, showing that high RI values resulted from the coincidence of subsidence hazard, exposed assets, and susceptible ground conditions rather than deformation magnitude alone. The moderate class covered 30.463% of the valid area. Hazard contributed 86.60% of the calculated RI variance, compared with 6.99% for exposure and 6.41% for susceptibility. Although hazard dominated spatial variability, E and V still modified the location and ranking of priority areas. Fourteen of the 73 villages with valid coverage were located in the high class. They contained 30,725 people and 9,308 households, representing 47.60% of the assessed rural population and 48.27% of rural households. High and very-high classes also coincided with 43.92% of valid railway length, 36.90% of power-transmission lines, 36.75% of freeways and highways, 36.40% of valid bridges, and 43.90% of industrial sites. These values indicate spatial co-location only and do not demonstrate observed damage or functional disruption. The high-plus-very-high share remained close under the AHP-consensus, equal-weight, and initial 0.60–0.30–0.10 scenarios, at 14.582%, 14.879%, and 12.954%, respectively. In contrast, removing individual components produced a wider range of 5.604–26.097%, indicating that model structure influenced the result more strongly than limited changes among the principal weighting schemes. Monte Carlo analysis yielded a mean high-plus-very-high share of 15.387%, a median of 14.708%, and a 95% simulation interval of 11.182–22.372%. The mean dominant-class probability was 0.963, and 92.20% of valid pixels had a dominant-class probability of at least 0.80; 115.236 km² had at least an 0.80 probability of remaining in the high or very-high classes. These probabilities represent classification stability under weight uncertainty, not damage probability. Temporal comparison showed high agreement between ascending and descending rates for P1 and P2, with r = 0.952 and 0.978, but lower agreement for Δv, with r = 0.462. The ascending track showed 19.83% intensified subsidence, 38.07% relative stability, and 42.10% reduced intensity; the descending track showed 0.13%, 90.53%, and 9.34%, respectively. Thus, the broad spatial pattern was more robust than the inferred inter-period change, and Δv should be treated as an exploratory temporal indicator rather than definitive vertical acceleration or independent validation. Heritage screening classified Tall-e Bakun as high, Toll-e Shoqa as moderate, and Persepolis as very low at representative points, while Naqsh-e Rostam, Naqsh-e Rajab, and Estakhr were outside valid coverage. NoData does not indicate low risk. Conclusion The framework provides a transparent basis for prioritizing subsidence monitoring in the Marvdasht Plain. High-priority areas emerge from the combined spatial influence of hazard, exposure, and physical–hydrogeological susceptibility rather than subsidence intensity alone. The concentration of nearly half of the assessed rural population and households in only 14 high-class villages also shows that management significance cannot be inferred from area percentage alone. Sensitivity and Monte Carlo analyses indicate that the main spatial pattern is relatively stable under reasonable changes in upper-level weights, while component removal produces larger changes, highlighting the importance of model structure. Nevertheless, RI remains a relative screening index. The absence of independent damage and control data prevented external ROC/AUC-type validation, and the model does not estimate structural failure, economic loss, or complete engineering and social vulnerability. Future work should incorporate independent damage inventories, groundwater-level and abstraction time series, GNSS and leveling observations, hydrogeomechanical modelling, infrastructure fragility information, and polygon-based heritage assessment. فرونشست زمین در دشت‌های آبرفتی زمانی برای مدیریت ریسک اهمیت بیشتری می‌یابد که شدت تغییرشکل، عناصر در معرض و حساسیت بستر به‌طور هم‌زمان ارزیابی شوند. هدف این پژوهش توسعه یک چارچوب بازتولیدپذیر و ممیزی‌پذیر برای پهنه‌بندی شاخص ریسک نسبی فرونشست در دشت مرودشت بود. مؤلفه خطر با استفاده از سری‌های زمانی Sentinel-1 پردازش‌شده در LiCSBAS و تلفیق هم‌تراز نتایج مدارهای صعودی و نزولی استخراج شد؛ مواجهه از ۱۹ لایه در چهار گروه و حساسیت فیزیکی–هیدروژئولوژیک از سنگ‌شناسی، نمایه فشار برداشت آب زیرزمینی مبتنی بر چگالی هموارشده چاه‌ها و مجاورت گسل‌ها ساخته شد. پایداری مدل با سناریوهای قطعی وزن‌دهی و ۱۰٬۰۰۰ تکرار مونت‌کارلو ارزیابی شد. از ۸۹۷٫۱۳۱ کیلومترمربع پوشش معتبر مدل، ۱۳۰٫۸۱۷ کیلومترمربع، معادل ۱۴٫۵۸۲ درصد، در طبقات زیاد و بسیار زیاد شاخص ریسک نسبی قرار گرفت. این پهنه‌ها ۱۴ روستا را دربر گرفتند که ۴۷٫۶۰ درصد جمعیت و ۴۸٫۲۷ درصد خانوارهای روستایی بررسی‌شده را شامل می‌شدند و با ۴۳٫۹۲ درصد از طول معتبر راه‌آهن، ۳۶٫۹۰ درصد از خطوط انتقال نیرو و ۴۳٫۹۰ درصد از واحدهای صنعتی هم‌مکان بودند. سه طرح اصلی وزن‌دهی نتایج نزدیکی تولید کردند، درحالی‌که حذف مؤلفه‌ها تغییرات بزرگ‌تری ایجاد کرد. بازه ۹۵ درصدی شبیه‌سازی برای سهم طبقات زیاد و بسیار زیاد ۱۱٫۱۸۲ تا ۲۲٫۳۷۲ درصد بود و ۹۲٫۲۰ درصد پیکسل‌ها احتمال تعلق به طبقه غالب حداقل ۰٫۸۰ داشتند. نتایج نشان می‌دهد پهنه‌های اولویت‌دار حاصل هم‌مکانی خطر، مواجهه و حساسیت هستند و اهمیت آن‌ها را نمی‌توان تنها از شدت فرونشست یا سهم مساحتی استنتاج کرد. بااین‌حال، اختلاف تغییرات دوره‌ای میان مدارهای صعودی و نزولی و نبود داده مستقل خسارت، دامنه استنتاج را محدود می‌کند. نقشه نهایی یک ابزار غربالگری و اولویت‌بندی پایش است و نباید به‌عنوان احتمال خسارت، خرابی سازه یا برآورد اقتصادی تفسیر شود.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Spatio-Temporal Analysis of Land Subsidence in Relation to Land Use Changes: A Geomorphological Approach (Case Study: Southern Plains of Alborz Province) تحلیل مکانی–زمانی فرونشست در ارتباط با تحولات کاربری زمین؛ رویکرد ژئومورفولوژیک ( نمونه موردی: دشت‌های جنوبی استان البرز) 252150 10.22034/gmpj.2026.582400.1611 FA ماهرخ سردشتی گروه جغرافیای طبیعی، دانشکده علوم اجتماعی، دانشگاه محقق اردبیلی عقیل مددی هیات علمی دانشگاه محقق اردبیلی صیاد اصغری سراسکانرود گروه جغرافیای طبیعی، دانشکده علوم اجتماعی ، دانشگاه محقق اردبیلی Journal Article 2026 05 19 Introduction Land subsidence as an anthropogenic geomorphological hazard has caused irreversible damage to infrastructure, water resources, and environmental stability in recent decades, especially in alluvial plains of arid and semi-arid regions of Iran. Alborz province, due to its transitional position between the Alborz mountain range and the central plains, the presence of fine-grained Quaternary sediments, excessive groundwater extraction, rapid urban and industrial development, and significant reduction of agricultural lands and water bodies, is considered one of the critical subsidence hotspots in the country. Despite scattered previous studies, few researches have simultaneously and integrally analyzed land use/land cover changes and subsidence rates with an anthropogenic geomorphology approach and feedback loop modeling. The main objective of this research is to analyze the spatio-temporal relationship between land subsidence and land use/land cover changes in the southern plains of Alborz province (Eshtehard, Hashtgerd, and Qazvin) during the periods 2016-2020 and 2020-2025, with emphasis on explaining causal cycles and developing a conceptual model of anthropogenic geomorphology. Methodology The present study is applied in terms of purpose and descriptive-analytical in nature, based on remote sensing and geographic information system data. To extract subsidence rates, 10 Sentinel-1 radar images (C-band, interferometric mode, ascending pass) were used during two time periods: 2016-2020 and 2020-2025. Radar data processing in SNAP software included orbit application, interferogram formation, Goldstein filtering, phase unwrapping, and topographic correction using the Range-Doppler method. The final phase was converted to vertical displacement, and the annual subsidence rate was classified into three categories: low (less than 10 cm/year), medium (10-20 cm/year), and high (more than 20 cm/year). To prepare land use/land cover maps for 2016, 2020, and 2025, Sentinel-2 (10 m resolution) and Landsat-8 (30 m resolution) optical images were classified using the maximum likelihood supervised classification algorithm in ENVI software. Four land use classes including barren lands, riverine areas, agricultural lands, and residential areas were defined, and validation was performed using 250 field control points and Google Earth images (overall accuracy above 85%). In the ArcGIS environment, subsidence and land use maps were integrated using spatial overlay and zonal statistics methods. The mean subsidence rate for each land use class was calculated, and land use transition matrix and hot spot analysis (Getis-Ord Gi* statistic) were performed to identify spatial autocorrelation patterns. Results and Discussion Land use change results showed that during 2016-2025, agricultural lands decreased by 50% (from 33,610 ha to 16,751 ha) and riverine areas decreased by 51% (from 4,032 ha to 1,950 ha). In contrast, barren lands increased by 43% (from 199,389 ha to 210,224 ha) and residential areas increased by 32% (from 15,356 ha to 20,319 ha). These changes indicate a transition from natural-productive functions to anthropogenic and non-productive functions, accompanied by reduced aquifer recharge, increased groundwater extraction, and intensified surface desiccation. The highest annual subsidence rate during 2020-2025 ranged between 23-25 cm/year, and the maximum cumulative subsidence was approximately 1.1 m in southern Fardis and northern Eshtehard, showing an 18 cm increase compared to the previous period (92 cm). Critical subsidence zones (more than 0.5 m) expanded from 35% in the first period to 45% in the second period, and the subsidence center migrated northeastward (along the Eshtehard-Fardis-Hashtgerd industrial axis). Spatial overlay analysis showed that the highest subsidence intensity occurred in barren lands (69,716 ha in the first period and 59,135 ha in the second period in the high-intensity class), which is attributed to aquifer desiccation, reduced pore pressure, and high compressibility of fine-grained sediments. Next, residential areas with 9,201 ha (first period) and 15,103 ha (second period) in medium to high subsidence zones indicate the role of increased surface load from urban structures in exacerbating layer compaction. Agricultural lands also contributed significantly to subsidence with 15,693 ha in the medium class (first period) and 15,253 ha (second period), resulting from excessive groundwater extraction for irrigation and water table decline. Spatial autocorrelation analysis (Gi* statistic) revealed that areas with positive correlation (hot spots) mainly coincide with barren land expansion and urban development in southern Fardis and northern Eshtehard (99% confidence level). The conceptual model (Figure 7 in the article) explains five main nodes including urban/industrial development, barren land expansion, groundwater extraction and water table decline, fine-grained layer compaction and subsidence, and ecological instability through two positive feedback loops: the first feedback loop (urban development ↔ increased groundwater extraction ↔ aquifer decline ↔ intensified subsidence) and the second feedback loop (subsidence ↔ runoff concentration ↔ vegetation degradation ↔ barren land expansion ↔ renewed subsidence intensification). Conclusion This research, by integrating Sentinel-1 radar data with Sentinel-2 and Landsat-8 optical data within a GIS and spatial statistics framework, and by presenting the first conceptual model of anthropogenic geomorphology for the southern plains of Alborz, demonstrated that land subsidence is a multi-causal phenomenon resulting from the interaction of land use changes, groundwater level decline, and compressibility of fine-grained sediments. The more than 50% reduction in agricultural lands and water bodies, simultaneously with a 43% increase in barren lands and 32% increase in residential areas, indicates a shift in the earth system equilibrium from a natural to an anthropogenic and unstable state. The highest subsidence rates (23-25 cm/year) and cumulative subsidence (1.1 m) are concentrated in southern Fardis and northern Eshtehard, and critical zones have expanded from 35% to 45%. The most important innovation of this research is the explanation of the positive feedback cycle among urban development, barren land expansion, groundwater extraction, and subsidence intensification, demonstrating that subsidence is not merely a physical phenomenon but an indicator of human-earth system instability in the region. Based on the findings, controlling subsidence in the southern plains of Alborz requires (1) integrated water resources management and a ban on excessive groundwater extraction, (2) revision of urban development patterns and prevention of horizontal urban expansion over subsidence-prone areas, (3) restoration of vegetation cover and agricultural lands to enhance aquifer recharge, and (4) continuous monitoring using InSAR techniques and machine learning-based prediction models. هدف این پژوهش، تحلیل روابط مکانی–زمانی میان تغییرات کاربری اراضی و نرخ فرونشست زمین در دشت‌های جنوبی استان البرز با رویکرد ژئومورفولوژی انسان‌زاد است. برای دستیابی به این هدف، از داده‌های ماهواره‌ای Sentinel-1 و Sentinel-2 و تحلیل‌های مکانی در محیط GIS استفاده شد. داده‌های راداری Sentinel-1 در نرم‌افزار SNAP پردازش و نرخ سالانه فرونشست زمین در دوره‌های ۲۰۱۶ تا ۲۰۲۰ و ۲۰۲۰ تا ۲۰۲۵ استخراج گردید. هم‌زمان، تغییرات کاربری اراضی از تصاویر اپتیکی به‌روش طبقه‌بندی نظارت‌شده و صحت‌سنجی میدانی به‌دست آمد. نتایج نشان داد که طی این دوره، اراضی زراعی با کاهش بیش از ۵۰ درصدی و پهنه‌های آبی با افت ۵۱ درصدی مواجه بوده‌اند، در حالی‌که اراضی بایر و مناطق مسکونی به‌ترتیب ۴۳ و ۳۲ درصد افزایش یافته‌اند. این تغییرات با افت قابل‌توجه سطح ایستابی و افزایش نرخ فرونشست هم‌زمان بوده است. بیشترین نرخ سالانه فرونشست بین ۲۳ تا ۲۵ سانتی‌متر و بیشترین نشست تجمعی حدود ۱/۱ متر در جنوب فردیس و شمال اشتهارد برآورد شد. نتایج نشان داد که بیشترین شدت فرونشست در اراضی بایر و زراعی رخ داده است که ارتباط مستقیم با خشکیدگی آبخوان و تراکم‌پذیری رسوبات ریزدانه دارد. مدل مفهومی تحقیق نشان داد که توسعه شهری، افزایش اراضی بایر و برداشت بیش از حد آب زیرزمینی، در قالب چرخه‌ای بازخوردی موجب تشدید فرونشست و ناپایداری مورفودینامیک سطح زمین شده است. در نهایت، نتایج پژوهش تأکید می‌کند که کنترل فرونشست در منطقه مستلزم مدیریت یکپارچه منابع آب، بازنگری در توسعه شهری و حفاظت از پوشش گیاهی برای حفظ پایداری ژئومورفولوژیک دشت‌های جنوبی البرز است.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Investigating the durability of snow cover using SWI, S3, NDSI and NDSII-1 indices and the effect of temperature changes on it in the period from 1994 to 2023 (study area: Belgeys Mountain) بررسی ماندگاری پوشش برف با استفاده از شاخص های SWI، S3، NDSI و NDSII-1 و تاثیر تغییرات دمایی بر آن در بازه زمانی 1994 تا 2023 (منطقه مورد مطالعه: کوه بلقیس) 253218 10.22034/gmpj.2026.405539.1442 FA مهدی فیض اله پور دانشگاه زنجان شعبان کرمی دانشجوی کارشناسی ارشد Journal Article 2023 07 05 Extended Abstract Introduction The advent of high-resolution remote sensing satellites has motivated various researchers to use multiple techniques and digital data sets to accurately map the snow-covered area on a regional and global scale. Indices are the most common automated methods used for snow and ice cover mapping. S3, NDSI and NDSII-1 indices are examples of indices that have the maximum power of discrimination between snow, ice, clouds and vegetation and have proven their role in snow mapping through satellite data. These indices use the reflective and absorptive properties of snow in the visible, NIR and SWIR bands. Although these indices are often used for snow cover mapping and their importance in extracting snow cover in different conditions has been proven, but water and cloud are the two main impurities that challenge the extraction of snow pixels from multispectral images. Methodology The method of this research includes choosing the right area, collecting data and satellite images of Landsat 8 and using spectral indices. OLI and TIRS images of Landsat satellite were used to extract snow cover in the studied area. For this purpose, Landsat 5 and 8 satellite images for the month of April of 1994, 2001, 2013 and 2023 were downloaded. All images were downloaded from the US Geological Survey website. Then, different bands of NDSI, NDSII-1, S3 and SWI indexes were used to estimate the snow cover, and then LST index was used to check the changes in ground surface temperature. In this research, the performance of snow spectral indices was tested using Pearson's correlation and coefficient of determination. Pearson's correlation is a statistical indicator for establishing linear relationships between two variables. Results and Discussion In 2023, all indicators obtained values close to each other. The lowest amount belongs to the S3 index and the highest amount of snow cover in these indices belongs to the NDSII-1 index. The difference between the lowest and the highest amount is equal to 3.27 square kilometers, which covers 1.3% of the surface of the studied area and shows that all the indicators have reached values close to each other. In 2013, the difference between the lowest and highest results in the indices was equal to 3.27 and it shows that the difference between this year and 2023 was at the same level. In the discussion of the use of indicators, the main goal is to accurately separate the complications that have interfered with the detection of snow cover. These phenomena in the studied area include water sources and cloud cover, which can be clearly seen from the produced images. In order to avoid this occurrence, several bands have been used in these indices so that this problem can be overcome. Therefore, in this section, Pearson's correlation coefficient and coefficient of determination have been used between each of the indices and bands used. These bands include green, red, NIR and SWIR bands. But these correlations were established between these bands and the SWI index. The reason for this is to find out the relationship between this index and the used bands. Among these bands, it can be seen that the SWI index has a strong negative relationship with the short-wave infrared or SWIR band. The correlation between these two indicators in the years 1994, 2001, 2013 and 2023 was equal to -0.93, -0.73, -0.89 and 0.87, respectively, in which the highest rate belonged to 1994. It means the same year with the highest amount of snow cover. The most positive correlation between this index was with the green band. Only in 1994, the red band had the highest correlation. In the rest of the years, the highest correlation was 0.86, 0.93, and 0.94 for the years 2001, 2013, and 2023, respectively. The lowest positive correlation belongs to the near infrared band. In the last part of the research, changes in ground surface temperature in this area were investigated to determine the effect of changes in ground surface temperature or LST on changes in snow cover. The lowest temperature in the region was -4.6 degrees Celsius on April 10, 1994. The expansion of snow covers this year has been a sign of heavy rainfall in this area during that time period. The highest recorded temperature for this region in the desired period of time is 34.46 degrees Celsius and it happened on April 29, 2001. Conclusion In this research, NDSI, NDSII-1, S3 and SWI indices were used, among which the SWI index is one of the newest indices in this field. In order to investigate the changes in snow cover, images from 1994 to 2023 were considered in a 29-year period. In 2023 and based on SWI indicators, the area of snow cover was equal to 28.41 square kilometers. The correlation coefficient between this index with the red, green, NIR and SWIR bands showed that the highest negative correlation between the SWI index and the SWIR band was established at -0.93 in 1994, and the highest positive correlation between this index and the green band was established in 1994. It was 0.94 in 2023. In the discussion of changes in the temperature of the earth's surface, it was also observed that the lowest temperature was -4.6 degrees Celsius in 1994 and the highest temperature was 34.46 degrees Celsius in 2001. در این تحقیق به بررسی قابلیت شاخص های NDSI، S3 و NDSII-1 در تمایز پیکسل های برف از ابر و آب در ارتفاعات کوه بلقیس پرداخته است. به منظور بررسی تغییرات پوشش برف، بازه زمانی 29 ساله از سال 1994 تا 2023 در نظر گرفته شد. بررسی ها نشان داد که پوشش برف در سال 1994 با افزایش قابل توجهی مواجه گردیده و معادل 07/160 کیلومتر مربع از سطح 16/253 کیلومتر مربعی منطقه مورد مطالعه را پوشش داده است. در سال 2023 و بر اساس شاخص های SWI مساحت پوشش برفی معادل 41/28 کیلومتر مربع برآورد گردید. برای تفکیک پوشش برف و مناطق فاقد پوشش برف در هر شاخصی باید از آستانه معینی تبعیت کند که این آستانه در شاخص SWI معادل 24/0 بوده است. ضریب همبستگی بین این شاخص با باندهای قرمز، سبز، NIR و SWIR نشان داد که بیشترین همبستگی منفی بین شاخص SWI و باند SWIR به میزان 93/0- در سال 1994 برقرار بوده و بیشترین همبستگی مثبت نیز بین این شاخص با باند سبز به میزان 94/0 در سال 2023 بوده است. در بحث تغییرات دمای سطح زمین نیز مشاهده شد که کمترین دما به میزان 6/4- درجه ساتیگراد متعلق به سال 1994 بوده و بیشترین دما به میزان 46/34 درجه ساتیگراد متعلق به سال 2001 بوده است. در نهایت چنین استدلال می شود که بر اساس نتایج همبستگی، شاخص SWI نسبت به سایر شاخص ها به نتایج بهتری در براورد مساحت پهنه برف دست یافته است.

انجمن ایرانی ژئومورفولوژی پژوهشهای ژئومورفولوژی کمّی 22519424 15 2 2026 09 23 Flood risk zoning under the influence of climate change based on CanESM2 scenarios using SWAT model in GIS environment (Case study: Nekarud watershed, Mazandaran) پهنه‌بندی خطر سیلاب تحت تاثیر تغییرات اقلیم بر اساس سناریوهای CanESM2 با استفاده از مدل SWAT در محیط GIS (حوضه آبخیز نکارود، مازندران) 253219 10.22034/gmpj.2026.502505.1543 FA صالح ارخی دانشگاه گلستان Journal Article 2025 01 27 Extended Abstract Introduction Climate change refers to irreversible changes in the average weather conditions in a region. Climate change as a change in the weather pattern is mainly caused by the emission of greenhouse gases from natural systems and human activities (Fawzy et al., 2020), which in the last few decades has caused an increase in the overall temperature of the earth, followed by the phenomenon that climate change has occured. According to the latest report of the Intergovernmental Panel on Climate Change, since 1971, the temperature of the earth's surface has been increasing at a faster rate than other fifty-year periods, at least during the last 2100 years (IPCC, 2022). In this research, it is tried to investigate the effects of climate change on the hydrological regime of the watershed and the condition of its floods in the future. CanESM2 climate model has been used under RCP2.6, RCP4.5 and RCP8.5 emission scenarios with SDSM statistical microscale model in the future time period of 2021-2050, and then geographic information system (GIS) and one of the decision methods. Multi-criteria decision making (MCDM) will be used for risk analysis and their ranking for this simulation. The resulting flood potential maps can be a suitable tool for strategic planning in the future. Therefore, the objectives of this research are as follows: 1) Simulating the runoff of the Nakarod watershed using the SWAT physical model, 2) Zoning the flood risk of the Nakarod watershed using the ANP model. Methodology The watershed of the Nekarud River is one of the watersheds of the Caspian Sea and is located in the geographical range of 53 degrees 17 minutes to 54 degrees 44 minutes east longitude and 36 degrees 28 minutes to 36 degrees 42 minutes north latitude. The basic data used in this research for the calibration and implementation of the micro-scale statistical model includes the minimum, maximum, precipitation and sunshine hours of the synoptic station of Baikla Nekarud in the period from 2001 to 2020. In this research, the three release scenarios of RCP2.6, RCP4.5 and RCP8.5 of the CanESM2 model in the statistical period of 2021 to 2050 were microscaled with the help of the observational data of the synoptic station of Baikla-Nakarod. In this research, three maps of distance from the river, basin shape coefficient and average maximum runoff in the upcoming period of 2021-2050 have been used for the flood risk zoning map due to climate change. To predict future temperature and precipitation variables, CanESM2 model under RCP8.5 scenario has been scaled with SDSM model, and SWAT model has been used to predict daily runoff in the future period. To complete the maps, the outputs of the SWAT model, including the construction of waterway lines from the map, which include DEM, the division of sub-basins, and the runoff of each sub-basin, which include other information on soil maps, land use, and precipitation, have been used. Finally, the layers are combined with the addition of ANP and the final flood risk map due to climate change is obtained. Results and Discussion The results show that both the maximum temperature and the minimum temperature of this station have increased in the future period. The range of temperature changes of the maximum station is between 0.08 and 0.28 degrees Celsius and the range of temperature changes of the minimum station is between 0.12 and 0.21 degrees Celsius. In the maximum temperature changes of this period, the highest temperature increase is related to this station under the RCP8.5 emission scenario with an increase of 0.28 degrees Celsius and the lowest temperature increase is related to the RCP2.6 emission scenario with an increase of 0.08 degrees Celsius. . Also, in the minimum temperature changes of this period, the highest temperature increase is related to the RCP8.5 release scenario with an increase of 0.21°C and the lowest temperature increase is related to the RCP2.6 release scenario with an increase of 0.12°C. The amount of runoff decreases in the hot months of the year, early autumn and late spring. Of course, the decrease in the hot season follows a uniform trend. There are exceptions in different scenarios. An increase is seen in late winter and early spring. The reason for this increase is directly related to the increase in temperature and the increase in rainfall in the coming period. The increase in temperature will increase the rate of snow melting in the region, and over time, the water sources in the high altitudes and cold areas will melt, which will lead to floods in the region. Most of the runoff occurred in late winter and early spring (months of January, February, March and April). So that the peak in RCP8.5 occurred in March, which can be justified due to the increase in temperature and the simultaneous snow melting in the basin of this peak. Also, the reason for the decrease in the scenario of RCP4.5 compared to both scenarios is more extreme rainfall occurred in this scenario in the summer and autumn months, but in RCP8.5, despite the decrease in runoff compared to RCP2.6, the runoff increased due to the effect of serious rains occurred in the months of March, April and May. Conclusion With the identification of the most critical climatic scenario in extreme events and maximum discharges, flood risk zoning of the region was carried out. For this purpose, from three average runoff maps of up to 30 years, basin shape factor and distance from the river were combined using ANP approach and classified into five classes. According to the zoning, 41.61 percent of the area of the basin is in very low and low risk (746.18 and 213.73 square kilometers), 20.78 percent in medium risk (479.46 square kilometers) and 35.71 percent in high and very high risk (368.65 and 455.55 square kilometers) in the range It is around the main river, which has many villages nearby Keywords: Climate change, flood risk, modeling, GIS, SWAT, Necarod. در عصر حاضر، با افزایش فعالیت‌های صنعتی و نادیده گرفته شدن مسائل زیست محیطی، اثرات تغییر اقلیم بیشتر نمایان شده و این پدیده را به عنوان یک معضل جهانی مطرح ساخته است. افزایش احتمال وقوع رخدادهای اکستریم اقلیمی مانند سیلاب و افزایش فرکانس و شدت آن از اثرات تغییر اقلیم است. منطقه شمال کشور، به دلیل داشتن اقلیمی مرطوب و معتدل و درنتیجه تغییرپذیری بالای بارش، ازجمله مناطقی است که در معرض سیلاب‌های مخرب قرار دارد. از اینرو در تحقیق حاضر تأثیرات تغییرات اقلیمی بر خطر سیل‌خیزی حوضه و پهنه‌بندی خطر سیل متأثر از این تغییرات در حوضه آبخیز نکارود واقع در شمال ایران مورد بررسی قرار گرفته است. به منظور پیش‌بینی پارامتر دما و بارش از مدل CanESM2 تحت سناریوهای انتشارRCP2.6 ، RCP4.5 و RCP8.5 با مدل ریزمقیاس گردانی آماری SDSM در دوره زمانی آینده 2021-2050 استفاده شده است. نتایج نشان دهنده افزایش 21/0 درجه سانتی‌گراد دما و افزایش شدت بارش در سناریو RCP8.5 است. بررسی وضعیت رواناب آینده با مدل هیدرولوژیکی SWAT انجام شده است. همچنین با مدل هیدرولوژیکی SWAT علاوه بر رواناب روزانه، نقشه حوضه‌بندی و خطوط آبراهه‌ها استخراج شده است. نتایج در بدبینانه‌ترین سناریو RCP8.5 گویای افزایش 1/3 درصدی رواناب نسبت به دوره پایه می‌باشد. در پایان نیز حوضه مورد مطالعه با تلفیق نقشه متوسط رواناب تحت سناریو RCP8.5، فاصله از رودخانه و ضریب شکل حوضه با استفاده از روش ANP در محیط GIS پهنه‌بندی شده است. با توجه به پهنه‌بندی، 61/41 درصد از مساحت حوضه در خطر بسیارکم و کم (18/746 و 73/213 کیلومتر‌مربع)، 78/20 درصد خطر متوسط (46/479 کیلومترمربع) و 71/35 درصد در خطر زیاد و بسیار زیاد (65/368 و 55/455 کیلومتر‌مربع) می‌باشد.