شناسایی کانال های گیسوییِ پخشِ سیلاب روی مخروط افکنه اسلام آباد در حوضه یزد اردکان (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
کانال های گیسویی به شاخه های فرعی جریان های رودخانه ای اطلاق می شوند که عمدتاً روی سطوح مخروط افکنه ها و در مواردی نیز در گستره دشت سرها شکل می گیرند. این کانال ها به عنوان مسیرهای اصلی توزیع سیلاب در مخروط افکنه ها عمل می کنند و شناسایی دقیق آن ها در پیش بینی مخاطرات طبیعی سیل، فرسایش و مدیریت زیست محیطی حائز اهمیت است. در مخروط افکنه اسلام آباد یزد، گسترش سیلاب به تشکیل لندفرم های فرسایشی ناشی از جریان های سطحی مثل آبکندها منجر شده که درصورت عدم مهار، خطری جدی برای زیرساخت های حیاتی مانند شبکه های حمل ونقل محسوب می شوند که مهم ترین مسئله این تحقیق است. هدف اصلی این پژوهش، تهیه نقشه مکانی این کانال های سیلابی با دقت بالا است. در این مطالعه با بهره گیری از تصاویر ماهواره ای GeoEye (چشم زمین) و الگوریتم تشخیص اختلاف کنتراست که روشی مبتنی بر قطعه بندی تصاویر است، توزیع فضایی کانال های گیسویی به صورت نیمه خودکار استخراج شد. نتایج پیاده سازی این روش در پلتفرم نرم افزاری E-Cognition با رویکرد پردازش شیءگرا نشان داد که خطوط چندضلعی یا پلی گون ترسیم شده، همسو با ریخت شناسی سطح مخروط افکنه است و جزئیاتی مانند الگوی آبکندها، شبکه کانال های گیسویی و مسیرهای پخش سیلاب را با دقت بالایی آشکار می سازند. همچنین بخش فعال مخروط افکنه اسلام آباد شناسایی و در قالب تصویر آورده شد. این یافته ها گامی مؤثر در مدیریت ریسک سیلاب و حفاظت از اکوسیستم های مناطق خشک محسوب می شود. براساس یافته های این تحقیق، می توان نتیجه گرفت که استفاده از روش قطعه بندی مبتنی بر اختلاف تباین، کارآیی بالایی در تشخیص فرم مخروط افکنه ها نشان داده است. این روش توانسته است علاوه بر تشخیص فرم واقعی تقریباً مخروطیِ مخروط افکنه اسلام آباد که با اریب به سمت چپ است، کانال های گیسویی سطح مخروط افکنه مذکور را هم به صورت نیمه خودکار با ترسیم خطوط پلی لاین به خوبی شناسایی کند. دلیل موفقیت این روش در تشخیص فرم و عوارض سطح مخروط افکنه اسلام آباد، بهره گیری هم زمان از اطلاعات طیفی و ویژگی های مورفولوژیکی مثل اندازه، شیب و شکل این لندفرم است.Identification of Braided Channels on the Islamabad Alluvial Fan in Yazd Ardakan Basin
Braided channels are subsidiary branches of riverine flows that primarily develop on the surfaces of alluvial fans and, occasionally, across floodplains. These channels serve as vital pathways for distributing floodwater across alluvial fans, making their accurate identification essential for predicting natural hazards, such as floods and erosion, and conducting effective environmental management. In the Islamabad Alluvial Fan of Yazd, the expansion of floodwaters has led to the formation of erosional landforms, such as gullies created by surface runoff. If not properly managed, these features pose significant risks to critical infrastructure, particularly transportation networks, which was a central focus of this research. The primary objective of this study was to create a high-accuracy spatial map of these flood channels. By utilizing GeoEye satellite imagery and a contrast difference detection algorithm, an image segmentation-based approach was employed to semi-automatically extract the spatial distribution of braided channels. Implementing this method on the eCognition platform using an object-based processing approach revealed that the generated polygons aligned well with the surface morphology of the alluvial fan. They effectively delineated features, such as gully patterns, braided channel networks, and floodwater dispersal routes. Additionally, the active area of the Islamabad Alluvial Fan was identified and visualized. These findings marked a significant advancement in flood risk management and the protection of ecosystems in arid regions. This research concluded that the contrast-based segmentation method had proven highly effective in detecting the morphology of alluvial fans. It successfully identified the approximately conical shape of the Islamabad Alluvial Fan, which was left-skewed, and detected the surface distributary channels in a semi-automatic manner through polyline drawing. The success of this method in capturing the form and surface features of the Islamabad Alluvial Fan was attributed to its simultaneous use of spectral information and morphological characteristics, such as size, slope, and shape of the landform. Keywords : Segmentation, Contrast, Flood Spreading, Braided Channels, Islamabad, Yazd. Introduction Braided channels are dynamic, branching subsidiary channels of river systems found in a variety of geographical environments, including the surfaces of alluvial fans, river deltas, mountainous gravel streams, sandy-bed rivers, and sedimentary plains. These channels are a significant focus for scientists across multiple disciplines. For geomorphologists, braided channels commonly found in high-altitude and periglacial regions are important for understanding erosion and sediment transport. For water resource and civil engineers, identifying these channels is crucial in the design of dams, water intake systems, and water transfer channels, all of which face threats from reservoir sedimentation due to high sediment transport and erosion rates. Braided channels serve as vital conduits for distributing floodwaters on alluvial fans, playing a central role in shaping hydrological and geomorphological processes. Economically, the alluvial deposits associated with braided channels form significant hydrocarbon reservoirs and are important sources of sand and gravel. These channels consist of numerous alluvial pathways that continuously separate and rejoin around sediment bars and islands, creating an intricate structure that resembles a braided rope. The formation and evolution of braided channels depend on water dynamics and watershed characteristics. As primary pathways for flood distribution on alluvial fans, their identification is critical for flood management. If not properly recognized and managed, braided channels can pose significant risks to regional infrastructure. Therefore, understanding and identifying these channels is essential for predicting natural hazards, such as floods and erosion, and conducting effective environmental management. Materials & Methods Modern techniques for analyzing landform features utilize object-based patterns and satellite image segmentation as essential tools for extracting and classifying geomorphological landforms. In this research, the spatial distribution of braided channels on the Islamabad Alluvial Fan in Yazd was semi-automatically detected using the contrast difference method. This approach employs a hierarchical segmentation strategy, breaking down images from whole to parts by dividing features into smaller, homogeneous sub-units. The method focuses on variations in pixel brightness, converting image features into binary regions (black and white). By determining spectral thresholds, dark and bright areas are separated through the delineation of distinct polygons. This technique enables identification of locations where braided channels form, which are critical for understanding areas prone to hazardous flood distribution and represent the active sections of the alluvial fan. The study concentrated on the Islamabad Alluvial Fan within the Yazd-Ardakan Watershed near Taft County, where frequent flooding had resulted in erosional landforms, including gullies formed by surface runoff. The uncontrolled expansion of these features posed significant risks to critical infrastructure, particularly transportation networks, highlighting the importance of this research. The primary objective of this study was to create a high-precision spatial map of flood channels to enhance flood risk mitigation strategies. To achieve this, GeoEye satellite imagery characterized by high spatial resolution and a contrast difference detection algorithm—an advanced image segmentation technique—were employed. This semi-automatic method was implemented on the E-Cognition platform using an object-based image analysis approach. The results demonstrated that the generated polygons closely aligned with the surface morphology of the alluvial fan, accurately depicting details, such as gullies, braided channel networks, and flood distribution pathways with remarkable precision. Additionally, the active section of the Islamabad Alluvial Fan was effectively identified and visualized. Research Findings One of the primary achievements of this method that captured researchers' attention was the radial pattern of braided channels observed in the output maps. This pattern illustrated the branching distribution of channels radiating from a central core and exhibited remarkable morphological consistency with Bristow's (1993) findings regarding the formation dynamics of alluvial fans in arid environments. Other significant findings of this study included: Spatial Mapping Accuracy: Integration of GeoEye imagery with object-based methods allowed for the precise extraction of braided channels at a spatial scale, significantly enhancing the resolution of flood hazard maps. Identification of Active Fan Section: By mapping the concentration of primary braided channels, the study revealed that the most active channels were predominantly located in the distal section (base) of the Islamabad Alluvial Fan. Morphological Alignment: The mapped polygons exhibited a strong correlation with the fan's surface features, confirming the method's reliability in complex arid environments. Geometric analysis revealed a left-skewed conical shape at the fan's base. Risk Reduction: Identification of active fan zones and erosion hotspots provided valuable insights for infrastructure protection and flood risk management. Discussion of Results & Conclusion Based on the results and findings of this research, it could be concluded that the majority of braided channels had formed in the distal section, or approximately at the base, of the Islamabad Alluvial Fan, which was also where the active section of the fan was located. The hierarchical whole-to-part algorithm based on contrast difference detection used in this study effectively identified the general morphology of the asymmetric conical alluvial fan and surface details, such as braided channels, through continuous polygonal lines. This method worked by gradually dividing image features into smaller subunits, continuing the segmentation process until spectral-spatial homogeneity criteria were met in each unit. The final homogeneous units were referred to as "image objects". Since these objects were derived from the hierarchical decomposition of large-scale features, this approach was also known as the "nested decomposition algorithm". The success of this method could be attributed to the integration of spectral data from electromagnetic reflection values across different bands with morphometric parameters, such as shape, density, and drainage patterns. While spectral data alone might not effectively distinguish landforms with similar reflectance, their combination with indices—such as the pyramidal shape coefficient characteristic of alluvial fans—significantly enhanced classification accuracy. These results signified a pioneering advancement in flood risk management and ecosystem protection in arid regions. The findings suggested that the presence of braided channels indicated high flow discharge and a significant concentration of coarse-grained sediments, such as sand, within the riverine system. Thus, identifying these channels in older sedimentary layers could be interpreted as evidence of past wet periods or frequent flood events. This characteristic positioned braided channels as vital tools for reconstructing historical climatic changes and conducting geomorphological studies as their formation patterns were directly linked to long-term moisture and hydrological fluctuations. This research underscored the importance of identifying braided channels as a critical factor in flood risk management, soil erosion reduction, and maintaining hydrological balance. Furthermore, these aquatic networks were not only valuable sources of construction materials, but their unregulated exploitation could lead to irreparable consequences for arid ecosystems. Consequently, this study demonstrated that the proposed method was an effective tool for monitoring morphological changes, identifying high-risk flood areas, and promoting the sustainable management of natural resources. Acknowledgements This study is part of the approved research project No. 210, which was dated April 23, 2023 and titled "Identification of Floodplain Hairpin Channels on the Eslamabad Alluvial Fan". The project was made possible through the generous financial support of the esteemed Provincial Research Council. We extend our heartfelt gratitude and appreciation to the honorable members of this council for their support.
تبلیغات
