طبقه بندی و آشکارسازی تغییرات لندفرم های ژئومورفولوژی و شبیه سازی تغییرات با سلول های خودکار مارکوف (مطالعه موردی: حوضه آبریز سجاسرود) (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
یکی از جدید ترین روش ها، بهره گیری از توانایی های تصاویر ماهواره ای در جهت استخراج و بررسی تغییرات لندفرم ها در بازه های زمانی مختلف و پیش بینی تغییرات آن هادر آینده است. در همین راستا برای شناسایی لندفرم های حوضه آبریز سجاسرود و بررسی روند تغییرات و پیش بینی تغییرات تا سال 2050، از تصاویر ماهواره ای TM و OLI در بازه زمانی 1986 تا 2018 بهره گرفته شد. در ادامه جهت شناسایی لندفرم ها از مطالعات و برداشت های میدانی و نرم افزار Google Earth و نقشه توپوگرافی و تجزیه مولفه های اصلی با تصاویر لندست استفاده گردید. پس از انجام تصحیحات لازم بر روی تصاویر ماهواره ای، روش طبقه بندی حداکثر احتمال برای طبقه لندفرم ها بکار گرفته شد. نتایج نشان داد پهنه های پوشش گیاهی و دشت آبرفتی به ترتیب 159/47 و 26/57 درصدافزایش مساحت و درمقابل کوه ها و تپه ها، پادگانه های آبرفتی، مخروط افکنه ها و آبرفت های جدیدبا کاهش مساحت مواجه بوده اند و همینطور نتایج الگوریتم MNF نشان داد بیشترین شدت و سرعت تغییرات مربوط به دشت های آبرفتی با مساحت 45/9892 و کوه ها و تپه ها 13/9481 هکتار و کمترین سرعت تغییرات مربوط به پادگانه های آبرفتی با میزان 36/10 هکتار است. در نهایت نتایج شبیه سازی CA-MARCOV نشان داد که در 32 سال آینده مخروط افکنه ها با مساحت 97/64- کیلومترمربع بیشترین تغییرات کاهشی، پوشش گیاهی و آبرفت های جدید به ترتیب با مساحت 96/120 و 23/0 کیلومتر مربع بیشترین تغییرات افزایشی را خواهند داشت.Classification and detection of changes in geomorphological landforms and simulation of changes with Markov automatic cells (Case study: Sajasrood catchment)
IntroductionIdentification of geomorphological landforms is very important in terms of the origin of their formation and the process of their changes in the past, present and future. Therefore, natural obstacles and the extent of the study areas and field studies are challenging. To solve such problems, the use of remote sensing technologies for environmental management and natural hazards is essential. One of the newest methods is to use the capabilities of satellite images to extract and study changes in landforms at different time intervals and predict their changes in the future. The main goal of this research is to classify, reveal and simulate the changes in the main and important landforms of the Sejasrud watershed using Landsat satellite images and the maximum likelihood supervised classification algorithm between 1986 and 2018, as well as predicting the changes using the Markov method for The year is 2050.MethodologyThe research method in this research is applied in terms of purpose and in terms of survey-analytical nature. Field and library methods were used to perform and collect information. LandsatTM and OLI satellite images from 1986 and 2018 were used to collect preliminary data .In this regard, in addition to extensive field studies, Google Earth images and topographic maps of the area were used to better identify land forms. After preparing the data required for better detection and detection of geomorphological landforms according to, different stages of image processing were performed. Then, for classification, educational samples that had previously been radiomatically and atmospherically corrected were used. Using trial and error process, the best method for educational examples was obtained by generating a vector layer on images and converting them into educational examples in ENVI5.3 software environment. In the next step, using the supervised classification method, the maximum probability of land classification of the surface forms of the studied area was studied. Finally, using the functions related to the study of changes in landforms, the rate of change and simulation until 2050 was done. In the process of this research, ENVI 5.3, ArcGIS10.1 EXCEL2013 TERRSE IDRISI software were used.Results What can be deduced from Tables (5, 4, 3) and Figure (7) is the extensive changes in the surface area of important geomorphological landforms of Sajasrood catchment, which shows the dynamics of effective processes in landforms and strong performance of tectonic forces, especially external dynamics and human activities. In the evolution of geomorphological forms. According to statistics and information obtained over 32 years, alluvial fans and mountain and river garrisons had the most decreasing change. Instead, the area of alluvial plains and vegetation (Figures 7 and 8) show a significant increase. The area of mountainous areas, hills and recent alluviums also show slight decreasing changes. The Markov model according to Table (4) showed that new alluviums, alluvial barracks and alluvial fans have the highest pixel transfer to the alluvial plain landform, respectively. Also, according to Table (5), the highest area displacement is related to alluvial fans and alluvial plain. The reasons for these reductions and increases are mostly due to the extensive industrial and agricultural activities in the last three decades and climate change. The proliferation of alluvial fans, alluvial garrisons, and new alluviums is closely related to the amount of precipitation, especially snowfall, which has unfortunately declined over the past three decades. On the other hand, agricultural activities have experienced tremendous growth in the last three decades, which has led to changes in land use in various geoforms such as alluvial fans, mountain slopes, hills, barracks and recent alluvial beds, as well as the physical, chemical and biological properties of land surface. Has changed and transformed. Because in the process of baseline pixel classification, the smallest change in ground surface effects in different spectral bands shows its satellite images. Therefore, spatial changes in terms of area and shape over a period of 30 years are quite natural and logical. Activities related to cryoclasts, thermoclasts, and snow and rain constantly cause erosion of mountainous and hilly areas. On the other hand, agricultural activities in the foothills and hills affect the reflection of satellite electromagnetic waves. As a result, parts of the landforms mentioned in the vegetation class may be included in the image classification. However, due to the active tectonic activity in the basin, the effective role of this factor should not be overlooked. Figure (8) shows the process of landform changes in terms of the intensity or severity of the changes. An increasing or decreasing trend means more intensity and speed of change. ConclusionThe mountainous region of Sajasrood catchment and its various geomorphological phenomena are changing due to different geomorphological processes. Landsat satellite imagery was used to study the geomorphological changes of the surface forms of this catchment and to discover its 32-year changes. After preprocessing, processing and post-processing steps, the geoforms were extracted using the supervised classification method using the maximum similarity method. Then, using two functions, Change Detection Statistics and Image Change Workflow in ENVI software environment, which are used in the field of change detection, changes were extracted statistically and map.The results of the research showed that the biggest incremental changes are related to alluvial plains and mountains and hills, and the biggest decreasing changes are related to alluvial cones. The results of table (6), which is a comparison between the 2018 map and the 2050 forecast map, indicate an increase in the area of new vegetation and alluvium and a decrease in the area of other landforms. This is completely logical because with the expansion of the physical development of cities and villages, the expansion of industries and the cultivation of areas related to the levels of alluvial cones and alluvial plains, foothills and hilly lands on the extent of covered areas.A plant will be added and other side effects will be reduced.Keywords: Land form, classification, Markov model, Sajasrud river basin
تبلیغات
