تحلیل و پیش‌بینی گسترش شهر مشهد با استفاده از تصاویر ماهواره‌ای چندزمانه و زنجیره مارکوف (طی سال‌های 1404-1379)

گسترش فیزیکی شهرها یک فرایند پویا و پیوسته است که در آن مرزهای شهر و فضای فیزیکی در جهت‌های عمودی و افقی از نقاط کمی و کیفی افزایش می‌یابد. روش تحقیق حاضر از نوع توصیفی تحلیلی و از نظر هدف کاربردی است تحقیق حاضر در پی ارزیابی و گسترش شهر مشهد در طی سال‌های 1379-1395 و سپس پیش‌بینی تغییرات تا سال 1404 است. داده‌های مورد نیاز پژوهش به روش اسنادی؛ و برای پی بردن به نوع و میزان تغییرات رخ داده در منطقه مورد مطالعه از تصاویر ماهواره لندست، سنجنده etm سال‌های، 1388، 1379 و سنجنده oli استفاده شده است. پس از عملیات طبقه‌بندی که از روش نظارت شده و الگوریتم حداکثر مشابهت استفاده شده است، در ادامه جهت پی بردن به تغییرات صورت گرفته در کاربری اراضی شهر مشهد که شامل کاربری‌های باغات و زمین‌های کشاورزی، محدوده‌های ساخته شده، اراضی بایر و مراتع مد نظر قرار گرفته از مدل زنجیره مارکوف استفاده شد؛ همچنین برای پیش‌بینی روند تغییرات تا سال 1404 از مدل ca استفاده شده است. همچنین در این پژوهش جهت اعتماد به طبقه‌بندی صورت گرفته از شاخص کاپا استفاده شده است. نتایج به دست آمده نشان می‌دهد که بیشترین تغییر، در سال‌های 79 تا 88 مربوط به محدوده باغات و اراضی کشاورزی بوده است، اراضی بایر در سال 1388 به نسبت سال 1379 کاهش یافته است اما در سال 1395 نسبت به سال 1388 افزایش یافته است. مساحت مراتع در سال 1388 به نسبت سال 1379 کاهش یافته است اما در سال 1395 نسبت به سال 1388 و 1379 افزایش چشمگیری داشته است. همچنین در طی 3 بازه زمانی 1395، 1388، 1379 بیشترین تغییر در کاربری‌ها مربوط به محدوده‌های ساخته شده است که بر اساس پیش‌بینی مارکوف در افق 1404 این کاربری حدود 121٫57% دچار تغییر خواهد شد؛ بنابراین نتایج مقاله حاضر می‌تواند به‌عنوان هشدار و تلنگری برای برنامه ریزان و مدیران شهری باشد تا با برنامه‌ریزی مناسب مانند سیاست‌های عمودی سازی و به‌عبارت‌دیگر گسترش در ارتفاع و نه در سطح، از گسترش بی‌رویه شهر به سمت باغات و اراضی کشاورزی جلوگیری نمایند.

Analyzing and Forecasting the Expansion of Mashhad City from 2000 to 2025 through Multitemporal Satellite Imagery and Markov Chain

1 Introduction;The physical expansion of cities is a dynamic and continuous process through which the city boundaries and the physical spaces in vertical and horizontal directions are increased both quantitatively and qualitatively. If urban growth is uncontrollable and unplanned, it may negatively affect spatial benefits and ultimately lead to urban expansion. The present study aims to investigate the expansion of Mashhad City in Iran from 2001 to 2017 and then forecast the changes until 2026. A Markov chain is a mathematical and probabilistic method which acts as a random process via which the future state of a pixel depends only on its predecessor and is predicted on it. The direct result of this model is the transmission probability matrix, but in this model no geographic perceptions are obtained. In addition, modeling a single map and representing the spatial distribution of classes is not generated. To solve this problem, the CAMarkov model was designed by John Von Neumann in the 1950s in order to add a spatial feature to the Markov model.;2 Materials and Methods;In the present research, the ETM images of the 2000 and 2009, and OLI images of 2016 from the Landsat Satellite were employed. The images were taken in 28.06.2000, 6. 6. 2009, and 25. 5.2016, respectively. Initially, Landsat Satellite images were geometric and radiometric corrections to reduce the satellite imagery errors. As such, the study area is separated from the images and it is attempted to classify the satellite data. The method used to classify the information is the monitoring method via which educational samples are used to classify the pixels. This means that by defining the specific pixels of each image for each of the classes, classification is performed in the form of the considered classes. Moreover, the maximum similarity algorithm is used for the classification of monitoring. In this method, the reflective value method and each pixel are unknown. In terms of the variance and covariance of a particular spectral reaction class, it is assumed that the distribution of the data of each class is based on the normal distribution around the pixel average of the given class. Practically, the variance and covariance, and the mean of the various classes of each satellite image are calculated for the classification of phenomena; thus, each pixel belongs to a class whose presence in that class is more likely to occur. In order to find out the changes in land use in the study area (Mashhad City), including gardens and agricultural land uses, builtup areas, grassland, and rangelands, the Markov chain module was employed. In the Markov chains, the landcover classes are used as the chain states. According to this analysis, we always use two raster maps called case models. In addition, the two maps illustrate the time interval between two the images and the predicted interval in the 1401 horizon in the CAMarkov model. The output of the Markov model also includes the possibility of turning the status and matrix of the converted areas in each class, and finally the images are probably conditional for different land use conversions. In this study, Cohen 's kappa coefficient (κ) was used for confirming the classification.;3 Results and Discussion;In this research, changes in land cover in gardens and agricultural land uses, and builtup areas, meadows, and rangelands using satellite imagery in a period from 2000 to 2016, and also maximum similarity algorithm, monitoring method and Markov chain model were employed. In the Markov chain model, the cover classes are used as the chain states. In fact, the transmission area matrix represents the number of pixels which are converted from a class to another and has changed from any landuse to another within the same time since 20002016.;Based on the CAMarkov model, four floors of the land cover in the land uses mentioned in the strategic perspective document for 1404 (2025) were forecasted. In this regard, it was determined that during the years of 2016, 2009, and 2000, the areas of land ​​uses of the builtup areas were significantly increased, and also the rangelands were expanded relatively; however, the areas of ​​ gardens and agricultural land uses have severely decreased. In addition, the areas of ​​meadows have decreased. The area of ​​land uses in the strategic perspective document for 1404 (2025) is somehow similar to those in 2016; therefore, the gardens and agricultural land uses will change as 90.74%, builtup land uses as 121.57%, meadows as 26.21%, and rangelands as 100%.  ;In fact, it was determined that the largest area of ​​transmission in different land uses was in the period between 1999 to 2009 for grasslands, gardens and agricultural land uses, builtup areas, and meadows. In addition, the highest probability of transmission area is in the period from 2000 to 2009 is related to meadows, rangelands, gardens and agricultural land uses, and builtup areas, respectively. The largest transmission area of the mentioned land uses is from 2009 to 2016, which is related to meadows, rangelands, builtup areas, and gardens and agricultural land uses, respectively. The highest transmission area of land uses is within the time interval of 20092016 related to meadows, rangelands, builtup areas, gardens and agricultural lands, respectively. It is also revealed that the area of ​​gardens and agricultural land uses was expanded in 2009 as compared to that of 2000, whereas it was significantly reduced in 2015.;The area of the builtup areas during the mentioned years was expanded, and in 2016 (as 359973900 m2) it was significantly increased. Meadows’ area was decreased in 2009 as compared to 2000, but it increased in 2016 as compared to 2009. In 2009, the area of rangelands was decreased as compared to 2000; however, it was again increased significantly in 2016 as compared to the years of 2009 and 2000.;4 Conclusion;According to the results obtained from the maps, in the three years of 2000, 2009, 2016, the most changes were related to builtup areas. Therefore, during this period the construction and physical growth of the city were mostly in the northwest direction. Moreover, because constructions usually are done on gardens and agricultural land uses, the decrease in the area of gardens and agricultural land uses, and consequently the increase of the builtup areas can be observed in this part of the city. According to the map of 2016, the gardens and agricultural land uses still remains in the southeast part of the city. One reasons for this issue may be the lack of growth of the city in this direction. One of the important issues in the field of urban planning is how urban spatial development and its resulting pattern are. The pattern derived from the spatial distribution of urban human activities called urban form is changing due to the dynamic and changing nature of cities. Urban growth is horizontal and vertical, and Mashhad has a horizontal urban growth. This form of urban growth over time has led to the disappearance of gardens and agricultural land uses and their transformation into builtup areas, resulting in economic, social and environmental crises. Therefore, land use management planning is one of the most important issues in Mashhad. If land use management planning is done correctly, optimally and accurately, many urban issues and problems may be solved.;