شبیه‌سازی تغییرات کاربری اراضی و دینامیک شهری (مطالعه موردی: شهر مراغه)

در دهه‌های اخیر همگام با رشد شهرنشینی، مدل‌های مختلفی جهت بررسی و پیش‌بینی رشد شهری بکار گرفته شده است.در این زمینه، ماشین‌های خودکار سلولی در چهارچوب رویکرد سیستمی وارد مباحث جغرافیایی شد. مدل‌های مختلفی به منظور ایجاد قوانین گذار در ماشین‌های خودکار سلولی ترکیب می‌شوند که در این زمینه می‌توان به زنجیره مارکوف اشاره کرد.در زنجیره مارکوف،حالت آینده یک سیستم تنها وابسته به حالت پیشین سیستم است، درحالیکه در ماشین‌های خودکار سلولی،حالت آینده یک سیستم نه‌تنها وابسته به حالت سابق، بلکه وابسته به حالت همسایگان نیز می‌باشد و همین ویژگی است که به آن ماهیت فضایی یا جغرافیایی می‌بخشد. پایش الگوی رشد شهر درطی تقریبا 30 سال گذشته روند توسعه شهر مراغه عمدتا در جهات شرقی و بر روی زمین‌های بایر رانشان می دهد.شهر مراغه در بین سال‌های 1369 تا 1379 یک رشد نسبتا سریعی را تجربه کرده و پس از آن نیز به صورت پیوسته عمدتا به سمت شرق توسعه یافته است. شبیه‌سازی فضایی الگوی رشد شهر نشان می‌دهد که در سال‌های آتی این روند همچنان تداوم خواهد یافت. بطوریکه در طی 17 سال آینده 774 هکتار از اراضی بایر و دیم و درحدود 417 هکتار از اراضی کشاورزی و باغات به ساخت وسازهای شهری تبدیل خواهند شد. همپوشانی فازی لایه‌های مختلف نشان می‌دهد که مکان‌یابی اولیه توسعه فیزیکی شهر مراغه به طور مناسبی صورت گرفته و بهتر است روند فعلی توسعه با تمایل بیشتر به سمت جنوب‌شرق، اجتناب از ساخت و ساز در شیب‌های تند و همچنین جلوگیری از تغییر کاربری اراضی کشاورزی به شهری تداوم یابد.

simulation of landuse changes and urban dynamics using ca markov hybrid model case study: maragheh city

introduction in coming decades, the rapid increase of large cities in the developing world and the transformation of urban landscapes in the developed world will be among the greatest challenges to human welfare and a viable global environment. in this context, there are several approaches for modeling urban development. in the past three decades, studies of nonlinear process and open systems have led to the emergence of new understandings of complex systems and their evolution. based on these understandings, cities are looked at as complex and open systems that have the capability of self organization. urban models based on the automata technique have also emerged under the paradigm of a self organizing system, with cellular automata (ca) being the simplest but most popular in action. in this research, the spatial expansion of maragheh city was simulated using cellular automata markov chain hybrid model. metodologywhite (1998) defined a ca as ‘‘a discrete cell space, together with a set of possible cell states and a set of transition rules that determine the state of each cell as a function of the states of all cells within a defined cell space neighborhood of the cell’’. in the ca framework, dynamics are represented as a change in the state of grid cells from one time step to the following time step. the cell need not, however, necessarily change its state. what happens to each grid cell is defined by a transition rule or transition rules. if the transition rule requires that the state of a grid cell is only dependent on its state at a previous time step, such a model is called a markov model, and is not considered a ca model. cellular automata models have one additional feature: the transition rules operate on cells based on the local neighborhood of those cells. for example, in a 2d grid, the state of a cell at time t+1 could be a function of the states of the cells to the north, south, east, and west of the cell of interest at time t.result and discussion transition probability matrixtransition probabilities express the likelihood that a pixel of a given class will change to any other class (or stay the same) in the next time period. for study area, the transition probability matrix of land covers have been presented in table1. in comparison with other classes, transition probability of class4 (barren lands and dry farming) state to other class states is higher. transition areas matrixthis expresses the total area (in cells) expected to change in the next time period. according to the transition areas matrix values (table2) is expected to the largest areas of transition in the study area related to transition of class2 and class4 to class1. so that, 8600 cells of class4 (about 774 hectares) and 4635 cells of class 2 (approximately 417 hectares) will change to class1 state. urban cellular automata and creating of suitability mapsin this study, another transition rules were designed using multi criteria evaluation and fuzzy membership functions. all layers that were obtained by applying these rules overlapped and suitability map was obtained for each class and then entered to the model. in relation to the neighborhood, von neumann neighborhood was used as a 5×5 proximity filter. according to output of cellular automata markov chain model, spatial expansion of city in the coming years will be almost similar to previous trend, and city mainly be expanded towards barren lands in the east of study area. conclusion results represent high efficiency of cellular automata markov chain in the urban spatial growth simulation. in the past three decades, development trend of maragheh city has been more towards barren lands. according to the output of the model, this trend will continue over the next 17 years. so that, the city will be expended due to the transition of barren lands cells state to urban cells state, and approximately 774 hectares from surrounding barren lands will be converted to urban lands. however, with continue of the previous trend, nearly 417 hectares of good agricultural lands will also change to urban lands.suggested that in order to maragheh city physical development prevent the agricultural lands destruction. especially the agricultural lands surrounding the city are mostly orchards, and have great importance in supplying the needs of city to the agricultural products, tourism, and reduce air pollution. in this context, it is desirable that city developed primarily in the form of town establishing in the south east of study area. however, it is necessary to consider the risk of flooding in these areas.