نوع مقاله : مقاله پژوهشی

نویسندگان

• شادمان درویشی ¹
• کریم سلیمانی ²
• مرتضی شعبانی ³

¹ دانش آموخته کارشناسی ارشد سنجش از دور و سیستم اطلاعات جغرافیایی، دانشکده علوم محیطی، مؤسسه آموزش عالی هراز، آمل

² استاد گروه آبخیزداری، دانشکده منابع طبیعی، دانشگاه علوم کشاورزی و منابع طبیعی ساری

³ دکتری جغرافیا و برنامه ریزی شهری، دانشکده علوم انسانی، دانشگاه علوم کشاورزی و منابع طبیعی ساری

چکیده

امروزه رشد مناطق شهری و تأثیر آن بر کاربری اراضی در جهان و بخصوص در کشورهای درحال‌ توسعه به یک مسئله مهم زیست ‌محیطی در علوم محیطی و برنامه‎ ریزی شهری تبدیل ‌شده است. هدف مطالعه حاضر تحلیل و پیش‌بینی رشد شهرهای بانه، مریوان و سقز بر کاربری‌ اراضی در یک دوره‌ی 45 سال (1987 تا 2032) با استفاده از مدل CA-Markov است. در مطالعه حاضر بررسی و مدل‌سازی نقشه‎های کاربری اراضی و رشد نواحی شهری با استفاده از تصاویر لندست و با اعمال الگوریتم Maximum Likelihood  و مدل CA-Markov در نرم‌افزارهای ENVI5.3 و IDRISI TerrSet انجام شد. نتایج این مطالعه با دقت بالای 80 درصد نشان می‌دهد در دوره‎ی 1987 تا 2017  حدود  897/39 و 790/38 هکتار (در شهر مریوان) و 801 و 772/29 هکتار (در شهر بانه) به ترتیب از اراضی پوشش‌گیاهی و زمین‌های کشاورزی به نواحی شهری تبدیل‌ شده است اما این روند برای شهر سقز کمتر بوده و رشد این شهر بیشتر نواحی بایر را به میزان 1249/29 هکتار تحت تأثیر قرار داده‌است.  همچنین نتایج مدل CA-Markov برای سال 2032 نشان داد در دوره‌ی 2017 تا 2032 همانند دوره‌ی قبل رشد شهر بانه در جهت شمال شرق، شهر مریوان در جهت شرق و شهر سقز در چهار جهت خواهد بود. در این دوره رشد نواحی شهری منجر به تخریب 511/29 و722/70 هکتار (در شهر بانه) و 1080 و2402/01 هکتار (در شهر مریوان) به ترتیب از اراضی زمین‎های کشاورزی و نواحی بایر خواهد شد. از طرف دیگر رشد شهر سقز بیشتر اراضی بایر را به میزان 1511/46 هکتار تخریب خواهد کرد. بدیهی است یافته‌های این مطالعه نقش مؤثری در برنامه‌ریزی‌های آینده دارد چراکه با آگاهی از روند رشد این نواحی می‌توان جهات توسعه شهر را به جهات بهینه هدایت نمود و تخریب اراضی ناشی از رشد شهر و درنتیجه تأثیرات منفی تغییرات کاربری اراضی را به حداقل رساند.

کلیدواژه‌ها

• رشد شهر
• زنجیره مارکف
• سلول خودکار
• حداکثر احتمال
• بانه
• مریوان
• سقز

20.1001.1.25883860.1399.29.114.10.4

عنوان مقاله [English]

Analysis and prediction of urban growth and its impact on land use using remote sensing and CA-Markov ; Case study: Marivan, Baneh and Saqqez cities

نویسندگان [English]

• SHadman Darvishi ¹
• Karim Solaimani ²
• Morteza Shabani ³

¹ MSc. Graduated of remote sensing and Geographic Information Systems, Department of Environmental Sciences, Haraz Institute of Higher Education

² Professor of watershed management, Department of natural resources, Sari agricultural sciences and Natural Resources University

³ Ph.D. in Geography and urban planning, Sari agricultural sciences and natural resources University, Sari, Iran

چکیده [English]

Extended Abstract
Introduction
 Urbanization is a continuous process and the spatial patternsof urban growth havealways played an important role in the transformation of human life throughout history. Urban growth has two dimensions: demographic and spatial, meaning that with increased urban population, the need for shelter increases and cities are faced with spatial growth. Expansion of cities in the spatial dimensions have several consequences,including changes in land use and land covers of areas surrounding cities.Land use change is currentlyone of the major concerns ofthe environmental approach, since land use changes in areas surrounding cities have led to changes in the economic structure of cities and the destruction of vegetation and agricultural lands as one of the main foundations of production in these areas. They have also seriously damaged other water resources, wildlife habitats, and resulted in the reduction of soil organic matter, changes in soil humidity and saltiness, increased energy consumption, increased urban heat islands, climate changes, as well as negative effects on the mental and physical health of urban residents. Nowadays, rapid growth in remote sensing technology and geographic information system, as well as the advancements in computer science and its application in environmental sciences and urban planning have created spatial modeling techniques such as Markov chain, Cellular Automata, intelligent neural networks and statistical models. Due to its dynamic nature, the capability of showing spatial distribution of land use changes, as well as its unique characteristics in modeling of natural and physical geographic featureson the ground and simpler adaptation with remote sensing data and GIS, a combination of Markov chain model and Cellular Automata are used as an important supporting toolfor decision making in urban planning and environmental sciences in many studies performedrecently.  Over the past few decades, the population of Iranhas increased from 27 million in 1955 to 79 million in 2016. And according to the 2016census, 74 percent of the population lives in urban areas. In recent years, the population of Kurdistan province has experienced a 1.42% (2011 to 2016)average annual growth rate (especially in Baneh, Marivan and Saghez), which isaround 0.18% more than the average annual growth rate of the country (1.24%). Investigating census data shows that Baneh, Marivan and Saqezhave experienced a higher urban growth rate as compared to other cities in the province, and thus monitoring this growth and predicting its negative effects on the surrounding land use seems crucial.Destruction of vegetation and agricultural lands not only results in climate change, but also directly affect the lives of residents in the region. Therefore, understanding the growth rate is necessary for properplanning and managementofthese areas.
 
Materials and Methodology
Images received from Landsat in 1987, 2002 and 2017 were downloaded from the US Geological Surveywebsite and used in the present study. Google Earth images, land useand topography maps, and ground control points (GCP) were also used to perform imagepreprocessing, classification operations, and accuracy assessment. The study area includesBaneh, Marivan and Saqqez cities, which have recently experienced a high level of population growth. Considering the impact of population growth on increased rate of construction and physical development of urban areas, it is therefore necessary to study urban growth. In order to reduce the city’s impact on land use in future, it is necessary to modelurban growth. Using these models, planners can guide the urban development back to the optimal and appropriate routes and minimize the destruction of the land use.Image pre-processing in the present research was performed in ENVI5.3 environment. Then, using Maximum Likelihood algorithm, the images were categorized into five classes of water, residential areas, vegetation, agriculture and open spaces. Then, the overall accuracy of the classification maps was assessed using ground control points. To predict the urban growth, CA-Markov model was used in the IDRISI TerrSet software.
 
Results and Discussion
Findings indicate that the classified images have an accuracy of above 80%, and thus, land use maps of the study areas are valid.Investigations shows that the growth inMarivan and Baneh has most severely affected vegetation and agricultural land use. In the time period of 1987 to 2017, 897. 39 and 801 hectares of vegetation in Marivan and Banehhave been transformed into urban areas, respectively.During the same time period, 790.38 hectares of agricultural land in Marivan and 772.29 hectaresinBanehhave changed into urban areas. It is also important to note that unlike Saqez, the degradation of vegetation and agricultural lands in Marivan and Banehwas more severe than bare lands. In other words, bare landsinSaqez were more severely affected (as compared to vegetation and agricultural land), and about 1249,29 hectares of bare lands have turned into urban areas, while only 121.50 hectares of vegetation, and 509.04 hectaresof agriculture lands haveexperienced such a change.Also, results of the CA-Markov model showed that the growth of Baneh and Marivan cities in the 2017-2032 period will be in the Northeast and East directions, respectively. Results also indicate that this urban growth will affect agricultural and bare landsmore significantly. It is predicted that about 511.29 hectares of agricultural lands and 722.70 hectares of bare lands (in Baneh city) and 1080 hectares of agricultural lands and 2402.101 hectares of bare lands (in Marivan city) will turn into urban areas in this time period.
 
Conclusion
Based on the findings, it can be concluded that planning urban growth inthe study areas should be performed in a way that vegetation and especially the surrounding agricultural lands are preserved, and the negative effects of land use changesare minimized. Also,plannerscan apply the results of the present study in their future plansto guide the development of Baneh, Marivan and Saqeztoward optimal ways and reduce land use degradation.

کلیدواژه‌ها [English]

• Urban growth
• Markov Chain
• Cellular Automata
• Maximum likelihood
• Baneh
• Marivan
• Saqqez