مدل سازی و پیش بینی روند گسترش و توسعه ی فیزیکی شهر بجنورد

Extended Abstract
 Introduction:
One of the most important problems of the urban network in Iran is the rapid expansion of cities and, consequently, the uneven development and growth of cities, which has occurred for various reasons, including increasing population and irregular migration. The first consequence of urban sprawl is land use change. The city of Bojnord, after being selected as the political-administrative center of North Khorasan province in 2005, faced a double population growth and demographic changes. The concentration of administrative, military, tourist, cultural and educational centers, along with the uncontrolled migration of villagers to the city, confronted the city with new urban development needs. The physical expansion of the city more than ever, the conversion of agricultural land into a city and the uncontrolled construction, is one of the main issues facing the city. In this regard, the use of cellular automation, as a technique with features such as simplicity, transparency and strong potential to simulate spatial dynamics, has caused more and more attention in modeling the spatial information system and urban affairs.
 
 Methodology:
The present study was performed by descriptive-analytical method and using cellular automation method by perceptron artificial neural networks and remote sensing data. Thus, satellite images related to the years 2011 and 2021 of Bojnord plain were the basis of the study and in order to image the spatial-temporal patterns of land use, the data of the remote sensing archive were used. Research data from satellite images related to 2011 and 2021 (a period of ten years) as well as topographic maps of the plain and the city of Bojnord along with some spatial data of the region including land price, distance from land uses, distance from fault, slope, Height, distance from the road and distance from the built areas are provided. The source and organization of this data is also related to the Surveying Organization, Google Earth and the Surveying Organization of Iran. In line with the objectives of the research, the multispectral images of Landsat 7 and Landsat 8 in relation to the city of Bojnord (path 161, row 034) have been downloaded. In order to increase the validity of the results, the output maps were adapted to Google Earth maps. Also, in the 2011 classification, the overall accuracy was 86 and the kappa coefficient was 81, and in 2021, the overall accuracy was 88 and the kappa coefficient was 85.
 
 Results and discussion:
The main findings include calculating and estimating land use changes and the share of each land use in the base years of this research, modeling and forecasting land use changes for 2031 based on mapping the potential for transfers for land uses and forming a Markov chain along with eight affecting factors: slope, height, Distance from faults, agricultural lands, main roads, main electricity network, built-up areas and land prices. In 2011, the largest area of the region was occupied by pastures, which accounted for 73.6% of the total land area. During this period, the built-up areas cover 4.2% of the area with an area of 53748901 square meters. In general, most of the pastures are in the southern and eastern regions of the area and the areas built in the central and northern part of the Bojnord plain, as well as agricultural lands are located in the central and eastern parts. According to the data obtained in 2021, the built-up areas include 65266522 square meters, agriculture 194519833 square meters, pastures 888136607 square meters and barren 130533045 square meters. The share of rangelands in land use of Bojnord plain is 69.4%, barren 10.3%, agriculture 15.2% and developed areas 5.1%. Decreasing pasture and agricultural levels and increasing built and barren levels are evident this year. In modeling land use change, the transfer force from one land use to another is modeled according to the variables; In the sense that each pixel of the image has the potential to change from one user to another. According to the calculation maps, it can be seen that the changes in the built areas in the central part of Bojnord plain and around the city of Bojnord have been mainly accompanied by changes in agricultural lands and pastures. In all parts, especially in the northwestern areas of Bojnord plain, rangelands have become agricultural lands and with the development of agriculture, these valuable lands, which are the habitats of animals and plants of different species, have been destroyed; Also, Bojnord-Esfarayen axis has become the most built areas.
 
 Conclusion:
This study to model and predict land use change in Bojnord, based on the two years based on 2011 and 2021 AD, divided land use into four main categories: constructed use, barren, agricultural and rangeland. The main findings with emphasis on the effect of eight factors affecting the slope, height, distance from the fault, agricultural lands, main roads, main electricity network, built-up areas and land prices, on the development of Bojnord based on data received from satellite images and the process of image correction to increase the validity and accuracy of research outputs shows that while the rangeland use level between 2011 and 2021 from 73.6% of total land uses decreased to 69.4% and agricultural land use level from 17.4%  decreased to 15.2%,  built-in (urban) user level increased from 4.2% to 5.1%. This increase has often occurred due to the greater impact of slope, height, distance from main roads and distance from built-up areas, and the greatest proportion of this change has occurred in the axis of Bojnord Esfarayen road (west and southwest of the city). Based on the formation of the Markov chain matrix and the created user class maps in 2011 and 2021, land use forecasting and modeling of Bojnord in the horizon of 2031 has been done.The results of the land use change forecast section confirm the increase of built-up areas and urban use and the decrease of agricultural and rangeland use in the horizon of Bojnord city development forecast in 2031.