کاوشهای جغرافیایی مناطق بیابانی (Aug 2019)
Prediction of Land Use Change and Determination of Urban Growth Patterns Using Markov Chain Model and Multi-Chip Satellite Images: A Case Study of Arak City
Abstract
The rapid changes in land use and land cover in the suburbs of large cities of Iran, including the city of Arak, as an industrial metropolitan city, have caused many problems in recent decades, including degradation of natural resources, environmental pollution and poor urban development. In most cases, this has been caused by an unjustified land use change. The study of changes and destruction of resources in past years, the feasibility and prediction of these changes and the pattern of urban growth in the coming years can be an important step in planning and optimal using of resources in the future. In this study, by using Landsat multispectral satellite imagery of 2000, 2018, and 2028, lateral information, Markov chain model, and artificial neural networks, the changes that have occurred in land use and land cover in Arak over the course of 28 years were investigated. To reveal the changes, the conversion rate for each course was obtained in four classes. To investigate the images, they were corrected, and then the best bands were selected for combining the bands. In the following, the classification of rates of maximum probability and evaluation of the accuracy of the classification as well as the overall accuracy were performed with kappa coefficient. Finally, by using the Markov forecasting model and the LCM modeling approach, land use change was predicted for the year 2028. The results indicated the excessive expansion of the built areas, severe degradation of agricultural lands during the period of 2000-2010, sharp decline of barren lands, their conversion to the lands built within the boundaries of the city, and the reduction of mountainous lands in the period of 2010-2018. Modeling the changes also showed that in the period of 2018-2028, the city will grow from the north and northeast to the south and southwest (i.e. toward the mountainous and garden lands of the two villages annexed to the city).
