Evaluation of land-cover changes in the city of Shahrekord, and predicting its future status, using remote-sensing data and CA-Markov

Land use mapping using remote-sensing data is one of the newest and most widely used methods for the provision of land use map, and making a comparison between the existent usages. The aim of this study therefore, is checking the trend of urban land cover, based on Landsat satellite images of ETM and OIL/TIRSS 8, between 2003-2016. To this purpose, after the separation of Sharekord City's boundaries, and providing the training samples, the land use map, using supervised classification method with maximum likelihood, was drawn. Based on training samples, the accuracy of extracted maps was checked, using Error Matrix. Finally, the land's area of 2003 and 2016 were compared. Then, we reached the adaptation map of two classified image, using the difference of these two images. Then using the adaptation matrix, the amount of changes between the evaluated years was determined. The Kappa coefficient for land use in 2003 and 2016 was 82% and 93% respectively. Results shows that the urban areas have been raised to 1632 hectares in 2003, and to 2220 hectares in 2016. Whereas, the amount of green space has increased to 577 hectares in 2003, and to 664 hectares in 2016. Results also shows that the amount of urban area in 2016 has been increased by 13%, in comparison to 2003, while the green space area increase was 11.6%. Finally, the amount of change in urban area -using CA-Markov Model- was predicted to be increased by 6.01% in 2026 (the next 10 years). Therefore, the urban development planning requires the green space development, to be in accordance with urban expansion through this city.