Predicting of Land Use Changes for 2030 By Using Remote Sensing and Landsat Multi-Timed Images (Case study: Mashhad city)

By predicting land use changes, the extent of the expansion and destruction of resources can be determined and future policies can be pushed in the right direction. In this study, the aim is modeling the land use changes process in Mashhad by using Landsat satellite images related to 1989, 2008, and 2014. Land uses were classified into six classes based on unsupervised classification. Then, the transfer matrix between 1989 and 2008 was computed to validate the Markov-CA model by using the 2008 land use map and by using the Markov-CA model, the 2014 land use map was predicted. Eventually, it was compared with the 2014 land use map with a Kappa coefficient of 0.91 that confirmed the model. Finally, the same model was used to mapping 2030 land use. Therefore, by entering the 2014 reference map as the base map, the 2030 land use map prediction map was extracted. The results showed that during the years 1998 to 2030 there will be an increasing trend in urban and arid lands and a decreasing trend in agricultural lands and gardens. The results clearly indicate that the Markov-CA model can contribute to the design of a sustainable urban system.