Predicting land use changes with emphasis on residential lands using CA-Markov model Case study (Bojnourd plain catchment)

Understanding temporal and spatial changes in land use is essential for decision makers and community planners. Land use requires knowledge of the current trend and forecasting future developments in land use and land cover. In this study, using Landsat 7, 8 satellite images and Ca-Markov model in EDRISI TerrSet software, simulation and prediction of land use changes in Bojnourd catchment area in North Khorasan province has been performed. After making atmospheric and geometric corrections on the images of 2001 and 2019, a map predicting land use changes has been produced for 2040.The validation of the model is done through the kappa coefficient, the value of which is 0.92 for the land use map of 2001 and 0.95 for 2019. The results of the model prediction show that in the study area, residential lands with the increase of more than 5 thousand hectares during the study period have the most changes. Also, most of the changes have been made around the city of Bojnourd.