Modeling Land Use Changes Based on Markov Chain in LCM(A Case Study of Ramhormoz)

Measuring is a key technology for assessing the extent and amount of land cover variations the knowledge of which is of fundamental importance for different planning. In this study, land use changes in the past 20 years using the TerrSat software and its potential for prediction in the future were evaluated using the Markov chain of Ramhormoz. For this purpose, TM Landsat 5, OLI Landsat 8 was used for 1996 and 2016, along with topographic maps and area coverage. Images of both time periods were classified into four categories of land use: lands made by human hands, arable land, agricultural and uncultivated lands. According to the results, the lands made with 39.1% of the most positive changes and lands with vegetation cover with the rate of -29.1% have the most negative changes in the region and the trend of forest areas has been a downward one. Then, using the prediction model and classified images (1996-2016), the 2016 coverage map was predicted using the Earth-change model. After evaluating the model, the overall accuracy was 83.09 and the Kappa coefficient was shown to be 0.79, which indicates the high adaptation between the predicted map and the classified map. With the introduction of the land cover plan for 2016, the model for land-changeover, a 2025 land cover map was prepared. The results showed that 102 hectares of agricultural land and 178 hectares of vegetation cover had been reduced.