Land use changes predicition in Zagros forest areas based on Markov Chain Model

Today, due to increasing population growth and land use changes and its impact on the environment, monitoring and modeling land use changes is one of the main prerequisites for optimum use of land and achieving sustainable development. The purpose of this study was to evaluate and predict the spatial and temporal dynamics of land use in the county of Sardasht, in order to obtain basic information for planning in line with sustainable forest management.

 Data from the Landsat 7 images (+ETM) 2003, and Landsat 8 (OLI) 2015 were analyzed. The Maximum Likelihood algorithm has been used to mapping the land use for the years. The analysis of the change dynamics using traditional Markov Chain and Cellular Automata was predicted for the next 25 years.

The overall accuracy of classified images in 2003 and 2015 was 89% and 94%, respectively, and the Kappa coefficient was 0.87 and 0.92. The results showed that during the period, bout 7% of the forest area has decreased and the agricultural lands has increased by 72%. According to the Markov transmission probability matrix the classes most affected by these changes is the forests and rangelands that changed to agricultural and residential.

 Changes in the extent of forests until 2040 show that the area of forests will decrease continuously. The results of current study could provide quantitative information, which represents a base for assessing the sustainability in the management of Zagros forest ecosystems and for taking actions to mitigate degradation.