Evaluation and Prediction of Changes in Vegetation Using Landscape Metrics and Markov Model (Case Study: Hamadan)

Landscape degradation and land use / land cover changes are constantly putting pressure on the environment of the country. Therefore, vegetation maps are very important in the production of information for urban planning. In this study, vegetation maps of Hamadan were prepared based on data obtained from Landsat ETM+ and OLI sensors, and then NDVI index value was calculated. NDVI is one of the most widely used indicators for monitoring vegetation changes, Which is achieved by red and infrared banding. In order to investigate qualitative changes in vegetation greenness, the numerical values of these index were categorized to 4 different classes of lands with excellent, very good, good and poor vegetation, and were applied to detect changes in classified maps using LCM model. Then in order to analyze landscape changes, 5 different metrics in class level, and 6 different ones in landscape level were considered. Predicting changes in vegetation cover in the next 15 years was done with Markov chain. Analysis of landscape metrics indicates that very good coverage declined from 12 percent in 2001 to 21 percent in 2016. Good vegetation cover, regarded as natural vegetation was also declined in the same period. Results obtained from Shannon diversity index in the landscape level, shows the values higher than one in the study area, which indicates high rate of diversity in the Hamadan landscape. Predicted results indicate that the most likely damage up to horizon of 2031, would take place on lands with excellent coverage (0.8329).