Document Type : Research Paper

Authors

• Zahedeh Heidarizadi ¹
• Ali Mohammadian Behbahani ²

¹ PhD student of Combat Desertification, Arid Zone Management, Gorgan University of Agricultural Science and Natural Resources, Iran

² Assistant Professor, Department of Watershed and Arid Zone Management, Gorgan University of Agricultural Sciences and Natural Resources (GUASNR), Iran

Abstract

     Protecting an ecosystem and conserving natural resources requires knowledge of the conditions and land-use changes. The purpose of this study was to monitor land-use changes in the past and evaluate the performance of GEOMOD and LCM models in simulating land-use changes in order to select a more appropriate model for predicting land-use changes in the future. Landsat satellite images were used during the periods of 1990, 2003, and 2016 and land-use changes were monitored by using these images. The simulation of land use status by the LCM model for 2016 was done using the maps of the years 1990 and 2003. Using MLP and Markov chain, the land use map was simulated for 2016. To run the GEOMOD model, the image of the single-user map of 2003 and 2016 were used and the map of "appropriateness of changes" was made by the use of variables affecting land-use change and they were introduced into the model. The results of the accuracy of the simulation map of 2016 showed that LCM and GEOMOD had Kappa coefficients of 81% and 71%, respectively. Therefore, the LCM model was chosen as the most appropriate model and the map of the year 2029 was predicted and prepared.

Keywords

• MLP Method
• Markov Chain
• Landsat
• Land use
• Prediction