Modeling land use changes in Gahvareh region, Kermanshah province

Land use change is an important global and local ecological trend and is one of the major challenges in the 21st century. The purpose of this study was to model land use changes in the Gahvareh region, Kermanshah province using the LCM model. This model has the ability to simulate several land use changes by utilizing and integrating Markov chain models, multi-layer perceptron neural network, logistic regression, and MLOP. In this research, Landsat 4, Landsat 5, and Landsat 8 satellite images were used to prepare land cover maps. After initial corrections, the images were classified using the Maximum likelihood method, and land use maps for the years 1986, 2000, and 2018 were prepared. After monitoring the changes of different land uses during two periods (1986-2000) and (2000-2018) and assessing the validation of the model, finally, the land use map of 2028 was predicted using the LCM model. The results showed that during the first and second calibration periods, 30% and 42% of forest land were reduced, 26 and 37 ha were added to agricultural land and 80% and 32% were added to residential land respectively. The most changes during the study period were the conversion of forest lands to pasture and thin forest (80%). Forecasts for the year 2028 showed that the dense forest would be destroyed and would be decreased in half in comparison to 2018. Evaluation of the accuracy of transmission potential modeling using artificial neural networks showed high accuracy in most of the scenarios.