Forest Change Modeling Using Logistic Regression in the Period 1988-2007 and Prediction of Future Condition in Gorgan Area

In this study، changes in forest cover in Gorgan area was determined for the period 1988-2007 and forest change modeling was performed using logistic regression model. To prepare forest change map، land use maps of 1988 and 1998 and 2007 were generated using satellite imagery classification. Then، change detection was performed using Post-Classification Comparison method in Idrisi software. Also، land use maps were used to prepare the independent variables. Results from the model showed that forest change has increased in the period of study. Validation of model results was performed using Pseudo-R2 and ROC values which were more than 0. 27and 0. 91 respectively. Then، the future condition of forest areas for the years 2016 and 2025 were predicted using the results from the logistic regression model. The results of this study can help managers and decision makers to monitor and prevent the destruction of forest areas. Output maps of Logistic regression model show high probability changes in land use and their location which are helpful for resource management towards efficient reduction and prevention of forest degradation in the study area.