Detection and prediction of forest level changes in Guilan province using satellite images and geomod model

Following the rapid economic and social development in recent decades, human activity to use natural resources has been reflected in the form of infrastructure and agricultural activities. This has severely affected forests as an important ecosystem which are considered potential environmental resources for future evolution. The purpose of this study is to detect changes of Guilan province forest levels during a period of 20 years (1996-2016), also modeling and predict these changes for the next 15 years using the geomod model. Landsat TM and OLI sensor images were used to prepare land use maps for 1996, 2006, and 2016 periods. Satellite images were classified into forest and non-forest classes using the maximum likelihood method and multiple educational samples. The geomod model was simulated based on the changes made in the period 1996-2006, changes in forest cover using the variables of height, slope, direction, distance from residential, distance from a road, distance from forest, with implementation for 2016. The predicted validation results of the forest cover map in 2016 is indicator the overall accuracy and value of the kappa index equal to 94.19% and 0.9159, respectively. Based on the results of detecting changes during the study period (1996-2016), 1054.97 hectares of forest area in Guilan province has been reduced and with the continuation of this trend and stable conditions in the next 15 years until 2031, another 871 hectares will be reduced from its level. Given the importance role of Hyrcanian forests, it is necessary to conduct multi-time studies to monitor and detect changes. Obviously, the information from such studies can be used in managerial and strategic planning.