Mapping of the distribution and canopy density of Northern Zagros forests using Sentinel-2 satellite images in the West Azerbaijan province, Iran

Since there are different statistics on the area, distribution, and density of Zagros forests, continuous monitoring and evaluation of these forests are facing implementation problems. The available statistics and information from Zagros forests are not effective in the management decisions of these forests due to many reasons such as their preparation in the traditional way or their oldness. In this research, a high-accuracy map of canopy density and distribution of Zagros forests as well as the determination of other land uses have been prepared using Sentinel-2 images of 2019 in the south of West Azarbaijan province. The image classifications were performed with the supervised method and ML and SVM algorithms in the ENVI 5.3 software environment. The Bing and Google Earth images were used to prepare the ground truth map to evaluate the accuracy of the output maps. In this research, the five non-forest land uses including ranges, gardens, agriculture, water sources, and barren and residential lands were classified. The results showed that the maximum likelihood algorithm with an overall accuracy of 87.3% and a kappa coefficient of 0.74 was the most accurate in preparing the canopy density map. The available statistics showed that the state of forest cover in the province is not in a favorable condition and the area of Zagros forest in the province is equal to 60200.55 ha, which is equivalent to about 90% of the updated statistics of the forests, rangelands, and watershed management organization of the country (67235,91 ha) in 2020. It can be stated that Sentinel-2 data has an acceptable efficiency in terms of accuracy and cost for preparing the canopy density and distribution map of forest areas and preparing the land use map.