Leaf area index is defined as the total area of one side of plant leaves per unit ground area in broadleaf canopies. The leaf Area Index (LAI), one of the most important structural parameters of forest ecosystems, is highly related to the forest dynamic and forest biological processes i.e., the photosynthesis activity, evapotranspiration, the Net Primary Production (NPP), energy and carbon exchange rate between the vegetation and the atmosphere. Remote sensing methods can compensate for these limitations of direct methods and other indirect terrestrial methods and estimate leaf area index at different scales and at large levels. In this study, the leaf area index was estimated using Sentinel-2 satellite imagery over a small part of the Baneh forests. A digital camera with a fish eye lens was used to collect the hemispherical photographs in 58 field reference plots with a size of 20m × 20m. The requiered digital image processing procedures were applied on the remote sensing data, and various vegetation indices were also calculated. Elevation, slope, and aspect maps were also used as an ancillary data. Spectral and non-spectral values were extracted from satellite imageries and ancillary data in each sample plot. Our results showed that the Red band and TNDVI (Transformed Normalized Difference Vegetation Index) have the highest correlation with LAI. The results of the regression analysis showed that considering only original spectral band as independent variable, a model based on the red and the near-infrared bands achieved the highest accuracy (R2= 0.753, RMSE= 22%). Considering a combination of original spectral bands, vegetation indices and non-spectral variables, a model based on TNDVI and DEM produced the highest accuracy (R2= 0.781, RMSE= 20%). The results indicated that the leaf area index in the Zagros forests can be estimated using remote sensing data with acceptable accuracy.
