Determination of stand volume with landsat and sentinel satellite images

In this study, "stand volume", one of the most important stand parameters, was tried to be determined by remote sensing methods. For this purpose, pure Anatolian Black pine, pure Turkish red pine and pure Oriental beech stands with in the borders of Amasya Regional Directorate of Forestry, Vezirköprü Forest Management Directorate were chosen as research object. In the study, Landsat 8 and Sentinel 2 satellite images were used as remote sensing data. Terrestrial data (ground control data) were obtained from field studies and local management plans. Stand volume was calculated in 371 sample plots with sizes ranging from 100 m2 to 900 m2. Pixel-based classification and band ratio methods have been aplied to satellite images. Band brightness values and various vegetation indices were obtained from satellite images. The relationship between the stand volume and the spectral data obtained from images was revealed and modeled by regression analysis. As a result of the supervised classification, Kappa statistics were determined as 71%, 77% and 71%, for Anatolian Black pine, Turkish Red pine and Oriental beech in Landsat 8 satellite image and 80%, 90% and 81% in Sentinel 2 satellite image. The most suitable volume model for Anatolian Black pine stands was obtained from the 20 m resolution data of Sentinel 2 satellite image, it was estimated by the (1/Bant6), (1/Bakı), (1/SR5), (CVI)2, LnEğim, LnBakı ve (1/SR2) variables, with a rate of 55%, with an error of 30.73 m3/ha. The most suitable volume model for Turkish Red pine stands was obtained from the 10 m resolution data of Sentinel 2 satellite image, it was estimated by the 1/DVI, 1/Bant34, (Bant35)2, (SR2)2 ve (Eğim)2 variables, with a rate of 61%, with an error of 26.70 m3/ha. Landsat 8 image gave the most suitable volume model for Oriental beech stands and it was estimated by the variables LnBakı, (SR3)2, (SR4)2, (1/Bant2) ve (1/GNDVI) with a ratio 54%, with an error of 58.97 m3/ha. It has been determined that there is a statistically significant relationship between independent variables and stand volume. Similar studies should be tried in different regions and for different tree species, it is thought that the success rates will be higher in studies using LIDAR, ICESat or Drone data. Key Words: Remote sensing, Forest management planning, Stand volume, Anatolian black pine, Turkish red pine, Oriental beech, Landsat 8, Sentinel 2