Key points are not available for this paper at this time.
In this study, a preliminary test was implemented with the purpose of evaluating Sentinel-2 MultiSpectral Instrument (MSI) imagery for forest growing stock volume (GSV) estimation, using spectral bands and vegetation indices, in a heterogeneous Mediterranean forest in northeastern Greece. This evaluation scheme was complemented with the comparative use of a Landsat-8 Operational Land Imager (OLI) imagery acquired also on June 2016.All Sentinel-2 MSI bands but the blue and red parts of the spectrum recorded by the sensor, provided negative but significant correlation results with growing stock volume. The use of traditional vegetation indices did not improve the strength of the correlation. The Difference Vegetation Index (DVI), Enhanced Vegetation Index (EVI) and Perpendicular Vegetation Index (PVI) were found to have the strongest correlation for the retrieval of GSV. The replacement of the Near InfraRed (NIR) band in the indices formula with the first rededge band improved the magnitude of the relationship significantly.The information content of Sentinel-2 MSI imagery was also found to be slightly higher compared to the Landsat-8 OLI imagery based on the results of the Random Forest regression algorithm that was used to model the relationship between the spectral information inherent in both sensors and the GSV.
Chrysafis et al. (Wed,) studied this question.