Open AccessDiscrimination of forest species using medium spatial resolution images
Author(s) -
Carla Talita Pertille,
Marcos Felipe Nicoletti
Publication year - 2021
Publication title -
advances in forestry science
Language(s) - English
Resource type - Journals
eISSN - 2359-6570
pISSN - 2357-8181
DOI - 10.34062/afs.v8i3.12525
Subject(s) - cunninghamia , remote sensing , normalized difference vegetation index , vegetation (pathology) , eucalyptus , environmental science , geography , forestry , cartography , physical geography , ecology , biology , botany , climate change , medicine , pathology
This research aimed to evaluate the potential of orbital images from the Landsat-8/OLI and Sentinel-2 /MSI sensors in the distinction of species from a forest stand located in Campo Belo do Sul, State of Santa Catarina, Brazil. A total of 53 plots were allocated in the field, in which the central coordinate of the plot was collected using GPS receivers. In SIG environment, two images were used, one from each sensor, closely dated to the field campaign and with no clouds and other atmospheric factors. Then, the images were processed, and 17 vegetation indexes were calculated for each one. The indices were compared statistically by the t-Student test for independent samples. The indices that provided the best species differentiation were: CRI, GNDVI, NDI11, NDI12, NDVI, RDVI, SAVI, and SR. In addition, the species with greater prominence in the Landsat-8/OLI images was Eucalyptus spp. whereas Cunninghamia lanceolata (Lamb.) Hooker was easily distinguished in Sentinel-2 images. It was possible to differentiate the species from remote data derived from the Sentinel-2/MSI and Landsat-8/OLI sensors. However, further studies using other Remote Sensing data sources and other species are suggested.