Annual Grass Biomass Mapping with Landsat-8 and Sentinel-2 Data Over Kruger National Park, South Africa

摘要: This study explores the potential of Landsat-8 and Sentinel-2 imagery for annual grass biomass mapping in savannas. To this end, three wet season image mosaics based on were created 2016, 2017 2018 over Kruger National Park (KNP), South Africa. For purpose calibration validation, use was made situ fuel values measured as part yearly veld condition assessment (VCA) KNP. The satellite reference data fed into a random forests machine learning approach to make park-wide predictions assess performance predictors (i.e., surface reflectance normalized difference vegetation index, NDVI). Examples sets used maps produced are provided together with obtained error statistics. latter suggest that NDVI from enable creation fairly reliable, These new estimates represent slight improvement recent efforts Sentinel-1 [1].