Using dual-polarization sentinel-1A for mapping vegetation types in Daklak, Vietnam

Sentinel-1A is a microwave satellite of European Space Agency (ESA) and was successfully launched in 2014. This satellite supports SAR (Synthetic Aperture Radar) data with C band, spatial resolution 10m, a cycle 12 days and free of charge. Sentinel-1A provide dual-polarization SAR image included in VV polarization and VH polarization data. Backscatter signal depends on the polarization, surface roughness of the objects and soil moisture. In this paper, the authors present the experience results using combinations of dual-polarization of Sentinel-1A images for mapping vegetation species. The combination of VV polarization and VH polarization used in this study included in |VV + VH|, |VV - VH|, VV |VH , RVI (Radar Vegetation Index) and PCA (Principal Component Analysis). The typical vegetations studied in this paper such as paddy rice, industrial trees, deciduous forest and evergreen forest. The classification SVM (Support Vector Machine) method was used in our study. The overall accuracy achieved 90.72% with Kappa index 0.8825. The study area is Ea Sup district, Dak Lak province in the Central Highlands region of Vietnam. © 2020 40th Asian Conference on Remote Sensing, ACRS 2019: "Progress of Remote Sensing Technology for Smart Future". All rights reserved.