Estimation of aboveground biomass of arboreal species in the semi-arid region of Brazil using SAR (synthetic aperture radar) images

Use of synthetic aperture radar data for the determination of normalized difference vegetation index and normalized difference water index

“…Using a deep-learning model (U-Net), Roßberg and Schmitt 15 observed a greater relative importance of VH polarization compared to VV polarization for the estimation of the NDVI in an area near the city of Gaoua in southern Burkina Faso. In a study developed by Jesus et al, 11 the authors observed that the variables that most contributed to an LR model for the estimation of aboveground biomass (AGB) in the semi-arid region of Brazil were the relationships between VH/VV polarizations, as well as DPSVI, H, and co-polarized VV. Ayari et al 13 noted that to determine the NDVI with RF in wheat fields in the Kairouan plain in central Tunisia, the VV and normalized cross-polarization ratio (IN) polarizations presented significant contributions.…”

“…The results indicate that horizontal-horizontal (HH) and verticalvertical (VV) cross-polarization can contribute significantly to the determination of these attributes. 3,8 Moreover, promising results were observed recently by Filgueiras et al, 9 Mandal et al, 6 Bai et al, 10 Bahrami et al, 1 Jesus et al, 11 Lu et al, 12 and Chen et al 2 when using Sentinel-1 SAR data for the determination of vegetation indices.…”

“…Based on SAR data, vegetation indices are determined using various approaches, such as physical and empirical models and data assimilation techniques. Empirical models are the most adopted and, in this method, the backscattering coefficient is related to vegetation indices through regression equations, such as linear regression (LR), 10,11 random forest (RF), 13 support vector machine (SVM), 14 and deep-learning model. 15 In all these studies, the results obtained are promising, indicating that SAR data can be used as a complement to data from optical sensors for crop monitoring.…”