Microwave scattering (1-10 GHz) from a vertically heterogeneous grass canopy

This study concerns the effects of considering vertically heterogeneous canopy structure when modeling microwave scattering from grassland with the Tor Vergata (TVG) model, which uses the matrix doubling method (MDM). The TVG model was extended with the M-volume approach to accommodate height-dependent variation in structure for every scatterer type. Used approach was to reproduce 1 – 10 GHz backscatter for all linear polarization combinations from an alpine meadow measured by a ground-based scatterometer with both the default (1-volume) and the M-volume approach with 3 volumes. Measured in-situ vegetation parameters were used to constrain the model. We found that both models were able to reproduce the angle-dependent backscattering for C- and X-band, measured on two afternoons, and the 31-day average measured radar return power for L-, S-, C-, and X-band within, or close to, the measurement uncertainty. Our analysis proved inconclusive on whether the 1- or 3-volume approach worked better for the considered grassland, but did show that the 3-volume approach allows for more flexibility in reproducing the actual angle-dependent backscattering for multiple frequencies, a flexibility that may prove necessary when more scattering angles are considered. Furthermore, predictions of the bistatic scattering coefficient at higher frequencies (C- and X-band) were significantly different between both models. For X-band with hh polarization differences up to 3 dB were found for the specular direction. We conclude that considering vertical heterogeneity of vegetation canopy structure with MDM leads to significantly different results than with the vertically homogeneous canopy approximations typically used.