On the Potential of Polarimetric SAR Interferometry to Characterize the Biomass, Moisture and Structure of Agricultural Crops at L-, C- and X-Bands

Abstract

Polarimetric SAR Interferometry (Pol-InSAR) has shown great promise for estimating the height of agricultural crops through the inversion of a scattering model of the plant canopy and the soil. The inversion also provides estimates of model parameters describing the microwave attenuation within the canopy and the relative scattering contributions from canopy and soil surface. Here, we investigate how vegetation characteristics including biomass, water content (VWC) and canopy structure are related to these parameters and provide a first assessment of the potential of estimating such characteristics using Pol-InSAR time series in L-, C- and X-Bands. The overall attenuation for maize is positively related to total VWC in L- and C-Bands. Furthermore, larger attenuation in VV than HH points toward the existence of anisotropic propagation effects due to vertical orientation of the stalks. Conversely, for wheat in C- and X-Bands there is no consistent relation between attenuation loss and VWC. Rather, structural changes occurring within the plant growth cycle appear to have an appreciable polarization-dependent effect on the observed attenuation changes. In addition, the estimated normalized volume backscattering power NVP (a measure of the relative scattering contribution from the canopy compared to the underlying soil) is associated with wet biomass. However, the contrasting sign of this relation (negative for maize in L- and C-Bands; positive for wheat in C- and X-Bands) indicates again the role of crop structural properties in the Pol-InSAR measurements. For instance, the NVP for maize in L- and C-Bands appears to decrease with increasing biomass due to the increasingly important double bounce ground-stalk scattering contribution as plants become taller and thicker. Overall, these results indicate the sensitivity of the Pol-InSAR parameters to canopy structure and biomass; this sensitivity is however dependent, amongst others, on crop type and radar frequency. When choosing an appropriate baseline/frequency configuration, the Pol-InSAR attenuation loss and NVP may complement the information of the estimated crop height, especially if the latter shows very little variation over the plant growth cycle (e.g. as for wheat).