Interpreting RADARSAT-2 Quad-Polarization SAR Signatures From Rice Paddy Based on Experiments

The objective of this letter was to interpret the spatial variation of rice backscattering signatures as a function of rice growth parameters. Two scenes of RADARSAT-2 quad-polarization images were acquired at two rice growth stages. In accordance with the acquisition dates, a wide range of rice growth parameters, such as leaf area index (LAI), biomass, canopy height, and stem density, were measured. Among them, six parameters were selected as impact factors. The correlation between impact factors and rice backscattering coefficients was analyzed before establishing regression models. Because of strong multicollinearity among the impact factors, a principal component regression method was applied to build the models for different polarizations. Results showed that the spatial variation of rice backscattering is most sensitive to the change of rice biomass and LAI at both rice growth stages. Compared with HH and VV, VH or HV has a better correlation with the spatial change of biomass and LAI, implying the advantages of RADARSAT-2 quad-polarization data in regional rice growth monitoring.