Potential of High-Resolution X- and C-Band SAR Sensors for Assessment of Biophysical Variables in Paddy Rice.

The objectives of this study were 1) to investigate the characteristics of high-resolution satellite SAR (synthetic aperture radar) sensors in X- and C-bands, and 2) to explore their unique capabilities in the assessment of key canopy variables in paddy rice. The X- and C-band SAR images were obtained by Cosmo-SkyMed, TerraSAR-X, and Radarsat-2 during the growing seasons of 2009-2012 in one of the major rice growing plains in Japan. All images were taken at the spotlight mode to ensure high spatial-resolution (1-2 m). Correlation analysis between backscattering coefficients (σ⁰) and biophysical variables showed clear and interesting results. In all SAR sensors, σ⁰ in rice canopies and nearby water surfaces were stable and consistent, although some obvious systematic shift was found between the X-band sensors. The X-band σ⁰ was not very sensitive to most of the other structural and morphological variables such as stem density and leaf size, but the σ⁰ during the maturing stage was best correlated with the panicle biomass. Results suggested a potential capability of X-band SAR for direct assessment of rice grain yield at regional scales from the space. The C-band σ⁰ had close positive relationships with the major volume-related variables such as LAI, plant height, and biomass. Structural variables such as leaf density and leaf size were moderately related to σ⁰. The response of σ⁰ to canopy biomass was expressed by an exponential equation with a high coefficient of determination, suggesting that the C-band σ⁰ may be able to detect the growth difference during the early stages. The close relationship of fAPAR with σ⁰ may be useful as a surrogate of the frequently-used NDVI-fAPAR relationship. SAR sensors with high spatial resolution would be useful for timely monitoring of biophysical variables, especially in monsoon regions, owing to their capability of all-weather observation.