A Preliminary Study of the Calibration for the Rotating Fan-Beam Scatterometer on CFOSAT

Abstract

The first rotating fan-beam scatterometer (RFSCAT) will be launched onboard the Chinese-French Oceanic Satellite (CFOSAT) in 2018. It provides a set of radar cross-section (σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> ) measurements at different azimuth/incidence angles over a wind vector cell (WVC), in order to determine the near-surface wind field using the backscatter model, i.e., the so-called geophysical model function (GMF). The accuracy of the retrieved wind vector is a sensitive function of the radiometric accuracy of the σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> measurements. Therefore, in-flight calibration, including the loop-back (internal) calibration and the external calibration performed with natural extended-area targets, is studied in this paper. Several homogeneous areas over land are first analyzed to check the stability and azimuthal dependence of the σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> over these areas. A new calibration mask of the homogeneous land areas is generated and will be used by RFSCAT calibration. Then a simple method of external calibration is proposed to eliminate the azimuthal-dependent σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> errors induced by the insertion loss of the rotating joint, which can be applied to both the rotating pencil-beam scatterometers and the coming RFSCAT. The “observed” σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> of RFSCAT is simulated using the SeasatA scatterometer (SASS) measurements and the “perturbed” azimuthal-dependent σ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> errors. The latter is then tracked by the proposed external calibration. The results show that the accuracy of gain corrections is up to 0.2 dB, ensuring consistency between different azimuthal measurements.