Motion bias analysis of sea surface height measurement by near-nadir interferometric Synthetic Aperture Radar (SAR) and its compensation method research

Abstract

ABSTRACT High-precision observation of ocean surface topography is of great significance in studying the ocean mesoscale and sub-mesoscale processes. Spaceborne near-nadir interferometric synthetic aperture radar (InSAR) is capable of measuring sea surface height with centimetre-level accuracy and over a wide swath. It differs from terrestrial height measurement because the constant movement of the dynamic ocean surface creates biases in synthetic aperture radar imaging and leads to height measurement inaccuracy. This study investigates the bias due to the motion of ocean surface waves and establishes a theoretical model of the height measurement motion bias related to the ocean wave spectrums and InSAR system parameters. The motion bias model was simulated in different SAR modes, including Stripmap and Terrain Observation by Progressive Scans (TOPS) modes in various oceanic states. The results showed that the motion biases are smaller in Stripmap mode than TOPS mode under the same oceanic conditions. By adopting the total zero steering technology in Stripmap mode, the motion bias can be significantly decreased. In the same oceanic state, the motion bias increases with increasing angle of incidence. In the TOPS mode, the Doppler centroid frequency of the antenna’s signal changes linearly with the azimuth time, and thus, the biases change significantly in the azimuth direction. The motion bias model is validated by the interferometric full-link experimental simulation and the results are consistent with the theoretical values. This study provides a compensation method with millimetre level accuracy to compensate for the height measurement bias caused by sea surface motion. The theoretical derivation of motion bias models and simulation results provide feasible suggestions for the design of InSAR systems and bias compensation for sea surface height measurement.