Current Measurements in European Coastal Waters and Rivers by Along-Track InSAR

Improved Method to Suppress Azimuth Ambiguity for Current Velocity Measurement in Coastal Waters Based on ATI-SAR Systems

Remote Sensing ◽

10.3390/rs12203288 ◽

2020 ◽

Vol 12 (20) ◽

pp. 3288

Author(s):

Na Yi ◽

Yijun He ◽

Baochang Liu

Keyword(s):

Coastal Waters ◽

Velocity Measurement ◽

Phase Fluctuation ◽

Ocean Surface ◽

Pollution Monitoring ◽

Speed Ratio ◽

Improved Method ◽

Eigenvalue Spectrum ◽

Practical Applications ◽

Along Track

Measurements of ocean surface currents in coastal waters are crucial for improving our understanding of tidal atlases, as well as for ecosystem and water pollution monitoring. This paper proposes an improved method for estimating the baseline-to-platform speed ratio (BPSR) for improving the current line-of-sight (LOS) velocity measurement accuracy in coastal waters with along-track interferometric synthetic aperture radar (ATI-SAR) based on eigenvalue spectrum entropy (EVSE) analysis. The estimation of BPSR utilizes the spaceborne along-track interferometry and considers the effects of a satellite orbit and an inaccurate baseline responsible for azimuth ambiguity in coastal waters. Unlike the existing methods, which often assume idealized rather than actual operating environments, the proposed approach considers the accuracy of BPSR, which is its key advantage applicable to many, even poorly designed, ATI-SAR systems. This is achieved through an alternate algorithm for the suppression of azimuth ambiguity and BPSR estimation based on an improved analysis of the eigenvalue spectrum entropy, which is an important parameter representing the mixability of unambiguous and ambiguous signals. The improvements include the consideration of a measurement of the heterogeneity of the scene, the corrections of coherence-inferred phase fluctuation (CPF), and the interferogram-derived phase variability (IPV); the last two variables are closely related to the determination of the EVSE threshold. Besides, the BPSR estimation also represents an improvement that has not been achieved in previous work of EVSE analysis. When the improved method is used on the simulated ocean-surface current LOS velocity data obtained from a coastal area, the root-mean-square error is less than 0.05 m/s. The other strengths of the proposed algorithm are adaptability, robustness, and a limited user input requirement. Most importantly, the method can be adopted for practical applications.

Download Full-text