A Novel Multi-Candidate Multi-Correlation Coefficient Algorithm for GOCI-Derived Sea-Surface Current Vector with OSU Tidal Model

Abstract

The maximum cross-coefficient (MCC) algorithm based on the template matching technique is a typical algorithm for obtaining the sea-surface currents (SSCs) in marginal seas. However, this algorithm has mismatches between images in highly turbid water. In this study, we implemented the MCC algorithm to Geostationary Ocean Color Imager-derived total suspended matter to obtain the SSCs in the Yellow Sea and the East China Sea. We propose a novel vector optimization algorithm, which is combined with the accurate estimate of tidal ellipses from the OSU tidal model. This method considers the three greatest candidate acquisitions from multi-correlation coefficients as potential vectors. The rotation direction of the vector within the tidal oscillation is used to identify and substitute for the spurious vector. The obtained average speed of SSC reached 0.60 m/s, which was close to the buoy-measured average speed of 0.58 m/s. Compared with the existing spurious vector eliminating method, the average angular error was improved by 20%, and the average relative amplitude error was improved by 4% in our case study. On the basis of ensuring data integrity, the inversion accuracy was improved.

Keywords

• geostationary ocean color imager (GOCI)
• OSU tidal model
• multi-candidate multi-correlation coefficient algorithm
• sea-surface current inversion