Deriving Ocean Surface Currents: A method using Geostationary Ocean Color Imager hourly data

Abstract

Tracking the movement of natural surface features across the time intervals between successive images has gained widespread acceptance for mapping ocean surface currents (OSCs). The 500-m and 1-h spatiotemporal resolutions of the Geostationary Ocean Color Imager (GOCI), launched in 2010, are ideal for observing the dynamics of mesoscale eddies and the diurnal changes dominated by tides. All reported works to date, however, are limited to a few occasions when there were cloudless skies and pixel-level accuracy. As a result, the abundant GOCI hourly data have not been put into an operational service to derive OSCs. This article revisits ocean feature detection and tracking techniques to discuss the development of a satellite image-matching system (SIMS). The results show that a SIMS can detect and track features from two consecutive GOCI images to approximately a 0.25-pixel level of accuracy in the presence of variable gaps due to land and clouds. Suspicious vectors can be ruled out by initially using size-irrelevant filtering and majority filtering methods with minimum preset thresholds. The products of pathlines and streamlines derived from GOCI hourly data make it possible to gain a better understanding of OSCs, both qualitatively and quantitatively.