Using Geostationary Satellite Ocean Color Data to Map Diurnal Hourly Velocity Field Changes in Oceanic Mesoscale Eddy

The mesoscale eddy structure in the upper ocean is difficult to observe with altimeter satellite data due to their coarse spatiotemporal resolution. However, the world’s first Geostationary Ocean Color Imager (GOCI) makes it possible to refine the eddy structure inversion with observations of high spatiotemporal resolution (500 m and 1 h, respectively). In this study, we proposed a new flow field retrieval algorithm recurrent all-pairs field transforms in mesoscale eddy flow retrieval (RAFT-MesoEddy) based on deep current optical flow architecture in the mesoscale eddy surface flow field for GOCI data. The validation results showed that the flow field matched the particle image velocimetry (PIV) data in both quality and spatial distribution, with mean end-to-end point error (MEPE) and mean average angle error (MAAE) of less than 0.07 pixel and 2.57°, respectively. The GOCI-derived flow field showed good consistency with drifting buoy trajectory data, with MEPE and MAAE values less than 0.34 m/s and 18.78°, respectively. The GOCI-retrieved flow field showed agreement with satellite altimeter geostrophic flow data, with the MEPE and MAAE being 0.28 m/s and 10.12°, respectively. The GOCI-retrieved flow field showed an agreement with Ocean Surface Current Analyses Real-time (OSCAR) surface flow field, with the MEPE and MAAE are 0.28 m/s and 9.62°, respectively. In addition, a specific mesoscale eddy to discuss the hourly dynamics of flow field and eddy kinetic energy (EKE) is taken. The significant diurnal dynamics revealed in the hourly GOCI observations suggest caution in mapping mesoscale eddy flow using conventional altimeter satellite data, as the temporal resolution is insufficient to capture diurnal variations.