| Based on the 2018—2022 Archiving, Validation and Interpretation of Satellite Oceanographic (AVISO) satellite altimeter
fusion products, this paper identifies mesoscale eddies in the South China Sea and the Northwest Pacific Ocean by applying the Okubo-
Weiss (OW) parameter method, the closed contour method, and the combination of sea level anomaly (SLA) extremum and closed contour method, and then compares and analyzes the detection ability of different eddy identification algorithms for mesoscale eddies. An analysis reveals an issue of excessive eddy identification and segmentation phenomenon when applying the OW parameter method regarding the count of detected eddies. The algorithms characteristic is confined to the core region of the eddy, resulting in underestimated measurements of eddy radius, area, and amplitude. The application of the closed contour method yields favorable outcomes. However, it fails to distinguish between single-core and multi-core eddies, nor does it recognize the internal structure of the eddies. Moreover, the application of the SLA extremum and closed contour method confines itself to the recognition of single -core eddies. It also fails to distinguish between multi-core eddies or the internal structure of the eddies. Therefore, to address the above problems of the combination of SLA extremum and closed contour method, this paper searches both the innermost closed contour and the outermost closed contour simultaneously by removing the inclusion of only one extremum as a restriction on finding the outermost contour and reducing the search increase. The improved algorithm not only recognizes multicore eddies, but also identifies both the external and internal structures of eddies, and effectively improves the problem of eddy omission in the recognition process of the combined SLA extremum and closed contour method. |
