Journalpaper

Temporal and spatial statistics of travelling eddy variability in the South China Sea

Abstract

The variability across decades of years of migrating eddy activities in the South China Sea (SCS) have not yet been documented. We employ a daily global eddy-resolving (0.1 degree) model product called STORM that covers a period of 1950-2010 to fill this gap. The frequency and pattern of eddy occurrence in the simulation is broadly consistent with satellite-based (AVISO) data. On average, annually 28 anticyclonic travelling eddy (AE) tracks and 54 cyclonic travelling eddy (CE) tracks with long travel lengths were derived from the discrete sea surface height anomaly fields of STORM. Eddy centers most frequently pass by the Luzon Strait and along the continental slope in the northern SCS to the Vietnam coast. The lifespans range from 6 days to 240 days for AEs and to 293 days for CEs, and the longest travel lengths are 1941 km and 1988 km, respectively. EOFs of the spatial fields of eddy diameter (ED), eddy intensity (EI) and eddy number (EN) show almost white eigenvalue spectra, when calculated on the model’s 0.1-degree grid, but when the data are coarsened to grids with 1-degree and 2-degree grid spacing, meaningful structures emerge. EI and ED are highly correlated on both seasonal and interannual time scales. In general, CEs are much more active than AEs, but the AEs with high intensities or large diameters are more frequent than similar CEs. The monthly ED, EI and EN exhibit annual cycles, which are, however, not very stable. The variabilities of annual means of ED, EI and EN is large at interannual time scales, little at interdecadal sales and exhibits hardly a trend. The sizes and intensities of eddies in the SCS are hardly connected to the ENSO-variability in the tropical Pacific. The EOFs, the weakness of the annual cycle stability and the absence of a correlation with ENSO, point to a massive presence of internal variability (as opposed to variability provoked by large-scale drivers).
QR Code: Link to publication