Thermal fronts of the southern South China Sea from satellite and in situ data

The major goal of this study was to find match-ups between thermal fronts mapped from satellite sea surface temperature (SST) imagery and from in situ data in the southern South China Sea (SSCS), using 11 ship surveys conducted by the South China Sea Institute of Oceanology (SCSIO) between 1987 and 1999. Fronts were automatically detected by the Cayula-Cornillon multi-image edge detection algorithm (CCA) applied to satellite-derived maps of the Advanced Very High Resolution Radiometer (AVHRR) SST obtained from the Pathfinder project (8364 twice-daily global fields with 9 km resolution between 1985 and 1996). Twice-daily near-instant frontal maps were composited without any averaging or smoothing to produce individual monthly frontal maps covering the period from January 1985 through December 1996 (144 maps in total). Although the SSCS is a tropical sea with little SST difference between water masses, the CCA turned out to be an effective tool for front mapping in the SSCS. Out of the 11 ship surveys analysed in this study, four surveys produced satisfactory match-ups. The percentage of match-ups is considered reasonably high given that (1) ship surveys were not optimized to cross fronts, therefore most in situ sections missed fronts; (2) satellite measurements of SST with AVHRR are hampered by cloudiness, therefore satellite-derived frontal maps might miss some fronts masked by persistent cloudiness. Fronts are more distinct in winter, when cross-frontal SST gradients are enhanced. From oceanographic vertical sections and horizontal maps, fronts are much sharper in the subsurface layer (represented here by 50 m level). Nonetheless, the CCA successfully detected SST fronts with a cross-frontal step as small as 1 degrees C.