Evaluation of Satellite-Derived Sea Surface Temperature (SST) Using Ocean Glider Data in the Agulhas Current System

Level 4 (L4) Sea Surface Temperature (SST) satellite products are appealing for a large number of users in the marine science community as they provide gap-free merged observations in near real time which are easy to use and can be assimilated in ocean and atmospheric operational models. Few validation studies of these products exist, particularly in regions where horizontal temperature and velocity gradients are high. In this study, in-situ sea temperature data collected from underwater gliders at the inshore front of the Agulhas Current are compared with SST data from three L4 SST products: the Multi-scale Ultra-high Resolution (MUR) Sea Surface Temperature analysis, the Operational Sea Surface Temperature and Sea Ice Analysis (OSTIA) and the microwave and infrared data (MW_IR) Remote Sensing System (REMSS) product, to gain some insight on the usefulness of these products to track and monitor ocean variability in western boundary currents Results show that all three products match glider measurements well, with correlation coefficients ≥ $0.85$. The L4 SST products which performs best at the inshore boundary of the Agulhas Current is the REMSS product with a correlation coefficient of $0.91$. Temperature observations collected at depths of 15m from surface drifters are generally used in validation satellite SST validation studies. Our research however shows that the strongest correlation between the observations and the L4 SSTs are obtained when comparisons are made with the surface temperatures (upper 3 m) rather than with temperature measurements undertaken at a 15m depth. In this study, we also show that the analysis errors provided in L4 SST products are not good indicators of the satellite SSTs accuracies. All SST product show a warm bias closer to the coast and fail to adequately capture temperature variations over the shelf break, where temperature gradients are high and increased oceanic variability occurs.