Challenges resolving the warm Agulhas Return Current extension lead to biases in satellite and reanalysis products

Abstract On 30 November 2010, a NOAA Ocean Climate Stations surface mooring with sensors to measure air–sea exchanges of heat, moisture, momentum, and CO 2 ; and near‐surface currents, temperature and salinity, was deployed at the inshore edge of the Agulhas Return Current. These data present the rare opportunity to observe and validate ocean–atmosphere exchanges in commonly used reanalysis products in this remote and undersampled region. Unexpectedly, 48 days after deployment, the mooring line broke loose and the mooring began to drift eastwards with the Agulhas Return Current until it was recovered on 8 March 2011. This study demonstrates that, while the reanalyses' sea surface temperatures (SST) and air temperatures have large biases compared to the mooring, the air–sea temperature differences and thus atmospheric boundary layer stability and air–sea fluxes (including turbulent and radiative heat fluxes and precipitation minus evaporation) compare reasonably well. This SST bias is evident in several satellite SST products. Biases are exacerbated with high cloud cover, particularly in infrared‐only SST products, but are also present in microwave radiometer SST. Biases are shown to be worse on the poleward side of the crests and troughs of Agulhas Return Current meanders, suggesting that increased resolution is needed to capture the warm ‘tongue’ structure and to avoid flattening out sharp temperature gradients. Furthermore, the mooring data reveal the cooling of the warm ‘tongue’ as the Agulhas Return Current extends eastwards past 36° E is predominantly due to net surface heat flux cooling. The implied weak mixing could explain how the ‘tongue’ is able to extend so far eastwards.