Investigation of model SST biases in southeastern tropical Atlantic through high-resolution regional climate simulations

The southeastern tropical Atlantic is one of the regions where global climate models consistently exhibit warm sea surface temperature (SST) biases that can exceed 5°C off the coast of Namibia and Angola near the Angola-Benguela Front (ABF). Previous studies suggest such SST biases may originate from poorly simulated surface winds and low-level clouds, as well as from errors in models' representation of both the coastal oceanic circulation and the sharp thermocline along the Angola coast. Furthermore, despite simulated SST biases show sensitivity to the surface wind-stress curl associated with the Benguela coastal low-level jet (BCLLJ), the origin of these biases is not fully understood.

We present results from a hierarchy of atmospheric, oceanic, and coupled regional model simulations with increasingly higher resolution (81km, 27km, and 9km) to investigate the origin and evolution of the SST bias. The impact of the atmospheric model resolution on the onset of the SST bias in uncoupled ocean simulations will be discussed with particular attention to the accuracy of the simulated BCLLJ and resulting coastal upwelling. Simulations from a high-resolution (9km) regional coupled model will be used to identify key air-sea feedback mechanisms that can play a key role in the development of the SST bias in southeastern tropical Atlantic.