North Atlantic climate mean state and variability: Local effects of mesoscale ocean dynamics

Western Boundary Currents, such as the Gulf Stream, are regions of vivid air-sea interaction. Mesoscale features of these currents play a fundamental role in global ocean heat transport and exchange with the atmosphere. Related processes and their interactions across scales have gained increasing attention in the last years, since high-resolution, mesoscale-resolving modeling became computationally feasible on climate time scales. Here, we show the impact of explicitly resolving the oceanic mesoscale in the coupled global climate model FOCI on North Atlantic and European climate. For this purpose, we use the ocean nesting capability in FOCI, which facilitates regional ocean grid refinement. We explore and compare pre-industrial simulations each extending over at least 150 years: a reference run without any grid refinement and an experiment with a nest in the North Atlantic. Technically, the regional ocean nest maintains frequent two-way exchange with the global host grid, which in turn is fully coupled to the atmosphere model. The ocean model NEMO has a global resolution of 1/2˚ model with 46 vertical levels and 1/10˚ refinement in the nest region, while the atmosphere model ECHAM6 has a 1.8˚ horizontal resolution (T63) and 95 vertical levels, including the strato- and mesosphere. Within the nest region, the increased resolution leads to a more eddy-rich simulation and an improved mean state. The North Atlantic Current is considerably better represented, which reduces the typical North Atlantic cold bias from -8˚C in the reference run without nest to -2˚C. Beyond local bias correction of the mean state, we will also discuss the impact of explicitly modeling ocean mesoscale dynamics on atmospheric variability on different time scales, such as the North Atlantic Oscillation or the Atlantic Multidecadal Variability.