The Gulf Stream (GS) is one of Earth’s strongest ocean currents and a cornerstone of global circulation, playing a critical role in climate regulation. However, coarse-resolution climate models poorly represent western boundary currents, and even 1/4° resolution simulations will remain inadequate without improved parameterizations. At medium resolution, a primary challenge is the GS separation from the coast, which depends on coastal curvature and the inertia needed to overcome topographic constraints. While the “volume penalization” method introduced by Debreu et al. (2022) for the CROCO model has yielded promising results for flow-bathymetry interactions, the GS at 1/4° resolution still lacks inertia. This limitation highlights the need to realistically represent the energy cycle of eddy activity and its feedback on the mean flow.

To address these challenges, we analyze a Lorenz energy cycle from high-resolution, coupled air-sea simulations of the Gulf Stream, using CROCO and WRF models. Our findings emphasize the dominant role of potential rather than kinetic energy routes in mesoscale to submesoscale energy transfer. This downplays the role of an interior route to mesoscale KE dissipation, with surface and bottom drag emerging as dominant mechanisms. Our results therefore suggest that modeling a realistic energy cycle at medium resolutions requires minimizing momentum dissipation through high-order discretization methods, while adequately addressing eddy-induced transport leading to potential energy dissipation. Additionally, recovering the inverse KE cascade would benefit from recent advances in energy backscatter parameterizations. Finally, accurate representation of surface and bottom drag — via air-sea coupling or appropriate parameterizations — is essential for capturing the GS dynamics and their broader climatic implications.