An eddy-energy based parameterization of subgrid variance contributions to resolved density

The Gulf Stream current plays a crucial role in the global climate, transporting warm and saline subtropical waters to high northern latitudes (i.e. the Nordic and Labrador Seas). However, current global climate models show large errors in sea surface temperature (SST) in the North Atlantic, largely due to errors in the path of the Gulf Stream as it leaves the east coast of North America and heads towards Europe. Upstream advective errors in heat transport then lead to errors in air-sea heat exchange.

We propose a new parameterization to take into account the subgrid scale density changes due to nonlinear equation of state in the mesoscale front regions. The parameterization depends on representing density fluctuations in the subgrid scale using a prognostic 2D eddy kinetic energy (EKE) and has following advantages:
i) self-contained (i.e. all parameters can be computed from model variables);
ii) based on physical processes (i.e. mesoscale eddies through EKE field);
iii) does not contain any dimensional coefficients (i.e. obeys the Buckingham Pi theorem).

We show two concurrent simulations using the Modular Ocean Model version 6 (MOM6) configured as global coupled sea ice-ocean model at 0.5 degree resolution forced with CORE2 atmospheric forcing. The new parameterization leads to significant improvements in the Gulf Stream, Kuroshio and Agulhas regions compared to the control simulation. SST biases in the North Atlantic are diminished significantly.