Mesoscale eddies are widespread in the ocean, and the mixing processes they induce play a crucial role in the transport of heat and substances, as well as in climate regulation. Consequently, the representation of eddy-induced mixing processes in non-eddy-resolving models significantly impacts the simulation of large-scale ocean circulation and climate change. Previous studies have shown that eddy-induced mixing coefficients vary greatly in space and often differ in different directions (such as meridional and zonal), a phenomenon known as the anisotropy of eddy-induced mixing coefficients. However, due to the lack of direct observational data on eddy-induced mixing in the ocean interior, the global spatial distribution of anisotropic eddy-induced mixing coefficients remains poorly understood. This study integrates the anisotropic isopycnal thickness diffusion scheme (GM scheme) and the along isopycnal diffusion scheme (Redi scheme) using the GECCO (German contribution to the Estimation of the Circulation and Climate of the Ocean) four-dimensional variational assimilation system. By assimilating data from the past 10 years of sea surface and ocean interior observations, we derived the spatial distribution of anisotropic eddy-induced mixing coefficients in the global ocean. The results indicate that both GM and Redi mixing coefficients exhibit significant spatial variability. Specifically, the GM mixing coefficient shows stronger anisotropy, while the Redi mixing coefficient shows weaker anisotropy. Further experiments demonstrate that incorporating anisotropic mixing coefficients can improve the accuracy of model simulations. This finding suggests that including anisotropic eddy-induced mixing coefficients in ocean models can enhance the simulation of large-scale ocean circulation and thermohaline structure, thereby improving our ability to predict and understand the global climate system.

 

eddy-induced mixing,parameterization,assimilation