2026/06/01 by Tian Jing, Ru Chen, Andrew Stewart +2
Earth and Planetary Sciences · #Oceanographic and Atmospheric Processes #Marine and coastal ecosystems #Ocean Waves and Remote Sensing
paper · doi:10.1175/jpo-d-25-0109.1
Abstract In eddy-permitting models, only mixing induced by unresolved subgrid-scale eddies needs to be parameterized. Therefore, it is essential to study the diffusivity induced by eddies with spatial scales smaller than a separation scale L * , i.e., scale-specific eddy diffusivity. Using Lagrangian particle trajectories from a high-resolution model solution [Massachusetts Institute of Technology General Circulation Model (MITgcm) latitude–longitude–polar cap (LLC4320) configuration], we estimate full-depth and scale-dependent eddy diffusivities in the Kuroshio Extension region, and then develop a novel empirical formula to accurately represent their variations. Our results reveal significant variations in both the magnitude and spatial structure of diffusivity with L * and depth. We find that Eady scale–based mixing length parameterization excels in the upper ocean, while GEOMETRIC and topographic GEOMETRIC parameterizations better capture scale-dependent particle-based diffusivity in the depth range away from the ocean surface. At middepths and for L * ranging from 0.8° to 1.8°, the parameterization based on eddy size effectively represents both the spatial structure and magnitude of diffusivities. By combining the strengths of these parameterizations, our empirical formula provides an accurate representation of eddy diffusivity across the entire ocean depth. This study offers a promising approach for improving subgrid-scale eddy mixing parameterizations in eddy-permitting climate models. Significance Statement Eddy-permitting climate models require scale-specific eddy mixing coefficients to parameterize mixing processes induced by eddies smaller than the smallest resolvable scale. In this study, we show that the eddy mixing rate has significant spatial variability in the Kuroshio Extension region, varying with separation scale (the resolvable scale; the cutoff between resolved and unresolved eddies) and depth. By extending total eddy mixing parameterizations to the scale-dependent case and evaluating their performance, we develop a novel empirical formula. Our findings reveal that this empirical formula, which combines the advantages of scale-dependent GEOMETRIC, topographic GEOMETRIC, Eady scale–based, and eddy size–based schemes, captures both the magnitude and spatial variability of scale-specific eddy mixing rates well. These insights provide a promising foundation for improving climate model performance by using full-depth, scale-dependent eddy mixing coefficients inferred from parameterization schemes.