Shedding LIGHT on eddy-induced mixing in an idealized circumpolar current via online particle tracking

Abstract

Many common offline diagnostics, such as Lagrangian particle tracking, will have to be computed online as computing architectures approach Exascale. In this talk I describe development of the online Lagrangian In-situ Global High-performance particle Tracking (LIGHT) diagnostic within the Model for Prediction Across Scales Ocean (MPAS-O) component of the Department of Energy’s Accelerated Climate model for Energy (ACME). LIGHT is fully parallel and capable of advecting the same number of Lagrangian particles as Eulerian grid cells. Lagrangian trajectories can be used to quantify mixing due to mesoscale eddies using two approaches: dispersion- and transport-based mixing metrics. In this talk, I highlight dispersion-based metrics and use filtering to quantify mixing occurring due to eddies, the mean flow, and nonlinear interactions in an idealized circumpolar current. Meridional diffusivity depth-variability is assessed via space-time dispersion of particle clusters over ten years’ worth of a million online particle trajectories. Diffusivity in the jet is largest near the critical layer supporting mixing suppression and critical layer theory. However, it attenuates less rapidly with depth in the jet than both eddy velocity and kinetic energy scalings suggest. A mean-eddy scale separation hypothesis is suspect because both nonlinear upper bound and lower bound diffusivity estimates are large and mixing arises from nonlinearity, i.e., eddy-produced initial filamentation that background mean shear further strains to produce enhanced mixing. Nonlinearity accounts for the majority of the total diffusivity. Eddy and mean flow interactions appear to reduce nonlinear upper bound mixing, but for the lower bound enhance diffusivity near the critical layer. Broadly, this work suggests that diffusivity parameterizations accounting for both diffusivity nonlinearity and depth variability are needed.

PHILLIP J. WOLFRAM

Los Alamos National Laboratory