Refraction and straining of wind-generated near-inertial waves by barotropic eddies

TL;DR

This paper investigates how wind-generated near-inertial waves are distorted by barotropic eddies, revealing that refraction causes linear growth of wavevector components over time, while strain effects are limited in steady flows but significant in unsteady flows.

Contribution

It provides a detailed analysis of wavevector evolution under eddy strain and refraction, highlighting the limited role of strain in steady flows and its importance in unsteady flows.

Findings

Refraction causes linear growth of wavevector components over time.

Strain is ineffective at the jet center and in steady flows.

In unsteady flows, strain modifies vorticity gradients and energy transfer.

Abstract

We analyze the distortion of wind-generated near-inertial waves by steady and unsteady barotropic quasi-geostrophic eddies, with a focus on the evolution of the horizontal wavevector $\boldsymbol{k}$ under the effects of mesoscale strain and refraction. The model is initialized with a horizontally-uniform ($\boldsymbol{k}=0$) surface-confined near-inertial wave which then evolves according to the phase-averaged model of Young and Ben Jelloul. A steady barotropic vortex dipole is first considered. Nearly monochromatic shear bands appear in the jet region as wave energy propagate downwards and towards anticyclone. As a result of refraction, both horizontal and vertical wavenumbers grow linearly with the time $t$ elapsed since generation such that their ratio, the slope of wave bands, is time-indepedent. Analogy with passive scalar dynamics suggests that strain should result in the exponential growth of $|\boldsymbol{k}|$. Here instead, strain is ineffective not only at the jet center, but also at its confluent and diffluent regions. Low modes rapidly escape below the anticyclonic core such that the weakly-dispersive high modes are dominant in the mixed layer. In the weakly-dispersive limit, $\boldsymbol{k}=- t \nabla \zeta(x,y,t)/2$ provided that (i) the eddy vertical vorticity $\zeta$ evolves according to the barotropic quasi-geostrophic equation; and (ii) $\boldsymbol{k}=0$ initially, as is typically assumed for near-inertial waves generated by large-scale atmospheric storms. In steady flows, strain is ineffective because $\boldsymbol{k}$ is always perpendicular to the flow. In unsteady flows, straining modifies the vorticity gradient and hence $\boldsymbol{k}$, and may account for significant energy transfers.