A One-Dimensional Model for Investigating Scale-separated Approaches to the Interaction of Oceanic Internal Waves

Kurt L Polzin,Yuri V Lvov

2023-10-26

Abstract:High-frequency wave propagation in near-inertial wave shear has been considered fundamental in setting the spectral character of the oceanic internal wave continuum and for transporting energy to wave-breaking. We compare idealized ray tracing numerical results with metrics derived using a wave turbulence derivation for the kinetic equation and a path integral to study such scale-separated interactions. At small inertial wave amplitudes, all three provide consistent descriptions for the mean drift of wavepackets in the spectral domain and dispersion about that mean drift. Extrapolating our results to the background internal wavefield over-predicts downscale energy transports by an order of magnitude. At oceanic amplitudes, however, the numerics support diminished transport and dispersion that coincide with the mean drift time scale becoming similar to the lagged correlation time scale. Despite this decrease, numerical estimates of downscale energy transfer are still significantly larger than oceanic derived metrics. Our results support replacing the long standing interpretive paradigm of extreme scale-separated interactions with a more nuanced slate of 'local' interactions in the kinetic equation.

Atmospheric and Oceanic Physics,Fluid Dynamics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the effectiveness and applicability of different scale - separation methods in ocean internal - wave interactions. Specifically, the paper explores the phenomenon of high - frequency - wave propagation in near - inertial - wave shear, which has been considered to determine the spectral characteristics of the continuous spectrum of ocean internal waves over the past 40 years and is crucial for energy transfer to the wave - breaking process. The author studies this scale - separation interaction by comparing the numerical results of idealized ray - tracing with the equations of motion and path integrals derived from wave - turbulence theory. These diagnostic indicators include the average drift of the vertical wavenumber, the dispersion of this average drift, the time - lag correlation estimate of the wavenumber, and the phase - locking of the wave packet and the background.

The paper points out that under small inertial - wave amplitudes, all three methods provide a consistent description of the average drift of the wave packet in the spectral domain and its dispersion. However, when extrapolating the results to the background internal - wave field, the predicted downward - scale energy transfer is an order of magnitude higher than the actual observations. Under ocean - amplitude conditions, the numerical results support reduced transfer and dispersion, which coincides with the average - drift time scale becoming similar to the lag - correlation time scale. The author interprets this as a transition to a non - Markovian process. Nevertheless, the numerically estimated downward - scale energy transfer is still significantly larger than the ocean - derived indicators.

The author believes that the residual differences between the ray - path formula and the observations stem from the inappropriate neglect of Bragg - scattering resonances, similar to the way in which the fourth - order cumulant provides an eddy - damping term in three - dimensional turbulence. Their results support replacing the long - standing interpretive paradigm of extreme - scale - separation interactions with a more detailed framework of "local" interactions.

In terms of formulas, the following key formulas are used in the paper:

1. **Ray equation**:
\[
\frac{dm}{dt} = -k\frac{\partial U(z,t)}{\partial z}, \quad \frac{dz}{dt} = \frac{\partial\omega}{\partial m}
\]
where \(k\) is the horizontal wavenumber, \(m\) is the vertical wavenumber, \(\omega\) is the intrinsic frequency, and \(f\) is the Coriolis frequency.

2. **Dynamic equation**:
\[
\frac{\partial n(m,t)}{\partial t}+\frac{\partial}{\partial m}\left(D_{33}(m)\frac{\partial n(m,t)}{\partial m}\right) = 0
\]
where \(D_{33}\) is the vertical - vertical component of the diffusion tensor.

3. **Diffusion coefficient**:
\[
D_{33}=\pi k^{2}f\int dp_{1}n(p_{1})m_{1}^{2}\delta(m_{1}\sigma_{m}-\sigma_{1})
\]

4. **Average drift**:
\[
\langle\frac{dm}{dt}\rangle=\frac{3}{4}kmm^{*}e_{0}\frac{1}{N}
\]

5. **Time - lag auto - correlation function**:
\[
C_{ij}(p)=\int_{t - \tau}^{t}dt'\left[\left(\frac{dm_{i}}{dt}(r(t))-\langle\frac{dm_{i}}{dt}(r(t))\rangle\right)\left(\frac{dm_{j}}{dt}(r(t'))-\langle\frac{dm_{j}}{dt}(r(t'))\rangle\right)\right]
\]

These formulas and analysis methods together form the core of the paper, aiming to reveal the complex dynamic processes in ocean internal - wave interactions.