An analysis of the Hines and Warner–McIntyre–Scinocca non‐orographic gravity wave drag parametrizations

Abstract

A column model based on CIRA wind and temperature profiles is employed to assess the characteristics of the Hines Doppler‐spread and Warner–McIntyre–Scinocca (WMS) non‐orographic drag parametrizations for internal gravity waves. The “Alexander–Dunkerton” variant of the WMS scheme is also briefly considered. This study goes into more detail than previous comparisons by performing a spectral analysis of the momentum deposition and drag, and by examining the ability of each scheme to reproduce a high vertical wavenumber tail consistent with atmospheric measurements. We find several undesirable characteristics in the drag produced by the Hines scheme. For typical midlatitude wind profiles, it produces an abrupt onset of large accelerations that vary strongly from one layer to the next. It is also unable to reproduce spectra consistent with observed wave saturation at high vertical vertical wavenumbers, even for the windless case. The WMS scheme has the ability to reproduce the observed spectral tail for the case of no background wind. In the presence of typical CIRA midlatitude background winds, however, it is demonstrated that the WMS saturation threshold generally does not follow the observed spectral behaviour at high vertical wavenumbers, except for one specific frequency dependence not considered in previous work. Doppler shifting is also found to interfere with the production of high‐wavenumber spectral tails consistent with observations.

Gravity Waves in Planetary Atmospheres: Their Effects and Parameterization in Global Circulation Models

“…The latter strongly distorts the vertical wavenumber spectra due to Doppler shifting. Majdzadeh and Klaassen [44] have thoroughly investigated the behavior of the parameterizations [42,43] and demonstrated that they never can reproduce the saturated m = −3 slope in the background shear, even if it was prescribed by the saturation criterion. This poses a question of whether the empirical threshold of the "universal" spectrum can be used in the presence of the mean wind at all.…”

“…McLandress and Scinocca [48] argued that the DSP could be further tuned, such that the averaged fields simulated by a GCM using different GW schemes were close. Majdzadeh and Klaassen [44] noted that the DSP could not reproduce the saturated m = −3 slope of the vertical wavenumber spectra in the absence of mean wind shear (similar to the scheme of [39]), because the large-wavenumber harmonics were progressively obliterated. The DST is based on the assumption that waves are conservative and non-interacting in the Lagrangian frame of reference and become nonlinear when transformed to the Eulerian frame [49].…”

A study of gravity waves on planetary atmospheres

“…But this approach does not account for saturation and requires an intermittency index to account for GW time scales. The spectral approaches rely on rather-arbitrary tuning of wave spectrum and abrupt critical-level absorption of harmonics resulting in a non-physical strain on simulated flow (Hoshino et al, 2013;Majdzadeh and Klaassen, 2019;Warner and McIntyre, 1996;Zalucha et al, 2013). Recently developed 'Whole Atmosphere' models extend GCMs to planetary exobase.…”