On gravity wave parameterisation in vicinity of low‐level blocking

Abstract

This note is framed as an open question to the community regarding parameterisation schemes using the blocking layer depth to reduce the orographic gravity wave drag. It is the purpose of this note to argue that the current orographic gravity wave drag parameterisation in the vicinity of blocking is inadequate. Reducing the gravity wave amplitude (and thereby reducing the gravity wave drag) by assuming an effective mountain height dependent on the blocking depth is not realistic. The arguments given here will hopefully spark a debate and new considerations, ultimately leading to improvements in current orographic gravity wave drag parameterisations. This note illustrates that low-level blocking can induce more gravity waves or gravity waves with a higher momentum flux (compared to the current parameterisation schemes). More realistic parameterisation schemes are likely to improve the models' performance. However, the fact is complex theories are needed to describe gravity wave excitation by orography so that it is difficult to represent gravity wave nature by a 'too simple' parameterisation scheme.

Observations of Gravity Wave Refraction and Its Causes and Consequences

“…The horizontal wavelength is visibly shorter in the mountain wave model. One possible explanation for this is that the mountain wave model only uses the width of the ridge to determine the horizontal wavelength and does not take the low‐level blocking width into account as suggested by Geldenhuys (2022b). An interference GW structure exists (Figure 7d) in the layer containing family 2 and only vaguely similar to the ALIMA observations.…”

“…This is a direct result of the limitation imposed in the models on the horizontal propagation of GWs. This triggered a number of studies; some reviewed parameterization schemes (e.g., Geldenhuys, 2022b; Plougonven et al., 2020), some increased drag from other known sources (e.g., Garcia et al., 2017; Polichtchouk et al., 2018a, 2018b; Richter et al., 2010), some studies mention islands might be the source (e.g., McLandress et al., 2012), some looked at new sources (e.g., Doernbrack et al., 2021; Geldenhuys et al., 2021) and others used GW intermittency to show increased drag (e.g., de la Camara et al., 2014, 2016). The large number (listed above) of different studies to solve one problem points to three possibilities: (a) the community is uncertain what the solution is, (b) there are a number of improvements required to our model parameterization schemes, or (c) improvements are required in our understanding.…”

“…This shows how much we rely on parameterization schemes. Dedicated studies (e.g., Geldenhuys, 2022b;Plougonven et al, 2020) and the large effort by the modeling community (e.g., Boeloeni et al, 2021;Kim et al, 2021;Polichtchouk et al, 2018aPolichtchouk et al, , 2018bSandu et al, 2016) show that parameterizations are still important even though computational developments allow to resolve larger parts of the GW spectrum.…”

Observations of gravity wave refraction and its causesand consequences

“…The horizontal wavelength is visibly shorter in the mountain wave model. One possible explanation for this is that the mountain wave model only uses the width of the ridge to determine the horizontal wavelength and does not take the low-level blocking width into account as suggested by Geldenhuys (2022b). An interference GW structure exists (Figure 7d) in the layer containing family 2 and only vaguely similar to the ALIMA observations.…”

“…This is a direct result of the limitation imposed in the models on the horizontal propagation of GWs. This triggered a number of studies; some reviewed parameterization schemes (e.g., Geldenhuys, 2022b;Plougonven et al, 2020), some increased drag from other known sources (e.g., Garcia et al, 2017;Polichtchouk et al, 2018aPolichtchouk et al, , 2018bRichter et al, 2010), some studies mention islands might be the source (e.g., McLandress et al, 2012), some looked at new sources (e.g., Doernbrack et al, 2021;Geldenhuys et al, 2021) and others used GW intermittency to show increased drag (e.g., de la Camara et al, 2014de la Camara et al, , 2016. The large number (listed above) of different studies to solve one problem points to three possibilities: (a) the community is uncertain what the solution is, (b) there are a number of improvements required to our model parameterization schemes, or (c) improvements are required in our understanding.…”