In this study, the excitation and propagation of non-orographic gravity waves (NOGWs) above tropopause depressions usually associated with propagating Rossby waves is investigated. Observations and analysis of NOGWs in the upper strato- sphere during research flight 25 of the 2014 DEEPWAVE campaign show hydrostatic wave patterns above a tropopause depression comparable to waves excited by orography at the surface. These gravity waves (GWs) stay above the depression as the depression travels east with the phase speed of the Rossby wave. In austral winter, the polar night jet (PNJ) develops in the Southern Hemisphere, leading to a much faster zonal stratospheric flow than the Rossby wave’s phase speed. Therefore, the source of these GWs could simply be the vertical displacement of the airflow above the tropopause mimicking an obstacle to the flow above.
The goal was to evaluate this potential NOGW source through idealized numerical simulations with the geophysical flow solver EULAG. Typical shapes of tropopause depressions and typical stratospheric winter conditions suggest the excitation of low-frequency or inertia-gravity waves (IGWs). To account for their oscillation in the vertical and horizontal plane, all simulations were conducted in 3D. Preliminary studies were performed to assess the reliability of the model by comparing its numerical solutions to analytical solutions of different mountain wave regimes. In particular, comparisons of the gravity wave drag and surface momentum fluxes were instructive and guided the setup of the following simulations. Within the framework of this thesis, the simulations are constrained to the stratosphere, with the tropopause represented as a transient, impermeable, and frictionless lower boundary. Multiple simulations with different idealized tropopause shapes and different idealized background conditions were carried out to characterise the sensitivity of the waves’ momentum and energy fluxes and to identify scenarios which reproduce the GW patterns observed during the DEEPWAVE case study. For example, varying the vertical ambient wind profile ue(z) revealed that increasing ue(z) by a constant speed Uc has a similar effect as decreasing the fold’s propagation speed by Uc. Furthermore, a local wind minimum significantly impacts upper stratosphere GW activity, even without introducing a critical level. This so-called valve layer still involves wave dissipation and first dissipates waves that propagate into the lee of the depression. Thus, in the presence of a valve layer IGWs look like hydrostatic non-rotating GWs in vertical cross-sections, with the main GW activity directly above the depression. Variations of the PNJ’s meridional location and strength showed that a southward position of the PNJ relative to the propagating tropopause depression and a very strong PNJ are necessary to reproduce the zonally elongated GW pattern detected in ERA5.
The idealized numerical simulations also demonstrated that GWs excited by a transient source like a propagating tropopause depression have a distinct pattern of ascending phase lines in time-height diagrams emulating the measurements of vertically staring ground-based lidar instruments. Since stationary mountain waves result in horizontal phase lines in these measurements, it might be possible to differentiate between these two sources and identify GWs excited by propagating tropopause depressions in lidar measurements. However, ascending phase lines in time-height diagrams can also be traced back to other processes (for example, transient conditions). Therefore, two measurement periods with upward-tilted phase lines were selected from the dataset of the Rayleigh lidar CORAL operating at the southern tip of South America and analysed based on ERA5. In these two cases, the analysis did not support a connection between the ascending phase lines and a passing tropopause depression but suggests an excitation by orography in the context of a transient wind forcing in the lower troposphere. On the other hand, the ERA5 analysis of the second case also indicated GWs above a tropopause depression, similar to the DEEPWAVE case study that motivated this thesis. The direct comparison of these NOGWs over the Pacific Ocean to the mountain waves above the Southern Andes and CORAL’s location showed that temperature amplitudes of NOGWs are more than a factor of 4 smaller than those of mountain waves. It can be challenging to isolate the contribution of small-amplitude NOGWs excited by tropopause expressions from the contribution of large-amplitude MWs to the total GW field measured by vertically staring lidars primarily influenced by mountain waves such as CORAL.
Next steps should involve a complete simulation of tropospheric and stratospheric airflows by simulating baroclinic instabilities in the presence of a PNJ in the upper stratosphere. In this way, the assumption of an impermeable tropopause can be dropped and it would be possible to generalise the conclusions of this thesis.