In this study, we apply the three-dimensional Stockwell Transform (3DST) to a novel dataset, namely airglow imager data from Rothera (68S, 68W). We use this approach to investigate small-scale high-frequency gravity waves (GWs) in the hydroxyl (OH) airglow layer, at a height $\sim$87 km in the mesosphere and lower thermosphere (MLT). MLT GWs are often underrepresented in models, being parameterised due to their small scale size and as such, the significant quantities of momentum and energy transferred by these small waves are missed. Better quantification of these waves is thus needed to support future model development. We find that the 3DST can identify waves and extract wave properties and their locations. Horizontal wavelengths are observed ranging from 10 to 40 km and vertical wavelengths of 15 to 40 km, with wave periods of 5 to 9 minutes, peaking at 7.5 minutes. These values are consistent with previous studies. Group speeds are found to be non-zero and large, implying that these GWs travel horizontally and fast. This case study demonstrates that the 3DST can be applied to airglow imager data and can successfully extract GW parameters. This is an important step in automating GW analysis in airglow.

The solar tides of the mesosphere and lower thermosphere (MLT) show great variability on timescales of days to years, with significant variability at interannual timescales. However, the nature and causes of this variability remain poorly understood. Here, we present measurements made over the interval 2005 to 2020 of the interannual variability of the 12-hour tide as measured at heights of 80 to 100 km by a meteor radar over Rothera (68S, 68W). We use a linear regression analysis to investigate correlations between the 12-hour tidal amplitudes and several climate indices, specifically the solar cycle (as measured by F10.7 solar flux), El Nino Southern Oscillation (ENSO), the Quasi-Biennial Oscillation (QBO) at 10 hPa and 30 hPa and the Southern Annular Mode (SAM). Our observations reveal that the 12-hour tide has a large amplitude and a clearly defined seasonal cycle with monthly mean values as large as 35 ms−1. We observe substantial interannual variability, exhibiting 2σ range in monthly mean 12-hour tidal amplitudes at the height of 95 km in spring of 13.4 ms−1, 11.2 ms−1 in summer, 18.6 ms−1 in autumn and 7.0 ms−1 in winter. We find that F10.7, QBO10, QBO30, SAM and time all have significant correlations to the 12-hour tidal amplitudes at the 95% level, with a linear trend also present. Whereas we detect very minimal correlation with ENSO. These results suggest that variations in F10.7, the QBO and SAM may contribute significantly to the interannual variability of 12-hour tidal amplitudes in the Antarctic MLT.