Evolution of Gravity Wave Distributions in the Polar Middle Atmosphere Observed by the Aura Microwave Limb Sounder

1. Introduction

The study of geographical distributions of gravity wave activity can provide valuable informations about the generation, propagation and dissipation of atmospheric gravity waves [1,2,3,4]. Atmospheric gravity waves are important for the transfer of energy and momentum through the atmosphere. Breaking and dissipation of atmospheric gravity waves essentially contribute to the residual meridional circulation of the middle atmosphere [5]. The present study analyses the distributions of inertia gravity waves with horizontal wavelengths between 200 and 825 km, periods between 2 and 12 hours, and vertical wavelengths between 6 and 30 km. These waves can be derived from temperature data of the satellite instrument Aura/MLS [6].

The most remarkable hotspot of gravity wave activity is found in the winter stratosphere above the Southern Andes [3,6,7,8,9]. The enhanced stratospheric gravity wave activity is due to surface wind interaction with the Andean mountain ridge and the Antarctic peninsula, so that orographic gravity waves are generated. During winter, the zonal wind is eastward in the troposphere and stratosphere, enabling the propagation of orographic waves from the ground into the stratosphere. It has been found that the gravity waves propagate against the polar vortex stream, so that there is enhanced gravity wave activity along the stratospheric polar vortex [3,10,11]. High-resolution studies of the gravity wave propagation from orographic and non-orographic sources in the southern hemisphere confirmed the picture of a gravity wave hotspot above southern Andes and enhanced gravity wave activity in a belt region over the Southern Ocean [12,13]. These studies are based on high-resolution observations of the AIRS satellite in the stratosphere and modelling of gravity wave propagation in high-resolution models.

While the gravity wave distribution is well explored in the stratosphere, there are less studies about the gravity wave maps in the mesosphere. A recent study gave an example of enhanced mesospheric gravity wave activity over the Weddell sea in summer 2010 [14]. In the present study, we will derive the temporal evolution and altitude evolution of gravity wave distributions in the middle atmosphere with focus on the gravity wave hotspots in the southern hemisphere during summer and winter from 2005 to 2021. Section 2 describes the Aura/MLS dataset and the data analysis. Section 3 presents the results on the evolution of the gravity wave distributions. Discussion and conclusions are given in section 4 and section 5.

2. Aura/MLS Dataset and Data Analysis

The study analyses temperature measurements in the middle atmosphere obtained by the Microwave Limb Sounder (MLS) aboard NASA’s Aura satellite. Aura was launched in 2004, and a detailed description of the MLS instrument is provided by [15].

The Aura satellite operates in a Sun-synchronous, near-polar orbit at an altitude of approximately 705 km, with an orbital inclination of about 98°. The satellite crosses the equator twice daily at approximately 01:45 and 13:45 local solar time (LST). Its orbital period is about 99 min. Atmospheric profiles are sampled along the satellite’s ground track at intervals of 1.48° in latitude, corresponding to approximately 165 km. Measurements are available over the latitude range from 82°S to 82°N.

The present study is based on version 5 Level 2 data from Aura/MLS. The data screening and quality control procedures follow the recommendations of [16]. Aura/MLS provides atmospheric temperature profiles on 42 pressure levels. Temperature is retrieved from measurements of thermal microwave limb emissions in the O₂ lines at 118 GHz and 234 GHz [17].

Using the same O₂ lines, Aura/MLS also measures geopotential height profiles [17]. These profiles are used to convert the temperature profiles from pressure coordinates, $T\left(p\right)$, to altitude coordinates, $T\left(z\right)$. The resulting $T\left(z\right)$ profiles cover altitudes from 13 to 90 km with a vertical spacing of 1 km. All $T\left(p\right)$ profiles are converted to $T\left(z\right)$ before further data analysis is performed.

Interpolation errors resulting from the limited vertical resolution of the Aura/MLS pressure profiles are expected to be negligible because the interpolation is performed from a relatively coarse pressure grid to a finer altitude grid.

The analysed dataset covers the period from 2004 to the end of 2021. In 2022, the number of pressure levels in the retrieved Aura/MLS profiles was reduced from 42 to 37 due to technical degradation of the MLS instrument. Therefore, Aura/MLS observations from 2022 onwards are excluded from the present study.

The precision of the temperature profiles is approximately 1 K in the stratosphere and 3 K in the mesosphere [17]. The precision of the geopotential height is 35 m between 316 hPa and 100 hPa, 44 m at 1 hPa, and 110 m at 0.001 hPa [17]. The vertical resolution of the temperature profiles is approximately 3 km in the stratosphere and 8 km in the lower mesosphere [17]. The along-track spacing of the sounding volumes is approximately 200 km.

In the following analysis, the temperature fluctuations along the suborbital track were high-pass filtered at each altitude level. The digital filter was applied in both the forward and reverse directions to avoid filter-induced phase delays. A non-recursive finite impulse response (FIR) high-pass filter with a Hamming window was used. The number of filter coefficients corresponds to a window length of three times the cut-off distance, i.e. 825 km, which is equivalent to the spatial distance covered by five consecutive temperature profiles.

The high-pass filter has a cut-off wavelength of 825 km. Thus, it retains temperature fluctuations along the suborbital track with horizontal spatial scales smaller than 825 km while suppressing fluctuations with larger spatial scales.

The angle between the gravity wave propagation direction and the suborbital track vector strongly affects the sensitivity of Aura/MLS to gravity waves [6]. If a gravity wave propagates parallel to the suborbital track, Aura/MLS can detect waves with horizontal wavelengths between 200 km, corresponding to the length of the sounding volume, and 825 km, which is determined by the cutoff of the high-pass filter. The filtering process effectively suppresses temperature fluctuations caused by tidal and planetary waves propagating in zonal directions.

For January and July of each year from 2004 to 2021, we calculated polar maps of the standard deviation of the high-pass-filtered temperature fluctuations $\delta T$, representing the mean gravity wave amplitude. The map grid points are spaced by 2.5° in latitude and 5° in longitude. Each grid cell has a width of 5° in latitude and 10° in longitude, resulting in overlapping grid cells. For each altitude and grid cell, the standard deviation $\Delta T$ was calculated from all high-pass-filtered temperature fluctuations $\delta T$ observed in January or July of the respective year.

3. Results

We explore the long-term evolution of gravity wave maps in the southern hemisphere at solstice in the years from 2005 to 2020. We selected the altitude 78 km in the mesosphere since there is a hotspot of gravity wave activity above the Weddell sea in January (summer) as well as July (winter). This hotspot in the mesosphere has not been explored in the literature yet, while the hotspot in the stratosphere was well investigated by AIRS observations and models [12].

Figure 1 shows the long-term series of gravity wave maps at 78 km altitude in January from 2005 to 2020. There is some interannual variation in the gravity wave maps. Most evident is a broad hotspot of gravity wave activity over the Weddell Sea, eastward of the Antarctic peninsula. It is a new result that this hotspot also occurs in the summer month January.

Figure 2 shows the long-term series of gravity wave maps at 78 km altitude in July from 2005 to 2020. In all years, there is an enhancement of gravity wave amplitude over the Weddell sea. This hotspot is separated from the hotspot in the stratosphere which is above the southern Andes in winter [12].

In the following we average all gravity wave maps of January or July from 2005 to 2021, in order to analyse the mean behaviour of gravity wave amplitude in summer or winter.

Figure 3 shows the evolution of the mean gravity wave activity in January (summer) from the upper stratosphere (z=44 km) to the mesopause (z=90 km), averaged for the years from 2005 to 2021. In the summer stratosphere, there is as expected no hotspot of gravity wave activity over the southern Andes. However, there is a hotspot of gravity wave activity over the Weddell sea in the mesosphere (z=74 to 86 km). We can see that there is no connection between this mesospheric hotspot and the stratosphere below.

Figure 4 shows the evolution of the mean gravity wave activity in July (winter) from the upper stratosphere (z=44 km) to the mesopause (z=90 km), averaged for the years from 2005 to 2021. As expected for winter, there is a hotspot of gravity wave activity over the southern Andes in the stratosphere and lower mesosphere from 44 to 66 km. From 70 to 90 km, the hotspot above the southern Andes diappeared and a hotspot of enhanced gravity wave activity appears above the Weddell sea (more south- and eastward of southern Andes). As in Figure 3, Figure 4 shows no connection between the mesospheric hotspot over the Weddell sea and the stratosphere below. It is a miracle, why the hotspot occurs in summer and winter above the Weddell sea.

Finally, we check if there are hotspots of gravity wave activity over the northern hemisphere in January (winter) (Figure 5) and in July (summer) (Figure 6). From these figures, it is clear that the northern hemisphere has generally no comparable strong hotspots of gravity wave activity as the southern hemisphere. There is only a region of enhanced gravity wave amplitude at z=82 km over the northern Pacific in summer and winter.

4. Discussion

Our study provided several new results. We found a hotspot of mesospheric gravity wave activity above the Weddell sea. This hotspot occurs in almost all years and in winter as well as in summer. It is not known yet if this hotspot is connected to the tropospheric wavemaker of the Antarctic peninsula. It is hard to explain why the mesospheric hotspot also occurs in summer. In case of the stratospheric hotspot over southern Andes, it only occurs in the winter season when the tropospheric orographic gravity waves can propagate from the surface to the stratosphere. In addition, we cannot find an evolution of the mesospheric hotspot over the Weddell sea arising from the stratospheric hotspot over southern Andes. In the literature, there are almost no discussions about the mesospheric hotspot over the Weddell sea. Past studies are focused on the stratospheric hotspot over southern Andes [12,13]. In the northern hemisphere, we do not find hotspots which are comparable in strength to those of the southern hemisphere.

5. Conclusion

We derived distributions of inertia gravity waves in the polar middle atmosphere from 2005 to 2021 by using temperature profiles observed by Aura/MLS. In the southern polar middle atmosphere, we found enhanced gravity wave activity in the mesosphere above the Weddell sea in summer and winter. The enhanced gravity wave activity occurs with small interannual variations in January and in July from 2005 to 2021. The mean vertical evolution of gravity wave distributions is analysed in January or July, showing the gravity wave maps separately at altitudes from 44 km to 90 km. In winter season (July), the well-known hotspot of gravity wave activity in the stratosphere over the southern Andes is isolated from the hotspot of gravity wave activity in the mesosphere over the Weddell sea. There is no evolution from the stratospheric gravity wave hotspot to the mesospheric gravity wave hotspot which is more south- and eastward of southern Andes. The cause of the mesospheric hotspot is unknown. It might be unlikely that orographic gravity waves from the troposphere can propagate to the mesosphere in summer because of wind filtering of the waves and the zonal wind reversal at the summer tropopause. In addition, there is a wide distance between the Antarctic peninsula and the mesospheric hotspot. We also presented the gravity wave distributions for the northern polar middle atmosphere which have no clear hotspots and generally smaller amplitudes compared to the southern polar middle atmosphere.