The influence of large-scale oscillations on long-term trend assessment in hydroxyl temperatures over Davis, Antarctica

• Published in: Journal of atmospheric and solar-terrestrial physics 2004-04, Vol.66 (6), p.493-506
• Main Authors: French, W.J.R., Burns, G.B.
• Format: Article
• Language: English
• Subjects: Atmospheric dynamics; Earth, ocean, space; Exact sciences and technology; External geophysics; Hydroxyl airglow; Mesospheric temperature trends; Physics of the high neutral atmosphere; Physics of the ionosphere; Physics of the magnetosphere; Planetary waves
• ISSN: 1364-6826; 1879-1824
• DOI: 10.1016/j.jastp.2004.01.027

Large-scale temperature oscillations are observed in 7 years (1995–2001) of night-time hydroxyl (6-2) band rotational temperatures, measured with a scanning spectrometer over Davis station, Antarctica (68.6°S,78.0°E). Variations of up to 30 K are evident in the 1439 nightly averages calculated, in quasi-periodic 10–60 day intervals with no apparent seasonal consistency year-to-year. We investigate the impact of this seasonal variability in large-scale temperature oscillations on long-term trend assessments. Lomb–Scargle periodogram analysis reveals that large-scale (quasi-periods 50–60 day) oscillations dominate March–April and July–October but activity generally decreases around midwinter (mid-May to late-June) when shorter-period smaller-magnitude oscillations are more characteristic. This is consistent with reduced stationary wave penetration into the upper mesosphere due to strong middle atmosphere eastward winds in the southern midwinter. A sliding-window multivariate fit with solar cycle and long-term linear trend terms shows most consistent coefficients of 0.06– 0.07 K/solar flux unit ( ∼8 K per solar cycle) and −0.5– −1.5 K/year for the 70 day windows centred between days 160 and 220 (across midwinter) each year when large-scale oscillations are minimal. Large uncertainties in the trend coefficients over the equinox intervals are obtained and we cannot determine if trends at these times are significantly different from zero. Enhanced cooling in the southern hemisphere upper-mesosphere may be linked with a trend toward stronger eastward flow in the southern high latitude polar vortex (due largely to the trend in photochemical ozone losses), reducing the penetration of waves into the southern winter mesosphere.