Jeffrey. M. Forbes

Department of Aerospace Engineering Sciences
University of Colorado, Boulder, CO, 80309

Nikolai Makarov and Yuri Portnyagin

institute for Experimental Meteorology
Scientific Production Association TYPHOON, Obninsk, Russia

A meteor radar was deployed at Amundsen-Scott Station at South Pole from January, 1995 through January, 1996, and is planned to operate mid-November, 1996, through mid-February, 1997. The radar measures winds near the earth's mesopause, the height in the atmosphere (~ 85-95 km) where the atmospheric temperature is coldest (~ 140-220 K). The experiment is a collaborative project between the University of Colorado and the Institute for Experimental Meteorology, Obninsk, Russia. The project is funded through the Office of Polar Programs of the U.S. National Science Foundation.

Winds are obtained by measuring the Doppler shift of coherent radio reflections from the ionized trails produced by meteor ablation in the upper atmosphere. The meteor trails are several kilometers in length, with an initial radius of about 1 meter. The measured hourly mean winds represent Gaussian-weighted averages centered near 95 km above sea level with a half-width of almost 10 km. The radar is located 500 meters from the geographic Pole along the 165 degree meridian; under normal operation, soundings are made along the four geographic longitude meridians of 0, 90E, 180E, and 90W. Centers of the sounding regions are removed from the pole in the horizontal plane about 180 km, yielding the following approximate coordinates of the wind measurements: (88S, 0), (88S, 90E), (88S, 180E), (88S, 90W). Line of sight winds are converted to horizontal winds by assuming the vertical wind contribution to be negligible. The only wind component measurable from the South Pole is the meridional wind, taken here to be positive in the northward direction.

The radar operates at 33.6 MHz and and average power of 120 Watts. The antenna system consists of four five-element transmitting/receiving Yagii antennae. In the search mode, the atmosphere is "interrogated" for the presence of meteor trails at a frequency of 100 Hz. After a wind measurement is made from a trail, which takes about 0.1s, the system returns to search mode. On the average, some 20-80 wind measurements are obtained to compute an hourly-mean meridional wind measurement in each direction. Occasionally the number of meteor returns exceeds 100, and occasionally there are 0-10 returns.

Wind observations commenced on January 19, 1995. Hourly values for the first 7 days of observations are illustrated in Figure 1. Spectral analysis shows the dominant periodicity in these data to be 12 hours. The least-squares fit of a 12-hour sine wave to the data is shown by the dotted line. During this period, two of the antennae were pointed along the 90W meridian, and sounded independent meteor trails. A comparison of these measurements were used to estimate the random error associated with any given wind determination. Typical random errors range between about 3 and 9 ms^(-1), with an average of about 5 ms^(-1). After February 1, 1995, the antennae were pointed in four orthogonal directions for future routine operations.

nice photo

Careful examination of the data in Figure 1 points to a westward-propagating 12-hour oscillation with a zonal wavenumber equal to one (i.e., around a constant latitude circle, the wave has one maximum and one minimum). This result is somewhat surprising, since according to classical tidal theory (Chapman and Lindzen, 1970) the predominant component of the solar semidiurnal tide is migrating with the apparent motion of the sun, corresponding to s = 2. However, the 12-hour s = 1 oscillation has been observed at the South Pole before (Hernandez, 1993) and appears to be a regular feature of polar mesopause dynamics.

A full analysis of the above phenomenon, focusing on the February 4-11, 1995 period, is provided by Forbes et al. (1995). These authors interpret the s = 1 oscillation to result from the nonlinear interaction between the migrating semidiurnal tide and a stationary (non-propagating) wave with s = 1. The proposed mechanism represents an alternative to the gravity-wave driven 'pseudotide' theory put forth by Walterscheid et al. (1986) to explain the occurrence of unexpectedly large semidiurnal tidal oscillations at high latitudes.

Analyses of data extending from January 19, 1995 through June 20, 1995, are presented in Portnyagin et al. (1996), with emphasis on the prevailing (24-hour average) wind component. The observed phenomena include (a) A regular month-to-month migration in the phase of the s=1 quasi-stationary component of the wind field; (b) During April, a 10-day period of strong net wind divergence from the pole; (c) evidence for upwelling associated with very high magnetic activity. It is speculated that the wind divergence event may be associated with the mesospheric penetration of momentum fluxes known to accompany intense wave 1 activity in the Southern Hemisphere during April.

nice photo

Figure 2 illustrates the wind divergence effect observed during April, 1996. By taking the average of the prevailing meridional winds in the four azimuth directions, we have constructed the daily zonal mean meridional winds, as illustrated in Figure 2. The large negative winds imply a net divergence away from the pole. Also shown is the daily magnetic activity index, Ap. We see that on April 7, a tendency towards wind convergence occurs in conjunction with a significant increase in magnetic activity. A reasonable explanation appears to be upwelling induced by Joule heating in the atmosphere just above the South Pole. We could find no evidence that suggests that our experimental determination of neutral winds is significantly affected by the increase in magnetic activity. In fact, it is noted that in Figure 2, the "wind divergence event" began long before the increase in magnetic activity.

Other phenomena currently under investigation with regards to the South Pole wind measurements include the seasonal variability of diurnal and semidiurnal tides, planetary wave modulation of the tides, the 10.1-hour normal mode, eastward and westward propagating waves with periods between 1.5 and 10 days, and continuation of the prevailing wind studies discussed above.

Opportunities for graduate students and post-docs include investigation of the above phenomena using the South Pole data, collaborative studies using wind measurements from other Antarctic radars, and theoretical/numerical modeling studies aimed at understanding the observations.


Chapman, S., and R.S. Lindzen,Atmospheric Tides, D. Reidel Publishing Company, Dordrecht, 1970.

Forbes, J.M., Makarov, N.A., and Portnyagin, Yu. I., First results from the meteor radar system at South Pole: A large 12-hour oscillation with zonal wavenumber one, Geophys. Res. Lett., in press, 1995.

Hernandez, G., Fraser, G.J., and R.W. Smith, Mesospheric 12-hour oscillation near South Pole, Antarctica, Geophys. Res. Lett., 20, 1787-1790, 1993.

Walterscheid, R.L., Sivjee, G.G., Schubert, G., and R.M. Hamwey, Large-amplitude semidiurnal temperature variations in the polar mesopause: evidence of a pseudotide, Nature, 324, 347-349, 1986.