LONG-PERIOD VARIATIONS OF WIND PARAMETERS IN THE
MESOPAUSE REGION AND THE SOLAR CYCLE DEPENDENCE
K. M. Greisiger, R. Schminder, and D. Kurschner
1. Heinrich-Hertz-Institute of Atmospheric Research and Geomagnetism,
Observatory of Atmospheric Research, DDR-2565 Kuhlungsborn, G.D.R.
2. Collm Geophysical Observatory of the Karl-Marx-University , Leipzig
DDR-7261 Collm, G.D.R.
A solar-cycle dependence of wind parameters below 100 km was found for
the first time, by SPRENGER and SCHMINDER (1969) on the basis of long-term
continuous ionospheric drift measurements in the l.f. range at the
observatories in Kuhlungsborn and Collm. They observed during the winter a
positive correlation of the prevailing wind with solar activity, whereas
the amplitude of the semi-diurnal tidal wind showed a negative correlation.
This result was confirmed later on by radar meteor wind measurements
(method D2) at Obninsk and further D1 -measurements at Kuhlungsborn and
Collm (PORTNJAGIN et al. , 1977) and, for the prevailing wind, by DARTT et
Prevailing wind : The continuation of D1 -measurements at Collm and the
beginning of continuous D2-mesurements at Kuhlungsborn in 1976 made
possible the further investigation of such long-periodic variations in the
middle atmosphere. During the years of maximum solar activity (1979-1981)
we did not observe the expected large westerly wind velocities in winter;
on the contrary, the velocities were as low as during the solar minimum of
1964/65. For the winter periods from 1975/76 to 1983/84, the correlation
of the average prevailing wind V (zonal component) with the 10.7 cm radio
emission of the sun F^q now became negative with correlation coefficients
as high as = -0.93 for Dz and -0.76 for D1 -measurements (As winter we took
the average of November, December and January).
For the meridional prevailing wind no significant variation was found
in this investigation. The same comparison as for winter was done for
summer (June, July, August) where the previous investigations gave no
correlation. Now the D2 values, too, showed a significant negative
correlation of the zonal prevailing wind with solar activity (r = -0.83)
for the years 1976-1983. The Dl-results of Collm have the same tendency
but a larger dispersion due to the lower accuracy of the harmonic analysis
because of the shorter daily measuring interval in summer (night-time
measurements). The same sense of the long-term variation of V in winter
and in summer let us expect that it is rather a manifestation ol an annual
characteristic. Therefore we have estimated annual averages approximating
the seasonal variation of V by a mean value V and an annual and
semi-annual harmonic which are the essential terms. The course of the
variation of V since 1964 (solar minimum) up to 1984 is reproduced in
Fig. 1 in the 'upper part and compared with annual averages of the solar
10.7 cm radio flux F 1 n 7 . Here we can see a positive correlation for the
years from 1964 to 1973 and a negative one later on with an indication of
new change in 1984. (The annual and semi-annual harmonics of the
prevailing wind showed no distinct long-term variation as also found by
7964 66 68 70 72 74 76 78 80 82 94 year
I I i— I i I i I i I i I i— I i__l i ■ r-
r ' i ' i ■ — i i i ' i ' — i i i i i i i i i
7964 66 68 70 72 74 76 79 80 82 94 year
Fig. 1 Variation of annual means of the zonal prevailing wind,
amplitude of the semi-diurnal tidal wind (both at 95 km and at
medium latitudes) and solar 10.7 cm radio flux for 1964-1984.
DARTT et al., 1983). The corresponding V values of Saskatoon for 1979
and 1982/83 (taken from the literature and analyzed in the same way)
confirm the low velocities during the last solar maximum found by the
measurements at Kuhlungsborn and Collm. Summarizing one can say that the
average circulation (annual mean) of the upper mesopause region exhibits a
distinct long-term variation which for some series of years seems to have
close connections with solar activity variations. But, in reality, such an
alternating mechanism of direct solar control seems to be very improbable.
We therefore conclude that the long-periodic variations found in the
prevailing circulation in 90-100 km at medium latitudes have essentially an
internal atmospheric cause and are of climatic character with a time-scale
of about 10 years. The source of these variations could be the recently
intensively investigated mechanism of momentum deposition and generation of
turbulence by breaking internal gravity waves (LINDZEN, 1981, MATSUNO,
1982, HOITON and ZHU, 1984). This mechanism is very likely the dominating
factor for determining the strength and direction of the relatively weak
prevailing circulation in the mesopause region.
Semi-diurnal tidal wind : In contrast to the prevailing wind variations we
found for the semi-diurnal tidal wind no change in the dependence on solar
activity. On the basis of the radar meteor wind measurements at
Kuhlungsborn we can now show that the same anti-correlation as in winter
exists also in summer. The D1 results for summer have a large dispersion
and give, as previously, no correlation. But from the D2 results one can
suppose that the solar cycle dependence of the semi-diurnal tidal wind is
not seasonal and should exist also for annual means. Concerning D1 results
one can expect that the relatively short summer period with values of large
dispersion will not qualitatively influence an annual characteristic. We
estimate (as for the prevailing wind) by harmonic analysis of the seasonal
variation an annual mean of the semi-diurnal tidal amplitude V_ Q as well as
an annual and semiannual harmonic. In the lower part of Fig. I we see the
course of V_q since 1964 from D1 results and since 1976 from D2 results.
The comparison with the variation of the solar 10.7 cm flux shows a clear
anti-correlation. The D1 results of V„ n give a significant correlation
coefficient r = -0.77 for almost two solar cycles (1964-1984). The D2
results confirm this with r = -0.75 for the last solar cycle.
Direct solar control of the amplitude of the semi-diurnal tidal wind
which is a manifestation of the thermally excited solar semi-diurnal
atmospheric tide seems to suggest itself. The thermal excitation takes
place mainly by absorption of the solar UV in the wave length range 200-370
nm due to ozone in the stratopause region. But after recent satellite
measurements we have only a weak enhancement of 0^ and UV (some percents)
from solar minimum to maximum and, moreover, our observations show a
decrease of the tidal wind amplitude with increasing solar activity. The
semi-diurnal tidal wave may be influenced further by the propagation
conditions between the excitation layer and the level of observations; that
is by the temperature profile and the background zonal wind. As can be
seen from recent numerical models of the semi-diurnal tide (WALTERSCHEID et
al., 1980) these factors must have variations during the solar cycle of the
same order as their seasonal ones in order to cause effects of the observed
magnitudes (percent variation 50-100 %). But solar cycle variations of
temperature and zonal wind in the upper stratosphere and mesosphere, found
up to now, are too small and not in the right sense to explain our results.
Continuous wind observations in the upper mesopause region over more
than twenty years revealed distinct long-term variations, the origin of
which cannot be explained with our present knowledge. But, for a
climatology of the middle atmosphere, they are of great interest. Further
studies are therefore desirable, both theoretical and observational. it
has been shown that lower ionospheric drift measurements and the radar
meteor wind method are valuable tools for a continuous monitoring of the
wind field at the top of the middle atmosphere over a time period of
1. Dartt, D.G. Nastrom, and A. Belmont, 1983, J. Atm. Terr. Phys. 45, pp.
2. Holton, J.R., and Zhu Xun, 1984, J. Atm. Sci. 41^, pp. 2653-2662.
3. Kaidalov, O.V., K.M. Greisiger, and K. Sprenger, 1977, Phys.
Solaterr., Postdam, pp. 91-96.
4. Lindzen, R.S., 1981, J. Geophys. Res. 86, pp. 9707-9414.
5. Matsuno, T. , 1982, J. Meteorol. Soc. Japan 60, pp. 215-226.
6. Sprenger, K. and R. Schminder, 1969, J. Atm. Terr. Phys. PP-
7. Waiterscheid, R.L., J. G. De Vore, and S. V. Venkatesvaran, 1980, J.
Atmos. Sci. 37, pp. 455-469.