3. Measurement Results and Discussion

3.1. Solar Eclipse on 11 August 1999

[21]  The first observation cycle was conducted during 3 days (10-12 August 1999) from 0900 to 1700 UTC+4. The aim of this measurement cycle was to study the lower ionosphere including the effects of the solar eclipse on 11 August 1999. The solar eclipse began on the Earth's surface in the observation point at 1407:28 UTC+4 and ended at 1627 UTC+4. The maximum phase of the solar disc occultation by the Moon was ~60%.

[22]  The effective power of the Sura facilities was 150 MW on 10 August and 60 MW on 11 and 12 August. A typical feature of this observation period was a disturbance in the E region. It was manifested in the occurrence of sporadic layers and intensification of the irregular structure. A layering was visually seen in the E region and the upper part of the D region. An interference of the signals scattered by APT, sporadic layers, and large-scale ionospheric irregularities were observed (which led to a multiplicity of reflections), and also there was a natural turbulence. These phenomena are described in detail by Belikovich et al. [2002a]. For example, in the day of the eclipse, there were sporadic layers: first at altitudes of 90 and 95 km and then above (at 99 km after 1400 UTC+4). At the end of this hour the structure of the E layer including E_s became vertically irregular up to a height of 108 km. Above the E -layer maximum, sometimes temporal changes in the signal amplitude were observed. The changes were caused by the horizontal motion of clouds of ionization. Such irregular phenomena make the determination of the atmospheric parameters more difficult. In particular, it was impossible to obtain reliable information from altitudes of 102-104 km in the 1400-1600 UTC+4 period.

Figure 2

[23]  Figure 2 shows the variations in density (the variations in the temperature are not shown because they are similar) at altitudes of 106 and 108 km between 1300 and 1700 UTC+4 for 10 August (points), 11 August (circles), and 12 August (crosses) 1999. Solid and dashed curves show the running means. The time of the eclipse on 11 August is marked by thick line on top of the time axis, and the maximum phase moment is marked by a cross. Figure 2 shows that on the day of the eclipse and in the adjacent days, variations of the density with periods ~60 min were observed. Variations of the temperature were also observed. The relative variations in the atmospheric parameters were 17-25 %. Even if there was an effect of the eclipse at heights of 106-108 km, it was masked by wave processes in the atmosphere and was less than the variations in the temperature and density. It has been indicated above that variations are a characteristic feature of the atmosphere in the considered height interval. The occurrence of these variations is related to the propagation of internal gravity waves.

[24]  On the same day in France (where the total solar eclipse took place), measurements of the variations in the atmospheric pressure were conducted [Farges et al., 2003]. The results of these observations showed that gravity waves are generated during the eclipse. Their source is located at two altitudes: in the lower atmosphere and at higher levels of the thermosphere. The period of the atmospheric IGW was ~60 min. One can see from our data that IGW with the same period were observed not only during the eclipse on 11 August but during the entire 3-day observation cycle. Therefore we cannot conclude unambiguously that the solar eclipse on 11 August originated the generation of IGW.

3.2. Dawn-Dusk Measurements

Figure 3

[25]  The aim of the next three cycles of measurements was studying the dawn-dusk phenomena. The measurements were conducted on 16-17 August 2000, 15-16 June 2001, and 31 July to 1 August 2002. The studies of the twilight D region on the basis of the 2000 and 2001 observations are described by Belikovich and Benediktov [2002]. We describe below the atmospheric parameters variations at the level of ~100 km. Figure 3 shows the time behavior of the temperature and density of the atmosphere on 16 August 2000 at a height of 99 km from 2000 to 2300 UTC+4. The sunset took place about 2022 UTC+4 and about half an hour later on the ground at an altitude of 100 km, respectively. The dashed curves show the running mean. Figure 3 visually shows the variations in T and r with periods of 30-60 min. The variation value D T/T was changing from 0.04 in the dusk period to 0.14 after it, whereas the disturbance in the density was D r/r 0.12. It is widely known that a generation of IGW may happen in the terminator region [see, e.g., Somsikov, 1983]. We think that the data obtained support this mechanism. We failed to determine the atmospheric parameters in the dawn hours on 17 August. That happened because of the sporadic layer at a height of 98 km and electron concentration irregularities between the sporadic and regular layers.

[26]  The next observation cycle in the periods of dawn and dusk was conducted on 15-16 June 2001. Both before the sunset and after the sunrise, variations of the atmospheric parameters with periods of 20-60 min and variations in the temperature and density of about (0.1-0.25) were observed. It was impossible to obtain reliable data on the atmospheric parameters in the time close enough to the sunset moment on 15 June 2001 because a strong sporadic layer was observed in the studied height interval during several hours. It should be noted that in this period, there was also atmospheric wave activity. The latter was manifested in the variations in the lower boundary of the signal scattered by the sporadic layer. These interesting events are described in detail by Bakhmet'eva et al. (2004).

[27]  Because of the ionospheric conditions of observations we failed to obtain the atmospheric parameters values in the periods close to dawn and dusk in the observation cycle on 31 July to 1 August 2002. However, on these days (both before dusk and after dawn), no time variations of the T(t) and r(t) values were observed unlike in the previous observation cycles. In our opinion, this fact also indirectly shows that the observed in the previous cycles internal gravity waves propagated from the terminator region. On 1 August, 1 hour after the dawn the IGW having been generated near the terminator region were already able to leave the observed region.

Figure 4

[28]  Figure 4 shows the vertical profiles of the temperature (Figure 4a) and density (Figure 4b) between 90 and 110 km for the evening hours on 31 July and morning hours on 1 August 2002. The points and circles show the values of T and r for 1 August and 31 July, respectively. The vertical profile of the density was approximated by the exponential function in the form r = r₀ exp [ -(h-h₀)/H]. Here h₀= 100 km, and r₀ is the density at the h₀ height. The approximated functions are shown by the dashed and dot-dashed curves. Similar curves show the averaged temperature profiles T(h). The thick lines show the T(h) and r(h) dependencies according to the CIRA 86 model. (The model SMI 88 used above gives close values.) One can see in Figure 4 that in the height interval between 95 and 112 km both before and after the dusk the vertical profile of the temperature have a minimum. One can also see that in the morning hours the density was by a factor of 2.7 higher than in the evening hours and the temperature profile was shifted up by a few kilometers. The height of the thermopause (the temperature minimum) was 100 and 103 km in the evening and in the morning, respectively. The value of the temperature within the thermopause was almost the same (120-125 K). The comparison to the data for the fall 1990 [Belikovich et al., 2002b] when the wave atmosphere activity also was low shows that the temperature value within the thermopause in the fall of 1990 was higher than in summer of 2002 (~140 K). Table 1 shows the main information: the time interval of the measurements, the time of sunrise or sunset and also the typical values of the temperature and density (the height of the temperature minimum (pause) and its value). For the sake of comparison, Table 1 shows also the corresponding values for the T(h) and r(h) profiles in the morning hours in the fall of 1990.

[29]  One can see in Figure 4 that the model profiles of the temperature do not show any minimum in the 90-110 km height interval. Only in the lower part of the height interval considered the measured temperature values are close to the model values. Belikovich et al. [2002a, 2002b] presented the results of the temperature measurements in August 1999. Then we observed the T(h) dependence with a maximum in the vicinity of 100 km, the wave activity being high. We suggested that the vertical profile T(h) with a maximum at a height of 100 km is caused by the seasonal variability. However, other explanations are possible. Taking into account that on 10-12 August 1999 both high wave activity in the studied height interval and irregular structure of the E layer were observed, one may assume that there was an inversion of the temperature caused by the mixing due to the transport from the more heated region. However, we consider more probable the distortion of the temperature vertical profile due to the propagation of IGW. The spatial scale (wavelength) of the IGW propagating at the considered heights may exceed our interval of measurements of the atmospheric parameters. Bakhmet'eva et al. (2004) estimated the spatial scales of IGW on the basis of the measurements of vertical motions. The scales were found to be 10-15 km. These scales are close to the height interval of the atmospheric parameters determination. Therefore the vertical profiles of the temperature T(h) and density r(h) could be distorted under IGW propagation.

[30]  The temperature minimum above 100 km may exist both because of the mesopause upward shift in the indicated time moments and due to the possible existence of the second minimum in the 100-110 km interval. Both points of view have their support while analyzing the experimental data. In particular one can see it in the data of the lidar observations at Fort Collins (1990-1999) presented by She et al. [2000].

[31]  In winter and summer the mesopause was observed at 100-102 and 65-87 km, respectively. In spring and fall, there was an intermediate regime when there existed both temperature minima (near 100 and 85 km). The two-level character of the mesopause in summer of 2002 was also noted by Mertens et al. [2004] and She and Von Zahn [1998] (measurements of the kinetic temperature by the SABER device).

[32]  Unfortunately, all these data have been obtained in different time at different latitudes and in different conditions. In particular, the lidar measurements are possible only at night and do not provide good accuracy above 100 km. Vice versa, it is easier to conduct our measurements in the daytime. So we think that currently it is hard to conclude finally on the lower thermosphere structure and specification of this structure is an important task.