3. Results of the Experiment and Simulation

Figure 2

[8]  It has been already mentioned that the experiment was carried out on 18-22 August 2003. This period was characterized by a complicated geophysical situation. Figures 2a and 2b show the time behavior of the geomagnetic indices Dst and Kp, respectively, for the period of the experiment (the data have been obtained from the World Data Center (http://www.wdc.ac.uk)).

Figure 3

[9]  The observation intervals are shown in Figure 2e by the cross-hatched rectangles. According to the common classification the storm sudden commencement (SSC) in the form of a sharp increase of Dst at 1400 UT on 17 August 2003 is shown in Figure 2 by vertical line. The maximum value of the Dst variations amplitude equal to -108 nT was registered at 1500 UT on 18 August 2003. On 19-20 August 2003 the Dst variations lied within 0-35 nT. Then on 21 and 22 August the Dst value began to decrease again but slightly, with the maximum Dst amplitude equal to 18. The maximum values of the Kp index during the magnetic storm were 6-7. The detailed picture of the behavior of the horizontal ( H ) and vertical ( Z ) component of the magnetic field and the declination D according to the data of the Tromsö ( j = 69.66^oN, l = 18.9^oE) station for 17-22 August 2003 are shown in Figure 3 (the data have been obtained from http://geo.phys.uit.no/). The beginning of the magnetic disturbance at 1430 UT on 17 August is clearly manifested by the sharp jump of H, Z, and D components of the magnetic field. After that, deep variations of the magnetic field were observed till 0500 UT on 19 August. Later, they changed to relatively smooth variations in its components till 1300 UT on 20 August and then again an increase of strong variations in the geomagnetic field occurred and continued till 23 August.

[10]  The magnetic storm in a significant way influenced the conductions of the experiment and its results. For example, from 1600 to 2000 LT on 18 August the critical frequencies of the ionosphere in the SURA facility location were below a frequency of 4.3 MHz, which is the least working frequency of the facility emission. That is why on this day the facility did not operate, and in the reception points, HF signals were received only in the presence of natural ionospheric disturbances. It is known from previous observations that during a magnetic storm, there occurs a broadening of the auroral oval into the midlatitude region. Thin curves in Figure 1 show the position of the equatorial boundary of the auroral oval in the Eurasian longitudinal sector for various values of the Kp index (the data have been obtained from http://sec.noaa.gov). This event was accompanied by an increase in the intensity of the total electron content variations [Afraimovich et al., 2003], an increase in the intensity of the small-scale irregularities at middle latitudes responsible for the backscatter of radio waves [Zherebtsov et al., 2002], a depletion of the f_oF2 critical frequency, and an increase of the short-wave absorption in the lower ionosphere [Brunelly and Namgaladze, 1988].

[11]  The ionospheric effect of the magnetic storm is clearly seen in Figure 2c where the deviations of the f_oF2 critical frequency ( d f_oF2(%) = 100 [f_oF2 - {f_oF2}]/{f_oF2} ) measured at Khabarovsk station on 17-22 August 2003 from the quiet mean values for 11-16 August 2003 are shown. Figure 2c shows that the first phase of the f_oF2 negative disturbance began at 0700 UT on 18 August and the decrease in the critical frequency reached 30-37% in the afternoon of 18 August and in the morning of 19 August. The time of the delay of the maximum deviation of d f_oF2 in the main phase of the magnetic storm on 18 August relative to the maximum magnitude of the Dst index ( -108 nT) was 4 hours. The recovery phase began approximately at 2000 UT on 19 August and lasted to 0900 UT on 21 August. Then a secondary strong negative disturbance began with the maximum deviation d f_oF2 reaching 45-49%. The decrease of the maximum usable frequency (MUF), increase in the absorption, and scatter of radio waves during the magnetic storm in the strongest way were manifested at the oblique LFM sounding paths Khabarovsk-Nizhny Novgorod, Irkutsk-Nizhny Novgorod, Khabarovsk-Rostov on Don, and Irkutsk-Rostov on Don. The MUF at these paths decreased by 25-30% and the lower usable frequencies (LUF) increased and that led to a considerable narrowing of the frequency range of the decameter wave channel as compared to quiet conditions.

[12]  The most visual effect of the AIT influence on the remote HF propagation was obtained on 19 August at the Irkutsk-SURA-Rostov on Don path. It was manifested in the appearance of an extra signal at frequencies above the maximum observed frequency (MOF) of the F region in the frequency band ~3-4 MHz during the operation of the heating facility. It is important to note that in this period fairly strong signal was received because of the reflection from the E_s layer with the maximum observed frequency of ~23-24 MHz. Its appearance could have also been related to ionospheric effects of the storm development. Figures 4a-4d show the sequence of ionograms at the Irkutsk-Rostov on Don paths obtained at 1720-1720 UT on August 19 during the facility operation (with the beginning at 0 and 10 min) and the pause (with the beginning at 5 and 15 min).

Figure 4

[13]  The extra signal appearing during the facility operation is marked in Figures 4a and 4b by SS (scattered signal). Figure 4 shows that SS was observed in the frequency interval ~14-18.5 MHz, whereas MUF at the propagation via the ionospheric F region by the standard hop way was ~14 MHz. In the pause, there was no scattered signal because its occurrence is due to the radio wave scatter at artificial small-scale irregularities, their lifetime in quiet ionospheric conditions not exceeding a few tens of seconds [Erukhimov et al., 1987]. The ionograms with the scattered signal during the SURA operation were observed from 1700 to 2000 UT on 19 August. On 21 August when the E_s layer was absent at the Irkutsk-Nizhny Novgorod and Irkutsk-Rostov on Don paths, the scattered signal was observed during the heating facility operation at the frequencies below MUF for the F region. This situation is shown in Figures 4e and 4f during the facility operation and in the pause, respectively.

[14]  During the operation of the LFM sounders at Khabarovsk and Inskip, no extra (scattered) signals (the appearance of which might have been related to the SURA facility) were observed. The most probable cause of that is a small power of these transmitters (200 W in Khabarovsk and 100 W in Inskip) as compared with 1.5 kW of the Irkutsk transmitter.

[15]  This factor is especially significant in observation periods with magnetic and ionospheric disturbances when the radio wave absorption at oblique sounding paths increases considerably and the intensity of the scattered signal is below the sensitivity threshold of the receiving equipment.

Figure 5

[16]  The diagnostics of AIT by the aspect scatter of radio waves was conducted using the bistatic Doppler HF radar at the Moscow-SURA-Rostov on Don path with the reception in Rostov on Don of the standard signals of the precise time RVM signals at a frequency of 9996 kHz. The method of registration and processing of Doppler spectra was described by Vertogradov et al. [1994]. Figure 5 (left) shows examples of the variations in the frequency Doppler shift during the operation of the heating transmitter and after its switching off in the conditions of relatively quiet ionosphere (19 August) and during the magnetic storm (20-22 August). The direct and scattered signals are marked as DS and SS, respectively. Figure 5 (right) shows the dependence of the amplitudes of the direct and scattered signals on time during the heating and after its end. The dashed vertical lines in Figure 5 (right) show the moments of switching off of the heating transmitter (300 s from the session beginning). Figure 5 (left) shows that the sonograms of the Doppler spectra of the scattered signals obtained in different geophysical conditions differ significantly from each other. For example, in the session at 1900 UT on 19 August (Figure 5a) the Doppler frequency shifts for the direct and scattered signals are close to each other: the difference is ~1.2 Hz, so the Doppler spectrum of DS is overlapped on the Doppler spectrum of SS. In this case the dependence of the amplitudes on time helps in separating the direct and scattered signals. Figure 5a (right) shows that the amplitude of the scattered signal decreases after the switching off the SURA facility, this fact being caused by the relaxation of scattering irregularities. A similar picture is seen for other sessions shown in Figures 5b-5d (right). Figures 5b-5d (right) show that during the 3 min after the switching off of the heating transmitter the scattered signal did not disappear completely. The latter shows that during geomagnetic disturbances the relaxation time of artificial small-scale irregularities increases. Three minutes after switching off the heating transmitter the sounding transmitter RVM was switched off. That is why we could not follow the evolution of the scattered signal till its complete relaxation after switching off the heating transmitter.

[17]  In the period of the magnetic substorm on 20-22 August the Doppler shift of the frequency (DSF) of the scattered signal reached values of ~8-10 Hz (see Figure 5c, left). It is worth noting that in the magnetically disturbed period sporadically appearing trains of quasiperiodic modulation of DSF of the scattered signal with a period of ~40-60 s and amplitude reaching 2 Hz (see Figures 5b-5d, left) were observed. In some cases both general increase of DSF (Figure 5b, left) and its depletion (Figure 5d, left) were observed. No quasiperiodic variations were observed in the Doppler spectrum of the scattered signal on a relatively quiet day (19 August, see Figure 5a). It should be noted that no such variations were registered also in the Doppler spectrum of the direct signal in all the days of observation.

[18]  Using the Doppler measurements for the bistatic location of the HF radar, one can determine the drift velocity of the irregularities responsible for the aspect scattering of radio waves in the direction orthogonal to the magnetic field along the bisecting line of the angle formed by the directions from the scatter region to the transmitter and receiver of the sounding signal [Uryadov and Ponyatov, 2003]:

(1)

where l is the wavelength, D F_D is the Doppler frequency shift, and q_s is the scattering angle.

Figure 6

[19]  For the given geometry of the location of the radar, depending on the sign of the Doppler frequency shift either southwest (an azimuth A 234^o) or northeast ( A 54^o) components of the irregularities drift velocity were measured. The estimates show that in the relatively quiet ionosphere (on 19 August) the northeast component of the irregularities velocity was measured and was found equal to ~20 m s^-1. During the magnetic storm (20-22 August) the measured direction of the drift velocity was changed to the southwest one and the velocity magnitude increased up to 186 m s^-1, that is up to the values typical to the high-latitude ionosphere [Boguta et al., 1989; Scali et al., 1995]. The evaluations of the electric field based on the velocity of the E B drift show an increase of the values of the electric field in the upper ionosphere from ~1 mV m^-1 in quiet conditions up to ~8.6 mV m^-1 during the magnetic storm. The analysis of the experimental data was performed taking into account simulation of radio wave propagation and their scattering at magnetically oriented irregularities concentrated in the region of the interaction of the pumping wave to the ionospheric plasma. The simulation showed that the key role in appearance of the extra signal at frequencies above MUF of the F region at the Irkutsk-SURA path was played by a strong sporadic E layer. Figure 6 shows a synthesized ionogram at the Irkutsk-SURA-Rostov on Don path calculated with the help of the International Reference Model (IRI) taking into account the Es layer at the Irkutsk-SURA path (this part of the path was controlled by the reception of LFM signals in Nizhny Novgorod where ionograms with the propagation mode via E_s were also registered) and the scatter at artificial small-scale magnetically oriented irregularities generated in the upper ionosphere by the powerful emission of the SURA facilities.

Figure 7

[20]  Comparison of Figures 4a, 4b, and 6 shows their good agreement. The calculations of the ray trajectories (see Figure 7) in the presence of the sporadic E layer show that the trajectories propagating at the first hop via E_s at frequencies above MUF for the F layer are realized at the Irkutsk-SURA path. At the second hop, part of the radio wave energy soaks through the E_s layer and reaches heights of the F region over the SURA heating facility. From there the radio waves are descended to the Earth surface due to the aspect scatter at the artificial small-scale magnetically oriented irregularities. It should be noted that the mechanism of radio wave capturing into the IWC at frequencies above MUF of the F region by other processes (in particular, by the negative gradient of the electron concentration ne) is not realized at this path. The latter is confirmed by the data obtained on 21 August when no E_s layer was observed and in the absence of considerable horizontal gradients of ne the scattered signal at the Irkutsk-SURA-Rostov on Don path was observed at the frequencies below MUF (see Figure 4e).