Extreme solar events and magnetic storms in October 2003 and...

3. Geophysical Aspects of the VLF Signals Registration on 28-30 October 2003

[30] The prominent solar flare X17.2 which began 28 October 2003 at 0951 UT and reached its maximum at 1110 UT and caused sharp changes in phases and amplitudes of the VLF signals received at the sunlit paths (WB, path 1, and path 2; see Figure 3). The estimates show that the strong X-ray radiation of the flare increased the electron concentration N_e in the D region at geomagnetic latitudes of 40-45^o N by a factor of 20 (WB). At latitudes above 50 ^o the increase in N_e was about a factor of 4 (paths 1 and 2). The EB path was in the nighttime conditions and did not react to this disturbance; however, from 1300 UT to 2200 UT, there was a decrease in the phase as compared to the undisturbed nighttime value on the average by 10 cc, the latter fact corresponding to the increase in N_e by about a factor of 5. This time interval was characterized by moderately disturbed conditions ( Kp and AE were about 4 and 400-600 nT, respectively). It is worth noting that according to the data of the Geotail spacecraft in this period till 2400 UT on 28 October there occurred an increase in the particle density in the solar wind up to a few tens of particles per cubic centimeter [Panasyuk et al., 2004]. That considerably intensified the solar wind pressure on the magnetosphere and might have stimulated precipitation of particles from the radiation belt [Tverskoy, 2004]. The EB path is located between 39^o and 45^o of geomagnetic latitude. Therefore it is natural to relate the observed increase in the electron concentration to injection into the ionosphere of electrons with energies of tens keV from the radiation belt. It is known that the development of auroral activity is accompanied by injection of energetic electrons into the ionospheric D region, the fact being registered by riometers [Isaev and Pudovkin, 1972]. So the shift of the region of auroral activity leads corresponding shift in the region of riometer absorption.

[31] The sudden commencement (SC) of this magnetic storm occurred at 0612 UT on 29 October. The geomagnetic disturbances that followed SC ( Kp=9 and AE up to 5000 nT) were not manifested at the paths at lower latitudes (EB and WB) but influenced path 1 and path 2 located at higher latitudes beginning from 0640 UT (Figures 5c and 5e) the northern end of the paths is located at a geomagnetic latitude of 56^o N. At these paths a phase decrease by about 30 cc (path 1) was detected at the amplitude decrease from 0.5 to 2 dB, these values corresponding to increase in N_e by a factor of 5 when there exists the C layer. The disturbances that followed continued till 1230 UT. The analysis of magnetograms at Moscow station ( F=50.4^o N) shows that the center of the electrojet was located in the vicinity of Moscow or even slightly southward from Moscow. The ionosphere vertical sounding data obtained at Moscow station manifest the same. According to these data the beginning of a strong disturbance is detected at 0730 UT. In this time till 0830 UT any traces of reflected signals were absent, the fact indicating that they were completely absorbed. Thus at this time typical features of high-latitude ionosphere were observed over Moscow.

[32] Thus the N_e increase in the ionosphere D region at path 1 and path 2 provides information on development of auroral activity at geomagnetic latitudes 50^o N. The data obtained at WB and EB indicate that the southward shift of auroral activity in this time interval did not reach the latitude of 46^o N.

Figure 9

[33] A different picture was observed in the dusk UT hours on 29 October at WB and also at path 1 and path 2, where no pronounced disturbances were observed except the quick variations in phase between 1900 UT and 2000 UT and about 2300 UT (Figure 5a). The magnetograms of midlatitude observatory Moscow ( F=50.4^o N) and auroral observatory Lovozero located in the European part of Russia indicate to a small geomagnetic activity about 1830 UT. In the interval 1900-2000 UT the disturbances at these stations sharply increased (up to 2500 nT and ~600 nT at Lovozero and Moscow, respectively) and then decreased to 2100 UT. The magnetograms of Moscow station show that the center of the electrojet was located much northward from Moscow. Thus, in the midnight (in local time) magnetosphere, where Lovozero and Moscow stations were located, the main intensification of auroral and geomagnetic disturbances were observed at auroral and subauroral latitudes (in usual meaning of these terms). Such position of the ionosphere electrojet is confirmed by the photo made from the DMSP satellite over Europe at 2200 UT on 29 October 2003 (Figure 9) (http://science.nasa.gov/spaceweather/aurora/gallery_01oct03.html). Figure 9 shows that the auroral processes were developed in the geographic latitude band approximately from 55 to 61^o N or in the 51-57 ^o band of geomagnetic latitudes. Therefore the center of the electrojet was actually located northward from Moscow. At the same time the data of the GOES 10 satellite (1200 LT of which correspond to 2100 UT) show that the magnetosphere was subjected to a strong compression. So from 1800 to 2300 UT GOES 10 was located outside the magnetosphere, that is, its boundary was located at a distance R<6.6 R_E [Panasyuk et al., 2004]. Because of that, the strong oscillations of the VLF signal phase at WB between 1900 UT and 2000 UT and about 2300 UT evidently contain information on the injection of electrons with energies of a few tens keV from the radiation belt into the ionosphere at geomagnetic latitudes of 41-44^o N. In this region short-time auroral processes and related to them geomagnetic disturbances and probably generation of geomagnetic pulsations might have been registered.

[34] A different picture of development of geophysical processes took place in the dawn LT sector of the magnetosphere, where at this time EB was located and observational means of Institute of Space Physics and Aeronomy (ISPA) and Institute of Solar-Terrestrial Physics (ISTP) were operating [Panasyuk et al., 2004]. The VLF data (Figure 8b) indicate that the midlatitude auroral activity was spread out at this time also to latitudes <45^o N, and the analysis of the observational data of ISPA and ISTP makes it possible to assume that at this time in the eastern sector there was a latitudinal splitting of two regions of auroral processes development.

[35] The idea on formation of two latitude-separated auroral zones is confirmed by the analysis of the events between 1900 UT and 2000 UT. At 1900 UT a sharp intense development of auroral processes and magnetic disturbances began: to 2000 UT the AE index increased from 500 nT up to 4000 nT. The VLF signals at EB responded to this disturbance by a sharp decrease of the phase down almost to the daytime values (Figure 5b). According to our estimates a lowering of the N_e profile in the D region from the nighttime values down to the daytime values occurred at latitudes of 38-45^o N. Such decrease in the phase continued till 2000 UT. Simultaneously an intense emission in the 630 nm line was registered by the photometer at the midlatitude observatory of ISTP ( F=41^o N). At higher latitudes the photometer of the ISPA observatory (Maiga, F=52.5^o N) registered two peaks of the emission intensity between 1900 and 2000 UT. At 2000 UT the emission almost disappeared. The all-sky camera at Zhigansk (about 60^o of geomagnetic latitude) at 1953 UT observed an auroral arc in the northwestern part of the horizon. The magnetograms of the ISPA eastern chain of stations manifest a quick shift of the center of the auroral electrojet northward from geomagnetic latitudes of sim 60^o (Zyryanka station). It is worth noting that this substorm is not seen neither at the ISPA western chain of stations nor at magnetograms of Moscow observatory. Thus there again occurred a splitting of the auroral zone to midlatitude one (detected by VLF signals and the ISTP observatory photometer) and high-latitude one (detected by optical and magnetometer means of ISPA).

[36] Intense increase of auroral luminosity was observed at ISTP till 2200 UT and reached a record value of ~4.3 kR in the 630 nm line. It is worth noting that the location of the ISTP observatory is shifted by 2^o southward from the VLF path WB and is located in the longitudinal band of WB. The signal phase at EB strongly decreased by 20 cc again and had almost the daytime values (see Figure 3), this fact manifesting an intensification of midlatitude auroral processes (at 38-45^o N). The eastern chain of the ISPA stations shows that to 2200 UT the electrojet center again was located at geomagnetic latitudes of about 60^o. The development of auroral processes at these latitudes was registered also by the photographic camera at Zhigansk [Panasyuk et al., 2004]. Thus about 2200 UT evidently in the dawn sector of the magnetosphere, there again existed two separated in latitude regions of auroral processes. The processes at geomagnetic latitudes of 50-52 ^o N were approximately by a factor of 5 more intense than at 1800-2000 UT.

Figure 10

[37] A possibility of meridian splitting of the auroral zone at night UT hours on 29-30 October 2003 is confirmed by the photo made from the DMSP satellite on the same night at 0200 UT over the North America (Figure 10) which clearly shows two separated in latitude regions of auroral processes development (http://science.nasa.gov/spaceweather/aurora/gallery_01oct03.html). One of these regions is centered at geomagnetic latitudes of 56-57^o N, and the other is centered at 60-63^o N. It should be noted that at 0100 UT on 30 October the Dst variation reached the maximum for this storm value of 363 nT (http://swdcwww.kugi.kyoto-u.ac.jp).

[38] Even more convincing proofs of the auroral zone splitting may be presented in our opinion on the basis of the complex analysis of the VLF signals and geophysical situation at night UT hours on 30 October. On this day at 1900-2000 UT distortions of the signal phase were registered. At EB the phase took almost daytime values, whereas at WB strong fluctuations with the amplitude of about 10 cc were observed. At path 1 the phase also was falling down below the daytime values and then increasing by 15 cc.

Figure 11

[39] The analysis of the Scandinavian chain of magnetometers including Borok observatory ( F=55^o N) performed by Kleimenova and Kozyreva [2005] manifests that a substorm develops about 2000 UT. The center of the electrojet of this substorm was located at 62-63 ^o geomagnetic latitude. The development of the next substorm is detected about 2200 UT. The center of the electrojet jumped down to a latitude of 58^o (Figure 11a). The magnetogram of Moscow observatory (50.4^o N) (Figure 11b) shows development of intense disturbances from 1900 UT 30 October to 0300 UT 31 October. The center of the electrojet of theses disturbances was located northward from Moscow. At the same time, the data of the Moscow ionosonde indicate to a formation in the ionosphere in the period from 2100 to 2330 UT of an auroral sporadic E layer, this fact manifesting injection of charged particles into the ionosphere over Moscow. Thus summarizing the results of the analysis of events development at the longitudes of Scandinavia and European Part of Russia, one can state that at about 2100 UT two auroral zones were formed. One zone was located at geomagnetic latitudes 41-45^o, the fact confirmed by the data of VLF signals at WB. The other zone was located at latitudes > 50^o. The break between these zones was located, evidently, at geomagnetic latitudes of 46-48^o.

[40] According to the data of the eastern chain of ISPA stations located at longitudes of the ISTP observatory, from 1900 UT to 2400 UT the center of the electrojet was located at latitudes of Zyryanka observatory (about 60^o N), drifting either southward or northward from it (Figure 11d). According to the data of the photographic camera at Zhigansk the aurora luminosity covered region by 250-350 km southward from the zenith, i.e., down to 56-57^o N. It is worth reminding that in this very longitudinal sector anomalous development of auroral processes was observed at the ISTP observatory at a geomagnetic latitude of 41^o N (Figure 11g). It has a record (for the entire time of observations at the observatory) intensity ~6-10 kR in the 6300 nm line. Thus the development of geophysical processes in the Siberian sector indicates that the splitting of auroral processes occurred at geomagnetic latitudes of 45-50^o N.

[41] Isaev [1962] was the first to draw attention to a possibility of splitting of the auroral zone to two regions. Later, Isaev and Pudovkin [1972] showed that the splitting of the auroral zone is related to the development of the DR current and that the probability of the splitting increases at an increase of the DR current value and intensity of geomagnetic disturbances. It follows from the above presented analysis that we detected such development of the events on 29-30 October 2003 on the basis of the analysis of features in VLF signals propagation at midlatitude paths and complex analysis of auroral and geomagnetic processes. According to Isaev [1962] in the years of solar activity maximum the low-latitude auroral luminosity region is formed at geomagnetic latitudes of 50-53^o N with the maximum of luminosity at 54^o. The DR current is located in this period at 3-3.5 R_E [Isaev, 1962]. The analysis of the auroral zone splitting shows that the splitting region is narrowing with development of the DR currents. According to Isaev and Pudovkin [1972], already at Dst= -(60-70) nT the splitting region is constricted to 2 ^o by latitude. It is worth reminding that in our case, Dst reached values of -400 nT and the DR current was located at 2.0-2.5 R_E. So one may expect a shift of the more low-latitude luminosity region to latitudes <50^o, which fact was actually registered in this work.