Auroral Disturbances During the January 10, 1997 Magnetic Storm

Transcription

1 Auroral Disturbances During the January 10, 1997 Magnetic Storm L. R. Lyons and E. Zesta J. C. Samson G. D. Reeves Department of Atmospheric Sciences Department of Physics NIS-2 Mail Stop D436 University of California, Los Angeles University of Alberta Los Alamos National Labs Los Angeles, CA Edmonton, Alberta, T6G 2E9 Los Alamos, NM Geophys. Res. Lett. Special Section on Non-substorm Geomagnetic Disturbances (in press) February, 2000 Abstract It is well known that intense and frequent auroral-zone disturbances, which are often attributed to substorms, occur during magnetic storms. We examine observations during the January 10, 1997 main phase and find that observed auroral-zone activity was dominated by a combination of global auroral and current enhancements, which are a direct response to solar wind dynamic pressure enhancements, and poleward boundary intensifications, which are localized in longitude and have an auroral signature that moves equatorward from the magnetic separatrix. Poleward and azimuthally expanding regions of auroral activity which accompany substorms are found to contribute significantly less to the observed activity. This suggests that poleward boundary intensifications and dynamic pressure responses may be an important cause of disturbances during periods of enhanced convection such as magnetic storms and convection bays. Introduction In addition to ring current formation and the associated Dst magnetic depression observed at low latitudes, intense and frequent auroral-zone disturbances occur during magnetic storms. These disturbances are often attributed to substorms, which are by far the most well-investigated auroral-zone disturbance. However, recent studies have shown that other types of disturbances, distinct from substorms, can also be very important within the auroral zone. Here we evaluate disturbances during the main phase of the January 10, 1997 magnetic storm. We find that the other types of disturbances contributed significantly more to auroral-zone activity observed during the main phase than did substorms. Substorms are associated with a localized region of auroral enhancement that initiates within the equatorward portion of the nightside auroral oval and then expands poleward and 1

2 azimuthally. This auroral bulge is associated with the substorm current wedge, which includes a reduction in the large-scale cross-tail current and a decrease in lobe magnetic field energy. In addition to substorms, disturbances (here referred to as poleward boundary intensifications, or PBIs ) having an auroral signature that moves equatorward from the magnetic separatrix are often observed. These disturbances occur during all levels of geomagnetic activity and occur repetitively, so that many individual disturbances can occur during time intervals of ~1 hr, [Lyons et al., 1998]. They appear to be associated with localized periods of enhanced tail reconnection [de la Beaujardière et al., 1994] and to be related to bursty flows in the tail [Kauristie et al., 1996; Lyons et al., 1999]. Some PBIs extend equatorward from the poleward boundary of the auroral oval as north-south features [Rostoker et al., 1987; Henderson et al., 1998]. PBIs are far more localized than substorms and are not associated with a large-scale reduction in the cross-tail current. In addition, recent studies of the main phase of the January 10, 1997 storm have suggested that variations in the solar wind dynamic pressure can directly lead to significant auroral-zone activity which is very different from both substorms and PBIs. Shue and Kamide [1998] found that the intensity of ionospheric currents increased strongly with solar wind dynamic pressure while the interplanetary magnetic field (IMF) was very strongly southward. Zesta et al. [2000] found that a strong pulse of enhanced solar wind dynamic pressure lead, without significant time delay, to a global enhancement of auroral-zone ionospheric currents and auroral emissions and to a large global poleward expansion of the auroral oval. Zesta et al. suggested that the ionospheric current enhancement connected to a globally enhanced region 1 field-aligned current system, consistent with the inference of Russell et al. [1994] that sudden increases in solar wind dynamic pressure significantly enhance dayside region 1 currents during periods of southward IMF. Increases in dynamic pressure have also been found to lead to an increase in the strength of the cross-tail current [Russell et al., 1994; Collier et al., 1998]. Observations IMF data for UT on January 10, 1997 are shown in Figure 1. Data is shown from WIND at (87, -58, -9 R E ) GSM and from Geotail which remained mostly within the dayside magnetosheath as it moved from the dawn side to the dusk side. Data is not shown for occasional 2

3 periods when Geotail entered the magnetosphere. The IMF measurements show unusually good agreement for two spacecraft with such large spatial separations. We can thus be quite confident that the IMF measured during this period quite accurately reflects the IMF that interacted with the magnetosphere. A time delay of 19 min was applied to the WIND data, giving good agreement with the time of the strong southward turning of the IMF observed at Geotail at 0500 UT. With this time delay, other IMF features agree in time to within ~5 min. Since Geotail was near the nose of the magnetosphere, we presume that the IMF impacted the dayside magnetopause within a few minutes of being observed at Geotail. The strong southward IMF that followed 0500 UT lead to the storm main phase, which had a minimum corrected Dst of 70 nt at ~10 UT [Jordanova et al., 1998]. Shaded regions in Figure 1 indicate periods considered below of enhanced solar wind density (proportional to dynamic pressure, since the solar wind speed was nearly constant) during the storm main phase. Global images of the auroral zone from the Polar spacecraft show two substorms prior to the start of the storm main phase at 0500 UT but following the initial impact of high-density solar wind at 0106 UT [J. Sigwarth, private communication, 1999]. Both substorms started with a localized night-side auroral enhancement that expanded into a typical substorm bulge. Images which include the first onset [at 0334 UT] can be seen in Tsurutani et al., [1998]; Pi2 pulsations and magnetic X-perturbations initiating at the time of the second onset (0427 UT) can be seen in Figure 2. As can be seen from Figure 1, these typical substorms followed intervals of southward IMF and occurred minutes following the observation of IMF substorm triggers (sustained northward turnings without an accompanying significant increase in B y, which are expected to lead to a reduction in the strength of large-scale convection [Lyons et al., 1997; Blanchard et al., 2000]). From UT, during the storm main phase, the IMF was quite steady. This time interval is ideal for studying geomagnetic disturbances during extended periods of relatively steady, strongly-enhanced convection when IMF substorm triggers are small or absent. To investigate the auroral activity during this interval, we show optical data in Figure 3 obtained by the CANOPUS meridian scanning photometers [Rostoker et al., 1995] from 5-12 UT. The CANOPUS photometers were in an excellent location for monitoring nightside auroral activity during this interval, at least one station being within 2 hr of magnetic midnight until 1015 UT. The 3

4 upper panels show emission intensities as a function of invariant latitude and UT as obtained from a merging of data from the photometers at Gillam (magnetic latitude Λ = 67 ) and Rankin Inlet (Λ = 73 ) along the same magnetic meridian. Emission intensities from Fort Smith (Λ = 68 ), ~1.5 hr in MLT to the west of Gillam, are shown in the lower panels. Intensities are shown for 6300 Å emissions, which result from ~ 1 kev electron precipitation, for 5577 Å emissions, which result from the precipitation of ~ > 1 kev electrons, and for 4861 Å emissions, which result from proton precipitation and are generally most intense within the equatorward portion of the plasma sheet. The 6300 Å emissions also respond to proton precipitation and thus often track the 4861 Å emissions within the equatorward portion of the plasma sheet. As can be seen by comparing the IMF B z trace that is plotted over the 6300 Å emission observations from Fort Smith, this inner plasma sheet region monotonically moved equatorward from UT as the IMF became increasingly southward and the strength of convection increased. After this time, this region remained at unusually low latitudes (Λ ~ 62 ) while strong convection was maintained by an increasingly strong negative IMF B y as the negative IMF B z gradually decreased. Only one poleward moving region of active aurora that initiates well equatorward of the poleward boundary of the auroral oval can be seen in Figure 3, that being in the Rankin Inlet- Gillam observations from UT. Ground Pi2 pulsations and magnetic X decreases in Figure 2 show the onset of this weak substorm at 0635 UT. (Entrances into the magnetosphere prior to this onset prevent possible IMF triggers for this substorm from being identified with the Geotail data.) The photometer data show significant auroral activity at other times, but no indication of the poleward propagating region of active aurora that occurs during substorms. While cloudiness interfered with observations from Gillam after 0700 UT, Fort Smith was at excellent MLT s to detect substorm activity for several hours after 0700 UT. Instead of substorms, activity is dominated by short time-scale bursts of equatorward drifting auroral enhancements that are identified as PBIs. These show most clearly in the 5577 Å emission observations. They can be seen in the Rankin Inlet data from UT, where they initiate at the poleward boundary of the auroral emissions, and they can be seen at Fort Smith from UT as they extend through the plasma sheet. The preponderance of equatorward drifting auroral disturbances, and the lack of poleward expanding regions of substorm aurora, also shows 4

5 very clearly in movies of the entire auroral oval, which are available after 11 UT from the Visible Imaging System s Earth Camera on the Polar spacecraft [Sigwarth et al., 1998]. Figure 2 shows magnetic X observations from ground stations covering a broad range of nightside longitudes. Data from Gillam (GIL) and Fort Smith (FMI) are included. Figure 2 shows that magnetic perturbations associated with the PBIs were often equal or larger than those associated with the substorms. Also, Figure 2 shows significant Pi2 pulsations during the period of PBIs, an association which has been seen previously [Lyons et al., 1998]. The poleward boundary of the 6300 Å emission occurs very near the magnetic separatrix [Blanchard et al., 1997], and in this example this boundary is clear in both the 6300 and 5577 Å emissions from Rankin Inlet. Zesta et al. [1999] found that the poleward boundary of the aurora moved rapidly poleward and then equatorward in direct response to the sharp pressure pulse at 1050 UT, and this motion can clearly be seen in the Rankin Inlet data (at ~0430 MLT) in Figure 3. Based on dual DMSP satellite crossings of the northern polar cap that occurred during the time of this pressure pulse, the pressure pulse caused the area of open, polar-cap field lines to decrease by ~40% as compared to the area that was observed by the DMSP satellites on their previous orbit (~100 min earlier). The brightening of the aurora, which Zesta et al. found to initiate throughout the auroral oval almost simultaneously with the magnetopause impact of the pressure pulse, can also be seen in Figure 3. Solar wind density measured by WIND is plotted in the same panel as the Rankin Inlet-Gillam 6300 Å emissions, and it can be seen that the poleward boundary of the emissions tracks the solar wind density remarkably well throughout the entire main-phase period shown in Figure 3. The only significant exception to this tracking is UT, when the poleward boundary moved poleward with the substorm auroral bulge. This suggests that, in addition to the substorm, the solar wind density had a large effect on the latitudinal width and poleward boundary of the auroral oval during this 8 hr period of strongly enhanced convection. Shue and Kamide [1998] have found that the solar wind density also exerted strong control on the strength of ionospheric currents during this period, and Zesta et al. [1999] found that, unlike substorms, the large ionospheric current enhancement associated with the 1050 UT pressure pulse was global. The ground magnetometer data in Figure 2 show that a global ionospheric current enhancement also occurred during the period of enhanced solar wind density from ~

6 UT. Control of global ionospheric currents by the solar wind density during the period of enhanced convection (after 5 UT) can be seen in the bottom panel of Figure 2, which shows the AL index obtained from 68 high latitude stations (light line) and the solar wind density measured on WIND (heavy line). (The high latitude stations available for this period had excellent coverage of regions of westward ionospheric current. There was little coverage of the MLT region of eastward current [Zesta et al., 2000, Figure 3], so that AU is not shown.) Thus the solar wind dynamic pressure had a major influence on auroral zone activity throughout this period of enhanced convection, strongly affecting the width of the auroral oval and the magnitude of the global ionospheric current system. These affects were observed nearly simultaneous throughout the auroral oval. Global auroral observations are not available for the ~07-09 UT pressure enhancement; however, based on the electrojet response and the response to the 1050 UT pressure pulse, the auroral response at ~07-09 UT was most likely global. Figure 2 also shows that, as with the PBIs, the ionospheric current enhancements associated with the dynamic pressure enhancements were greater than occurred following the three identified substorm onsets. In addition to auroral zone emissions and currents, particles at geosynchronous orbit respond to geomagnetic activity. Energetic electron and proton data from LANL and , and available energetic electron data from LANL , are shown in Figure 4. A typical substorm electron injection is seen following the 0427 UT substorm. LANL , located just after midnight, saw significant flux enhancements a few minutes after the onset, and this flux enhancement was later (~0445 UT) seen with energy dispersion near noon by the other two satellites as the freshly injected electrons drifted azimuthally around the Earth. Injected energetic protons were also seen by these two satellites. A small electron enhancement is seen at ~6 min following the 0635 UT onset, followed at ~0700 by electron enhancements at the other two satellites. At 0654 UT, a brief, sharp increase is seen nearly simultaneously in the electrons and in the and protons, and this is followed by a series of proton flux spikes at on the afternoon side seen most clearly in the highest two energy channels. These spikes, which are noticeably sharper and shorter in duration than the substorm flux enhancements, are also seen at after 0900 when that satellite reached 14 MLT. 6

7 (Superposed on the spikes, the protons fluxes also show a more prolonged increase that initiated at ~1050 UT in response to the large pulse in solar wind dynamic pressure.) Vertical dotted lines in the proton panel Figure 4 identify the beginning of clearly identifiable spikes. Notice that there are several spikes per hour, which is comparable to the occurrence rate of PBIs and very different from substorms. Also, allowing for a 15 min time delay from magnetic drift, the spikes are seen during the same time interval when PBIs observed by the CANOPUS photometers at Fort Smith PBIs extended to Λ < 67, and can thus be expected to have affected geosynchronous field lines. The spikes seen on have a one-to-one correspondence with those seen on This can be seen by the vertical dotted lines in the panel, which are the same as those from ~09-11 UT in the panel but with a 5 min delay to allow for magnetic drift from to These data suggest that the extension of PBIs onto geosynchronous field lines is associated with the proton flux spikes. This is in agreement with Henderson et al. [1998] and Sergeev et al. [1999], who reported transient geosynchronous particle injections in association with north-south auroral structures that extended to near the equatorward boundary of the auroral oval. Discussion and Conclusions Observations during the January 10, 1997 storm main phase, when the IMF remained strongly southward and convection is expected to have remained strong, show auroral activity that was dominated by a combination of global enhancements in direct response to solar wind dynamic pressure enhancements and PBIs. The available observations show far less contribution from substorms during the main-phase. This suggests that the PBIs and dynamic pressure response may dominate geomagnetic disturbances during periods of steady, enhanced convection (e.g., magnetic storms and convection bays), a possibility that warrants further evaluation. Both PBIs and the dynamic pressure enhancements give responses that resemble substorms in that they lead to auroral and electrojet enhancements. However PBIs are much more localized than substorms and do not have the poleward and azimuthally expanding region of active aurora that characterizes substorms. Some PBIs extend to low latitudes and appear to lead to spikes of enhanced particle fluxes at geosynchronous orbit. Since PBIs are believed to be associated with bursty flows in the tail [Lyons et al, 1999], the geosynchronous injections suggest that these flows can penetrate as far 7

8 earthward as geosynchronous orbit [Sergeev et al., 1999]. The auroral-zone response to dynamic pressure enhancement (ionospheric current enhancement and poleward expansion of the region of auroral emissions) is far more global than substorms and, unlike substorms, occurs without appreciable time delay or azimuthal expansion. Also, based on previous work, dynamic pressure increases lead to an enhancement of the cross-tail current and the lobe magnetic field, which is opposite to that which occurs during substorms. Acknowledgments. Research at UCLA was supported by NSF Grant OPP and NASA Grants NAG5-6243, NAG CANOPUS data have been obtained with support of the Canadian Space Agency. Research by J. C. Samson was supported partly by the Natural Sciences and Engineering Research Council of Canada. We thank the WIND processing team for the WIND magnetic field, S. Kokubun for providing Geotail magnetic field data through DARTS at the Institute of Space and Astronautical Science in Japan, and G. Lu for providing the Al-68 index. References de la Beaujardière, et al., Quiet-time intensifications along the poleward auroral boundary near midnight, J. Geophys. Res., 99, 287, Blanchard, G. T., L. R. Lyons, and J. C. Samson, Accuracy of 6300 Å auroral emission to identify the separatrix on the night side of the Earth, J. Geophys. Res., 102, 9697, Blanchard, G. T., L. R. Lyons, and J. F. Spann, and G. D. Reeves, On the predictability of substorms and intensifications following northward turnings of the IMF, J. Geophys, Res., 2000 (in press). Collier, M. R., et. al., Multi-spacecraft observations of sudden impulses in the magnetotail caused by solar wind pressure discontinuities: Wind and IMP 8, J. Geophys. Res., 103, 17, Henderson, M. G., G. D. Reeves, and J. S. Murphree, Are north-south structures an ionospheric manifestation of bursty bulk flows?, Geophys. Res. Lett., 25, 3737, Jordanova, V. K., et al., Effect of wave-particle interactions on ring current evolution for January 10-11, 1997: Initial results,. Geophys. Res. Lett, 25, 2971, Kauristie, K., et al., Study of the ionospheric signatures of the plasma sheet bubbles, in Substorms 3, ESA Publications Division, Noordwijk, p.93,

9 Lyons, L. R., et al., Coordinated observations demonstrating external substorm triggering, J. Geophys. Res., 102, 27,039, Lyons, L. R., et al., Near Earth plasma sheet penetration and geomagnetic disturbances in New Perspectives on the Earth s Magnetotail, ed. by A. Nishida, S. W. H. Cowley, and D. N. Baker, American Geophysical Union, Washington, p. 241, Lyons, L. R., et al., Association Between Geotail Plasma Flows and Auroral Poleward Boundary Intensifications Observed by CANOPUS photometers, J. Geophys. Res., 104, 4485, Rostoker, G., A. T. Y. Lui, C. D. Anger, and J. S. Murphree, North-south structures in the midnight sector auroras as viewed by the VIKING imager, Geophys. Res. Lett., 14, 407, Rostoker, G., et al., CANOPUS-A ground-based instrument array for remote sensing the high latitude ionosphere during the ISTP/GGS program, Space Sci. Rev., 71, 743, Russell, C. T., M. Ginskey, and S. M. Petrinec, Sudden impulses at low latitude stations: Steady state response for southward interplanetary magnetic field, J. Geophys. Res., 99, 13,403, Sergeev, V. A., et al., Development of auroral streamers in association with impulsive injections to the inner magnetotail, Geophys. Res. Lett, 26, 417, Shue, J.-H., and Y. Kamide, Effects of solar wind density on the westward electrojet, in Substorms-4 edited by S. Kokubun and Y. Kamide, Kluwer Academic, Boston, p. 677, Sigwarth, J. B., and L. A. Frank, Optical determination of the polar cap area and open magnetic flux with the Visible Imaging System (abstract), Eos, Trans., AGU, 79(45), Fall Meet. Suppl., F760, Tsurutani, B. T., et al., The January 10, 1997 auroral hot spot, horseshoe aurora and first substorm: A CME loop?, Geophys. Res. Lett., 15, 3047, Zesta, E., et al., The effect of the January 10, 1997 pressure pulse on the magnetosphereionosphere current system, AGU Monograph Magnetospheric Current Systems, 2000 (in press). 9

10 Figures Figure 1. IMF y- and z- (GSM) components from WIND and Geotail and solar wind density from WIND for UT on January 10, WIND UT s have been delayed by 19 min. Figure 2. X component and Pi2 observations from auroral-zone ground magnetometers for UT on January 10, For each station, solid triangles identify the UT of magnetic midnight, and magnetic latitude is given along the right-hand edge of the figure. Main phase time periods of enhanced solar wind dynamic pressure are shaded. Lower panel shows the AL index [G. Lu, private communication, 1999] from 68 high-latitude stations and the solar wind density observed by WIND shifted in UT as in Figure 1. Figure 3. CANOPUS meridian-scanning photometer data for UT on January 10, Figure 4. Geosynchronous energetic particle data for UT on January 10, Main phase time periods of enhanced solar wind dynamic pressure are shaded. 10