K. Sigsbee1, C. Cattell1, R. Lysak1,
F. Mozer2, C. Carlson2, R. Ergun2,
J. McFadden2, K. Tsuruda3, T. Yamamoto3,
T. Okada4, S. Kokubun5, D. Fairfield6,
R. Strangeway7, G. Parks8, M. Brittnacher8
1School of Physics and Astronomy, University of Minnesota, Minneapolis, MN USA
2University of California, Berkeley
4Toyama Prefectural University
5STELAB, Nagoya University
8University of Washington, Seattle, WA
Some of the outstanding questions of magnetospheric physics are associated with the relationship between processes in the plasma sheet and the aurora. No specific signatures of quiet auroral arcs have been identified in the plasma sheet and no definite connections have been made between auroral features and magnetic reconnection in the plasma sheet. It is not clear whether this is due to a lack of experimental data or because there are flaws in current paradigms of magnetospheric convection and substorms [Kennel, 1992].
The answers to these questions cannot be found using measurements from a single satellite or ground station. During January and February 1997, the orbit of FAST [Carlson et al., 1998a] was situated with its 4000 km apogee over the northern auroral zone, while Geotail reached radial distances of up to 30 RE in the magnetotail. The locations of FAST and Geotail, combined with the availability of data from Polar, provided an excellent opportunity to study simultaneous data from the magnetotail and auroral zone. Emphasis was placed on examining the role of low frequency waves in magnetosphere-ionosphere coupling. Observations of low frequency waves in the magnetotail and the auroral zone are well documented, but there have not been any systematic studies of possible relationships between waves seen in these two regions of the magnetosphere.
2. Low Frequency Waves in the Magnetotail
Waves over a broad range of frequencies from Pi 2 pulsations and field line resonances to waves excited by kinetic cross-field current instabilities have been proposed to play a critical role in plasma sheet processes associated with substorms. Higher frequency waves near the lower hybrid frequency have also been found [see e.g., Cattell et al., 1994], but they will not be discussed here.
By performing Morlet wavelet analysis on electric and magnetic field data from the GEOS-2 geostationary satellite, Holter et al.  discovered fluctuations near 3 mHz associated with substorm breakup. These fluctuations were interpreted as oscillations of entire field lines that developed as second harmonic standing waves. Field line resonances similar to those observed in the GEOS-2 data, which correspond to standing shear Alfvén waves on magnetic field lines connecting the northern and southern ionospheres, may be closely related with auroral arcs [Streltsov and Lotko, 1996].
During one substorm when AMPTE/CCE data were available, Takahashi et al.,  reported a large amplitude oscillation of the total magnetic field with a frequency of 0.08 Hz and southward turnings of the magnetic field during most cycles of the oscillation near a radial distance of 8 RE. Possible explanations for these oscillations included wavy current sheet motions, oscillatory motions of the neutral line, and periodically spaced tearing islands. In a survey of AMPTE/IRM data, Bauer et al.  found low frequency waves between 0.15 to 1 mHz related to flapping motions of the magnetotail at distances between 9 and 20 RE from Earth. Neutral sheet oscillations with large compressional components from 8 to 15 mHz accompanied by magnetic field dipolarizations and high speed earthward flows were also observed. Observations of waves in similar frequency ranges have been made by the Geotail satellite. During a substorm on April 26, 1995, a magnetic field dipolarization and high speed earthward flows of 2000 km/sec were observed by Geotail at a radial distance of 12.7 RE and approximately 23:00 MLT [Fairfield et al. 1998]. Waves below 0.1 Hz, including a brief compression in the Pi 2 frequency range at substorm onset, were recorded by Geotail [Sigsbee et al., in progress].
One class of disturbances associated with substorms has been called current disruption events. During current disruption events near the onset of substorm expansions and intensifications, enhancements in wave power from 0.1 to 4 Hz were observed by AMPTE/CCE [Lui et al. 1992]. Current disruption takes place on time scales comparable to or shorter than the ion gyroperiod and is a transient, spatially localized phenomenon [Lui et al., 1996]. The north-south component of the magnetic field often briefly reverses sign during current disruption. Lui et al.  proposed the frequent sign change of the north-south component and fluctuations in the other components were turbulence seen by a spacecraft located in the current disruption site. Takahashi et al.  had also suggested their observations were related to disruption of the tail current sheet, but chose to interpret the data as the signature of a substorm neutral line. Observations of these disturbances have motivated development of the current disruption model of substorms. This model suggests a kinetic cross-field current instability, such as the ion Weibel instability, causes current filamentation and electromagnetic turbulence near and above the ion gyrofrequency [Lui, 1996].
3. Low Frequency Waves in the Auroral Zone
The low frequency electric field fluctuations observed by satellites such as Viking and FAST above the auroral zone are not well understood. Evidence of both spatial and temporal variations has been found, but there is often uncertainty in interpretation of the data. According to Aikio et al., four possible sources of electric field variations in the auroral zone are spatial structures, spatial structures with a parallel potential drop, traveling shear Alfvén waves, and interfering Alfvén waves. Classifying field variations can be difficult since spatial structures observed by a moving satellite can be indistinguishable from Doppler shifted temporal variations. Both spatial and temporal structures can coexist in the same data set. For example, Aikio et al.  found evidence in Viking data for a spatial structure with a parallel potential drop below the satellite and interfering Alfvén waves.
Interfering Alfvén waves may be important in the dynamics of auroral arcs [Lysak, 1993]. Because the Alfvén speed rises dramatically above the ionosphere, Alfvén waves reflected from the ionosphere can become trapped in a resonant cavity, called the ionospheric Alfvén resonator [Polyakov and Rappaport, 1981]. A model of the resonator presented by Lysak  suggests electron inertial effects caused by waves trapped in the ionospheric Alfvén resonator could accelerate suprathermal electrons. Acceleration of particles by Alfvén waves or plasma instabilities driven by currents associated with Alfvén waves could produce ion conics and provide a mechanism for the excavation of the auroral plasma cavity [Lysak, 1998]. Waves trapped in the resonant cavity appear as fluctuations in the electric and magnetic fields from 0.1 Hz to 1 Hz. These fluctuations can be observed on the ground as Pi B pulsations at substorm onset [Bösinger et al., 1981].
The interaction between field-aligned currents and the ionosphere has resulted in the development of magnetosphere-ionosphere coupling models to explain substorm onsets. The formation of the substorm current wedge may occur because the current path through the ionosphere becomes more conductive than the cross-tail current. Models of magnetosphere-ionosphere coupling consider that the ionospheric response to currents generated in the magnetosphere is not instantaneous and takes place after shear mode Alfvén waves communicate the field-aligned currents to the ionosphere [see e.g., Kan et al., 1988; Lysak et al., 1992].
4. Selection of Conjunctions
The models and observations discussed in the previous sections raise many questions about the role played by low frequency (< 1 Hz) waves in substorms. In order to study possible connections between waves in the magnetotail and the auroral zone using data from FAST and Geotail, we selected broad time intervals when Geotail was located in the magnetotail within 3 hours of local midnight. We then focused on shorter intervals during these time periods when FAST was also above the nightside auroral zone.
The Morlet wavelet transform [see e.g., Kumar and Foufoula-Georgiou, 1994 and references therein] was the primary technique used to analyze electric and magnetic field data from FAST and Geotail. Traditional Fourier transform methods are not ideal for impulsive signals, such as the electric field fluctuations associated with substorms. and the aurora. Wavelet analysis overcomes the limitations of Fourier analysis and is an excellent tool for identifying features that occur on different frequency and time scales.
Electron and ion data from FAST were used to study the relation between features seen in the FAST fields data and regions of auroral particle acceleration and precipitation. Ultraviolet images from Polar [Torr et al., 1995] helped provide a more comprehensive picture of the state of the magnetosphere during these events.
5. January 18, 1997 Conjunction
An overview of electric and magnetic field data from Geotail and Polar for the January 18, 1997 conjunction is shown in Figure 1. At the start of the conjunction interval at 03:00 UT, Geotail was located in the northern tail lobe outside of a thinned plasma sheet. Examination of the local plasma densities, as indicated by the Geotail spacecraft potential, revealed a sudden expansion of the plasma sheet around 04:20 UT. Earthward flows of 200 km/sec were also observed at this time. The large scale plasma sheet expansion was followed by a brief encounter with the plasma sheet boundary layer and a second earthward flow burst at 04:30 UT. At this time Geotail was located at 23:46 MLT. Approximately 15 minutes later, the same expansion of the plasma sheet was observed by Polar at 02:31 MLT and an altitude of 6.6 RE, providing a dramatic reminder of the global nature of magnetospheric processes.
The conjunction between FAST and Geotail on orbit 1620 occurred just after a partial thinning of the plasma sheet at 05:30. Using the Tsyganenko 89 field model with a KP of 3, the FAST and Geotail footpoints were traced down to an altitude of 100 km in the ionosphere. During orbit 1620, the Geotail footpoints remained more or less stationary, moving only approximately 0.10 in latitude. FAST came within a few degrees of the Geotail footpoints on this pass through the northern auroral zone. An overview of FAST data with the despun electric field, magnetic field in field aligned coordinates, electron energy and pitch angle spectrograms, and the ion energy and pitch angle spectrograms is shown in Plate 1. The FAST magnetic field data shows signatures of field aligned currents as well as more rapid fluctuations starting at 05:41:40 UT. These fluctuations occurred in the same region as an ion conic and low energy counterstreaming electrons. Images taken by the Polar UVI instrument show the auroral oval was quiet with a persistent recovery remnant located between 700 and 730 geomagnetic latitude and 0 to 0.5 MLT, which corresponds to the location of the counterstreaming electrons observed by FAST.
At the edges of the region of counterstreaming electrons, two short bursts of wave activity near 1 Hz were observed by FAST. Raw data and Morlet wavelet scalegrams are shown in Plate 2. The E/B ratio for these waves was 2.8x104 km/sec, consistent with the Alfvén speed of 3.5x104 km/sec calculated from FAST magnetic field and ion composition data. Similar bursts of waves were also observed in the magnetotail by Geotail throughout the conjunction interval.
Lower frequency oscillations of the fields were also recorded by both FAST and Geotail. Analysis of the Geotail spin averaged (3 second resolution) electric and magnetic field data was performed on an interval beginning several minutes before FAST was in the auroral zone to allow time for propagation of waves from the magnetotail to the auroral zone. Wavelet analysis of the electric and magnetic fields within the plasma sheet from 05:30 to 05:40 shows peaks in the spectra of BX GSE and EY GSE in the Pi 2 frequency range between 0.01 and 0.02 Hz similar to those observed on GEOS-2 by Holter et al. . Oscillations of the electric field at 0.03 Hz were observed by Geotail near 05:36 UT. The E/B ratio for these fluctuations was roughly 500 km/sec, consistent with the estimated Alfvén speed of 600 km/sec. Higher frequency oscillations were observed near 0.05 Hz and 0.1 Hz with E/B ~ 500 km/sec. Geotail traveled at 1 km/sec through the magnetotail, so peaks from 0.03 to 0.05 Hz could represent the transit of the satellite through spatial structures on the order of 20 to 30 km. For comparison, a typical proton gyroradius in the plasma sheet might be on the order of 1000 km. Power is also seen in the frequency range 0.03 to 0.05 Hz and near 0.1 Hz in wavelet scalegrams of FAST spin averaged (5 second sampling rate) fields for orbit 1620. Because FAST was moving at close to 6 km/sec, peaks from 0.03 to 0.05 Hz might represent spatial structures of 120 to 200 km. Spatial structures of this size in the auroral zone should map to structures on the order of 1 RE at Geotail, assuming they are on the same flux tube. Observations of power at the same frequencies by both FAST and Geotail and the lack of agreement between the estimated spatial scales, is consistent with the idea that the variations are temporal, and not due to Doppler shifting of spatial structures.
6. February 3, 1997 Conjunction
Figure 2 shows an overview of the Geotail electric and magnetic field data for the February 3, 1997 conjunction. On this day, Geotail was in the plasma sheet until about 20:45 UT, when thinning of the plasma sheet caused it to move towards the boundary of the tail lobes. Inside the boundary region, Geotail encountered large electric field fluctuations up to 50 mV/m. Strong tailward flows of 500 km/sec were observed between 20:55 and 20:59 UT. Several bursts of earthward flow were observed starting at 21:00 UT, suggesting the neutral line was initially located earthward of Geotail but later moved tailward. From about 21:00 to 21:40 UT, Geotail was in a thin, active plasma sheet. During this time Geotail made multiple crossings of the current sheet as shown by the large fluctuations in Bx GSE. Polar images from 21:00 to 21:50 UT show intensifications in the auroral zone related to the activity observed by Geotail. Between 21:05 and 21:12 UT, the electric field observed by Geotail became quiet again and the Polar images showed very little activity.
A conjunction between FAST and Geotail occurred during FAST orbit 1800 from 21:22 to 21:30 UT. At this time Geotail was located at 00:01 MLT. Polar images taken at the start of the conjunction show bright regions near midnight and on the dayside in the postnoon sector. Throughout the conjunction, images from Polar showed intense activity in the auroral zone from 21:00 MLT to midnight. Tracing of the FAST footpoints into the ionosphere shows FAST entered the auroral oval near the conjunction region at midnight and skimmed along the poleward edge of the oval. During this conjunction, FAST took 8 minutes to travel across the auroral zone, in contrast to the brief 4 minute encounter at the January 18, 1997 conjunction. This is due both to the position of FAST on orbit 1800 and the greater level of activity. An overview of FAST fields and particle data for orbit 1800 is shown in Plate 3. In the coordinate system used in Plate 3, BY is perpendicular to the magnetic field in the FAST spin plane and positive deflections in BZ correspond to downward currents. On this pass, inverted V structures with electron energies of several keV were observed. FAST traveled through several regions of up and down going electrons [see Carlson et al., 1998b]. Ion conics persist over the whole auroral zone. No ion beams were observed, indicating that a potential drop occurred only above the spacecraft. The spreading of the pitch angle distribution of the ion conic towards 900 near the poleward edge of the oval is indicative of the transit of FAST through a region of local ion acceleration.
Fluctuations in the electric and magnetic fields between 0.01 and 0.1 Hz were observed by Geotail. Some of the peaks in the wavelet scalegrams appear to be related to sudden decreases in the ion gyrofrequency, however the exact meaning of this correlation is unclear, since Geotail was also moving in and out of the current sheet at this time. Unfortunately, comparison with spin fit electric field data (frequencies below 0.1Hz) from FAST during this time was prevented by data gaps. Higher frequency fluctuations from 1 to a few Hz were observed by both Geotail and FAST during this conjunction. Raw data and Morlet wavelet scalegrams of examples of fluctuations in this frequency range are shown in Plate 4.
Waves observed in the magnetotail near 30 RE by Geotail and by FAST above the nightside auroral zone during the two conjunctions studied may represent field line resonances, traveling Alfvén waves, and interfering Alfvén waves. The waves observed by FAST near 1 Hz during the January 18, 1997 conjunction were compared to the model of the ionospheric Alfvén resonator developed by Lysak  using an Alfvén speed profile with maximum of 80,000 km/sec at 6000 km and a value of 28,000 km/sec at 4000 km to match the local Alfvén speed observed by FAST. An eigenmode of the ionospheric Alfvén resonator with E/B ratios close to the FAST observations at 0.9 Hz was obtained for a conductivity of 10.0 mho and perpendicular ionospheric wavelength of 100 km.
The electric and magnetic field oscillations below 0.1 Hz observed by FAST and Geotail on January 18, 1997 are within the frequency range of the field line resonances which can be excited by disturbances of the near-Earth plasma sheet discussed by Streltsov and Lotko . Fluctuations in this frequency range were also observed by Geotail on February 3, 1997. Many other observations of waves below 0.1 Hz have been made in the magnetotail, but there is a great deal of disagreement on how to interpret the data. Some authors have related their observations to field line resonance structures [e.g, Holter et al., 1995], while others have chosen to interpret their data as flapping motions of the magnetotail or as turbulence related to disruption of the cross-tail current. Further studies of conjunctions between FAST and Geotail and more detailed comparisons to various models are needed to clarify this issue.
Acknowledgments. This work was supported by NASA Grants NAG5-3596 for FAST, NAS5-2718 for Geotail, and NAG5-3182 for Polar.
Aikio, A. T., et al., On the origin of the high-altitude electric field fluctuations in the auroral zone, J. Geophys. Res., 101, 27157-27170, 1996.
Bauer, T. M., et al., Low-frequency waves in the near-Earth plasma sheet, J. Geophys. Res., 100, 9605-9617, 1995.
Bösinger, T., et al., Correlations between PiB type magnetic micropulsations, auroras, and equivalent current structures during two isolated substorms, J. Atm. Terr. Phys., 43, 933-945, 1981.
Carlson, C. W., et al., The Fast Auroral Snapshot Mission, Geophys. Res. Lett., in press, 1998a.
Carlson, C. W., et al., FAST observations in the downward auroral current region: energetic upgoing electron beams, parallel potential drops, and ion heating, Geophys. Res. Lett., in press, 1998b.
Cattell, C. et al., Geotail observations of spiky electric fields and low-frequency waves in the plasma sheet and plasma sheet boundary, Geophys. Res. Lett., 21, 2987-2990, 1994.
Fairfield, D. H., et al., Geotail observations of substorm onset in the inner magnetotail, J. Geophys. Res., 103, 103-117, 1998.
Holter, O., et al., Characterization of low frequency oscillations at substorm breakup, J. Geophys. Res., 100, 19109-19119, 1995.
Kan, J. R., L. Zhu, and S.-I. Akasofu, A theory of substorms: onset and subsidence, J. Geophys. Res., 93, 5624-5640, 1988.
Kennel, C. F., The Kiruna Conjecture: the strong version, Proceedings of the First International Conference on Substorms (ICS-1), Kiruna, Sweden, 1992.
Kumar, P., and E. Foufoula-Georgiou, Wavelet analysis in geophysics: an introduction, in Wavelets in Geophysics, E. Foufoula-Georgiou and P. Kumar (eds.), Academic Press, Inc., 1994.
Lui, A. T. Y., et al., Current disruptions in the near-Earth neutral sheet region, J. Geophys. Res., 97, 1461-1480, 1992.
Lysak, R. L., The relationship between electrostatic shocks and kinetic Alfvén waves, Geophys. Res. Lett., in press, 1998.
Lysak, R. L., Generalized model of the ionospheric Alfvén resonator, in Auroral Plasma Dynamics, R. L. Lysak (ed.), AGU Monograph 80, p. 121, 1993.
Lysak, R. L., J. Grieger, and Y. Song, Fast ionospheric feedback instability and substorm onset, Proceedings of the First International Conference on Substorms (ICS-1), Kiruna, Sweden, 1992.
Polyakov, S. V., and V. O. Rappaport, Ionospheric Alfvén resonator, Geomag. Aeronomy, 21, 816, 1981.
Sigsbee, K., et al., FAST-Geotail correlative studies of magnetosphere ionosphere coupling in the nightside magnetosphere, Geophys. Res. Lett., in press, 1998.
Streltsov, A., and W. Lotko, The fine structure of dispersive, nonradiative field line resonance layers, J. Geophys. Res., 101, 5343-5358, 1996.
Takahashi, K., et al., Disruption of the magnetotail current sheet observed by AMPTE/CCE, Geophys. Res. Lett., 14, 1019-1022, 1987.
Torr, M. R., et al., A far ultraviolet imager for the International Solar-Terrestrial Physics Mission, in The Global Geospace Mission, ed. by C. T. Russell, Kluwer Academic Publishers, Boston, 1995.