Flux transfer events simultaneously observed by Polar and Cluster: Flux rope in the subsolar region and flux tube addition to the polar cusp

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113,, doi: /2007ja012377, 2008 Flux transfer events simultaneously observed by Polar and Cluster: Flux rope in the subsolar region and flux tube addition to the polar cusp G. Le, 1 Y. Zheng, 2 C. T. Russell, 3 R. F. Pfaff, 1 J. A. Slavin, 1 N. Lin, 4 F. Mozer, 4 G. Parks, 4 M. Wilber, 4 S. M. Petrinec, 5 E. A. Lucek, 6 and H. Rème 7 Received 1 March 2007; revised 7 September 2007; accepted 25 September 2007; published 12 January [1] In this paper we present observational evidence of a flux transfer event observed simultaneously at low-latitude by Polar and at high-latitude by Cluster. This event occurs on 21 March 2002, when both Cluster and Polar are located near local noon but with a large latitudinal separation. During the event, Cluster is moving outbound from the polar cusp to the magnetosheath and Polar is in the magnetosheath near the equatorial magnetopause. The observations show that a flux transfer event occurs between the equator and the northern cusp. Polar and Cluster observe the FTE s two open flux tubes: Polar encounters the southward moving flux tube near the equator and Cluster encounters the northward moving flux tube at high latitude. The low-latitude FTE appears to be a flux rope with helical magnetic field lines as it has a strong core field and the magnetic field component in the boundary normal direction exhibits a strong bipolar variation. Unlike the low-latitude FTE, the high-latitude FTE observed by Cluster does not exhibit the characteristic bipolar perturbation in the magnetic field. However, the plasma data clearly reveal its open flux tube configuration. It shows that the magnetic field lines have straightened inside the FTE and become more aligned to the neighboring flux tubes as it moves to the cusp. Enhanced electrostatic fluctuations have been observed within the FTE core, both at low and high latitudes. This event provides a unique opportunity to understand high-latitude FTE signatures and the nature of time-varying reconnection. It shows that existing FTE models cannot accommodate all the features in global observations, and coordinated measurements from largely spaced multiple spacecraft place important constraints which are crucial to the development and refinement of FTE models. Citation: Le, G., et al. (2008), Flux transfer events simultaneously observed by Polar and Cluster: Flux rope in the subsolar region and flux tube addition to the polar cusp, J. Geophys. Res., 113,, doi: /2007ja Introduction [2] The phenomenon known as flux transfer events (FTEs) is widely accepted as the manifestation of timedependent reconnection [Russell and Elphic, 1978]. Near the low-latitude magnetopause, the signature of an FTE is generally most discernible in the magnetic field along the expected normal to the boundary, which is a strong bipolar 1 Heliophysics Sciences Division, NASA Goddard Space Flight Center, Greenbelt, Maryland, USA. 2 Applied Physics Laboratory, Johns Hopkins University, Laurel, Maryland, USA. 3 Institute of Geophysics and Planetary Physics, University of California, Los Angeles, California, USA. 4 Space Sciences Laboratory, University of California, Berkeley, California, USA. 5 Space Physics Laboratory, Lockheed Martin Advanced Technology Center, Palo Alto, California, USA. 6 Department of Space and Atmospheric Physics, Imperial College London, London, UK. 7 Centre d Etude Spatiale des Rayonnements, Toulouse, France. Copyright 2008 by the American Geophysical Union /08/2007JA variation outward then inward or vice versa. Other components of the field vary in concert but with a more varied signature. The characteristic bipolar magnetic field signature can be readily explained by the motion of an open flux tube produced by transient reconnection [Russell and Elphic, 1978]. A northward moving flux tube produces a bipolar signature with normal polarity (outward/inward or +/ ) and a southward moving flux tube with reverse polarity (inward/outward or /+) [Berchem and Russell, 1984; Rijnbeek et al., 1984]. Studies of FTEs observed at the low-latitude magnetopause, which are identified solely based on their bipolar signatures, have confirmed their association with magnetic reconnection as supported by magnetic field, plasma and energetic particle observations [e.g., Berchem and Russell, 1984; Daly et al., 1984; Rijnbeek et al., 1984; Thomsen et al., 1987; Farrugia et al., 1988; Kuo et al., 1995; Kawano and Russell, 1997; Le et al., 1999]. As a result, the characteristic bipolar magnetic signatures have been used as the working definition to identify FTEs in the data. [3] Since FTEs were first discovered, several models have been proposed to explain the nature of the time- 1of13

2 varying reconnection, including original patchy reconnection [Russell and Elphic, 1978], bursty reconnection along a single extended X-line [Scholer, 1988; Southwood et al., 1988; Ku and Sibeck, 1997], and multiple X-line reconnection [Lee and Fu, 1985; Fu and Lee, 1985; Sonnerup, 1987; Ku and Sibeck, 2000]. These models invoke time-varying reconnection on the magnetopause with X-line(s) of different locations and/or spatial extents. Although the magnetic field topology within the FTE flux tubes is different in different models, the predicted local signatures overlap: they all produce characteristic bipolar magnetic field signature. Thus it is difficult to test these models and to distinguish different reconnection mechanisms using local studies. As a result, our knowledge about FTEs remains poor. We do not understand the magnetic topology of the FTE flux tube, its evolution as it propagates on the magnetopause, or its global connectivity and ionospheric signatures. It is of great interest to study the time-varying reconnection using observations from multiple spacecraft that are separated by large distances (much larger than 1 R E, the scale size of the FTE flux tube). [4] While the FTEs observed in the low-latitude subsolar region have consistently exhibited the characteristic bipolar signature in the magnetic normal component, it is not immediately obvious that the same is true at high latitudes. Since data from the Cluster spacecraft become available, several studies have been focused on FTEs at high latitudes to understand their structures and generation mechanisms [e.g., Lockwood et al., 2001; Wild et al., 2001, 2005; Owen et al., 2001; Zong et al., 2005; Thompson et al., 2004; Fear et al., 2005; Dunlop et al., 2005; Asikainen and Mursula, 2006]. For the high-latitude FTEs observed in these studies, many indeed exhibit classical bipolar signatures resembling the low-latitude ones [Wild et al., 2001, 2005; Dunlop et al., 2005; Asikainen and Mursula, 2006] or unbalanced bipolar signatures [Owen et al., 2001]. However, in some other cases, the high-latitude magnetic signatures do not completely resemble the low-latitude ones even though there are indicative FTE characteristics in the simultaneous plasma or ground-based observations. For example, during a close conjunction between the Cluster spacecraft and the EISCAT Radar station, Lockwood et al. [2001] have found clear evidence of an FTE in the ground-based observations. These ground signatures include transient erosion of cusp to lower latitudes, transient and traveling enhancements of the flow into the polar cap, and poleward moving events into the polar cap. The coordinated in situ observations from Cluster equatorward of the polar cusp have showed transient magnetic field perturbations and an enhanced core field with a scale size of 1 R E. However, there is no bipolar signature in any one of the three magnetic field components. In another case of FTEs observed poleward of the cusp in the northern lobe, the transient magnetic signature appears to be a series of periodic reversals in the GSM B z component and again does not resemble those at low latitudes [Thompson et al., 2004]. [5] To our knowledge, there are only a few reports to date on the FTEs simultaneously observed by two spacecraft separated by a large distance. Elphic and Southwood [1987] have reported a case when FTEs are observed by two spacecraft with much different latitudes, AMPTE UKS in low-latitude northern hemisphere and ISEE in midlatitude Figure 1. The spacecraft orbits in the (top) noon meridian plane (XZ plane) and (bottom) equatorial plane (XY plane). The Polar orbit track is in black. The Cluster orbit track is in green. The separation distances among the three Cluster spacecraft with data available (C1 in black, C3 in green, and C4 in red) are amplified by a factor of 100. southern hemisphere. The bipolar signatures of the FTEs at UKS exhibit the normal polarity associated with northward moving flux tubes. The FTEs at ISEE show the reverse polarity associated with southward moving flux tubes. This work provides the first observational evidence of the pair of open flux tubes produced by time-varying reconnection. Wild et al. [2005] have reported FTEs observed simultaneously by Cluster at high latitude and Geotail at low latitude. At high-latitude postnoon magnetosheath, Cluster observe a series of bipolar magnetic perturbations in the boundary normal direction, each associated with an en- 2of13

3 Figure 2. Overview of the Polar observations at low latitudes showing data from the Magnetic Field Experiment (MFE) and the electric field investigation (EFI). A low-latitude FTE observed by Polar is marked by the shaded interval. hancement in the total field strength and an enhanced energetic electron flux. During the same time interval, Geotail, located at the low-latitude southern hemisphere and near local noon, also observe a series of FTEs characterized by the bipolar magnetic field perturbations. By applying these multipoint observations to a model of the motion of open flux tubes, they are able to infer that the reconnection site is located at near equatorial latitude, between Cluster and Geotail. In another observation using Cluster and Double Star data, Dunlop et al. [2005] report a case where Cluster is in the northern high-latitude region and Double Star TC-1 spacecraft south of the equator, and both spacecraft see a series of FTEs. Their results provide further observational evidence that time-varying reconnection produces pairs of open flux tubes moving in opposite directions. [6] In this paper, we present observational evidence of a flux transfer event observed simultaneously near the lowlatitude magnetopause by Polar and the high-latitude cusp by Cluster. This event occurs on 21 March 2002, when both Cluster and Polar are located near local noon but with a large latitudinal separation: Cluster constellation is moving outbound from the polar cusp to the magnetosheath, and Polar is in the magnetosheath near the equatorial magnetopause. This event provides us a unique opportunity to understand high-latitude FTE signatures and the nature of time-varying reconnection. In the next two sections, we will present the observations. 2. Event Overview [7] Figure 1 shows the spacecraft orbits in the noon meridian plane (XZ plane, top) and equatorial plane (XY plane, bottom). The black traces are the orbit track of Polar, which was near the subsolar magnetopause moving from southern to northern hemispheres in its skimming orbit. The green traces are the orbit track of Cluster 3. Since the data are only available from three of the four Cluster spacecraft, 3of13

4 Figure 3. Overview of the Cluster 1 observations at high latitudes showing data from Cluster Ion Spectrometer-Hot Ion Analyzer (CIS-HIA) and Fluxgate Magnetometer (FGM). Cluster 1 (C1), Cluster 3 (C3), and Cluster 4 (C4), the Cluster triad is shown with the separation distances amplified by a factor of 100. The separation distance is 100 km between C1 and C3 and 110 km between C1 and C4. The magnetopause in Figure 1 is based on the shape of the empirical model by Petrinec and Russell [1993]. [8] Figure 2 is an overview of the Polar observations at low latitudes showing data from the Magnetic Field Experiment (MFE) [Russell et al., 1995] and the Electric Field Investigation (EFI) [Harvey et al., 1995]. Unfortunately, this event occurs during a period of spacecraft orbit maneuver and data from the plasma instruments are not available. In Figure 2, both the magnetic field and electric field data are averaged to 6 s resolution. For this interval, the electric field data along the spacecraft spin axis is perturbed. The spin axis component of the electric field data is set to zero, and the electric field data from the two spin plane axes are used in the calculation of the GSM components. This mainly affects the GSM Y-component since the spin axis of Polar is approximately normal to the orbital plane. From the magnetic field data, it is clear that Polar went from the magnetosphere to the magnetosheath at 0208 UT and back into the magnetosphere 0346 UT (marked by two dashed lines). This is supported by the plasma density inferred from the spacecraft potential shown in Figure 2 (bottom). The magnetic field in the subsolar magnetosheath fluctuates between north and south directions in the B z component, and is predominantly northward with a duskward B y component for the entire magnetosheath interval. During this interval, ACE is located 221 R E upstream and 20 R E from the Earth-Sun line and Wind 56 R E upstream and 97 R E from the Earth-Sun line. Despite the fact that both ACE and Wind have returned interplanetary magnetic field (IMF) data during this interval, neither of the two IMF data sets shows good correlation with the subsolar magnetosheath field in terms of the clock angle (the angle in the YZplane) with a convection time delay nor do they correlate with each other within this time range. Thus neither ACE nor Wind data represent the near-earth IMF conditions for this interval due to the spacecraft s large distance from the 4of13

5 Earth-Sun line. Since Polar is in the subsolar magnetosheath, we can infer that the near-earth IMF fluctuated between the north and south direction in a similar way as the subsolar magnetosheath field did. [9] In the Polar magnetic field data in Figure 2, the outstanding feature is an isolated bipolar perturbation in the B x component at 0242 UT. Since Polar is near the subsolar magnetopause, the X-direction is mainly aligned to the magnetopause boundary normal direction. The bipolar perturbation is also accompanied by a strongly enhanced total magnetic field strength and decreased plasma density. They are clearly classical FTE signatures. The peak-to-peak perturbation of the FTE is greater than 80 nt, ranking it as one of the largest FTEs ever published. The duration of the FTE is about 3 min. The magnetosheath magnetic field is southward immediately before the FTE. [10] Figure 3 is an overview of the Cluster 1 observations at high latitudes showing data from Cluster Ion Spectrometer-Hot Ion Analyzer (CIS-HIA) [Rème et al., 2001] and Fluxgate Magnetometer (FGM) [Balogh et al., 2001] instruments. From top to bottom, Figure 3 shows the ion energy spectrogram, the ion density, the ion bulk velocity, the ion temperature, the magnetic field components in GSM, and the total magnetic field strength. Combining with the Cluster orbit trajectory in Figure 1, we can determine that the Cluster spacecraft are initially in the northern tail lobe moving outbound, move into the polar cusp at 0020 UT, cross the magnetopause at 0210 UT, and are in the magnetosheath thereafter. [11] We note that the Cluster magnetopause crossing from the polar cusp to the magnetosheath at 0210 UT occurred nearly simultaneously with the Polar magnetopause crossing from the magnetosphere to the subsolar magnetosheath at 0208 UT. In Figure 1 the model magnetopause scaled by the solar wind dynamic pressure well predicts the subsolar magnetopause position, as Polar is just slightly outside the model magnetopause at 0208 UT. However, Cluster is located significantly inside the model magnetopause at 0210 UT. This is due to the presence of the polar cusp, which causes an indentation of the magnetopause, and is generally not modeled with simple magnetopause models. Owing to the existence of the cusp indentation, the actual magnetopause boundary normal direction may show large deviation from those predicted by the magnetopause models. 3. Observations of the Flux Transfer Event [12] To examine the simultaneous Cluster data around the Polar FTE time, Figure 4 shows an expanded view of the magnetic field data from both the Polar and Cluster 1 spacecraft for a 30 min time interval from 0230 to 0300 UT, during which both spacecraft are in the magnetosheath. The magnetic fields are displayed in GSM coordinates and are both spin-averaged data with time resolutions of 6 s for Polar and 4 s for Cluster. In the Cluster data a transient perturbation stands out due to its temporal proximity to the Polar FTE time. The transient magnetic field perturbation at Cluster is observed about 2 min after the Polar FTE time. The two dashed lines in Figure 4 mark the center time of Polar FTE and Cluster transient signature, respectively. The time delay in the order of 2 min is a reasonable time delay for a signature convecting from the subsolar region to the Figure 4. Simultaneous observations of the magnetic field at (top) Polar and (bottom) Cluster 1 from 0230 to 0300 UT. The dashed lines indicate the center times of the lowlatitude and high-latitude FTEs. The magnetic fields are displayed in GSM coordinates and are both spin-averaged data with time resolutions of 6 s for Polar and 4 s for Cluster. cusp region by the magnetosheath flow. The Cluster perturbation occurs in all three magnetic field components with nt in amplitudes and about 1 min in duration. However, none of the three magnetic field components shows a bipolar signature for this transient perturbation. The Cluster perturbation is unipolar in all three GSM components and thus would be unlikely to exhibit a bipolar signature in the boundary normal direction. This Cluster event would probably not be classified as a flux transfer event due to its lack of a characteristic bipolar signature if using the Cluster data alone. However, the nearly simultaneous Polar observation of a large FTE at low-latitude makes it an interesting event to be further investigated as it is a good candidate for a high-latitude FTE. [13] We now show the internal structure as revealed in the high-resolution magnetic and electric field data as well as plasma data to demonstrate that this Cluster transient event is likely to be the high-latitude signature of the transient reconnection correlated to the low-latitude FTE at Polar. To display the high-resolution Polar data, we rotate the data from GSM into a boundary normal coordinate system which is ordered by the core field in the center of the FTE, or the flux tube axis, as shown in Figure 5. In this system, N is along the magnetopause boundary normal direction; L is on 5of13

6 Figure 5. Boundary normal coordinate system for the low-latitude FTE. The thick arrow is the orientation of the flux tube axis inferred from the Polar data. the magnetopause surface and parallel to the FTE s flux tube axis (at the maximum of the core field); and M is on the magnetopause surface and perpendicular to the FTE s core field to complete the right-handed coordinate system. Since Polar was located very close to the subsolar point, the boundary normal direction is (0.9991, , ) in GSM, which is closely aligned (2.4 ) to the GSM X-axis. The Polar FTE s core field is found to point southeastward, at an angle of 59 from the due southward direction. [14] Figure 6 shows 8 min Polar magnetic field and electric field data around the low-latitude FTE in the boundary normal coordinate system defined in Figure 5. The time resolution of the magnetic field is 0.12 s, which is the highest available; and the electric field displayed here has the same time resolution, obtained by averaging the data with higher resolution (0.025 s or 40 samples/s). The magnetic field inside the FTE has a strong bipolar signature with reversed polarity () and is accompanied by a strong core field. This implies a flux rope structure with helical magnetic field lines within the FTE, and the flux rope appears to move southward based on the polarity. Since Polar was slightly northward of the equator, the reconnec- Figure 6. High-resolution (0.12 s) magnetic and electric field data within the low-latitude FTE observed by Polar. The data are displayed in the boundary normal coordinate system defined in Figure 5. 6of13

7 Figure 7. Power spectra of the electric field fluctuations within the low-latitude FTE (0241: :40 UT in black) and in the neighboring magnetosheath (0238: :40 UT in gray). The Nyquist frequency of the electric field data is 20 Hz. tion site should also be located northward of both Polar and the equator. [15] We will not attempt to make any conclusion on the FTE s E B convection velocity here due to the fact that the Polar electric field data in Figure 6 are calculated only from the measurements of the two spin-plane components. Instead we concentrate on the electric field fluctuations within the FTE. In the core region of the FTE (shaded interval), we observe much enhanced electric field fluctuations with amplitude up to 10 mv/m in high-frequency range (>1 Hz). These high-frequency fluctuations appear to be mainly electrostatic since we do not see evidence of magnetic fluctuations at the same frequencies as the time resolutions for the magnetic field and electric field data are the same in Figure 6. Using the highest-resolution (40 samples/s) electric field data, we show the power spectra of the electric field within the FTE (0241: :40 UT) and in the magnetosheath (0238: :40 UT) in Figure 7. The enhancement of the electric field fluctuations is broadband up to 20 Hz, the Nyquist frequency of the data. The spectral power inside the FTE core region is about two orders of magnitude higher than that in the magnetosheath in the whole bandwidth. [16] Now we examine high resolution data within the high-latitude FTE observed by Cluster. For this particular case, the Cluster magnetic field and electric field data are available at very high resolution. The magnetic field data are taken at the burst mode with a time resolution of s, or 66.2 samples/s. The time resolution for the electric field is s, or 450 samples/s. Figure 8 shows 2 min of these high resolution field data around the high-latitude FTE from the three Cluster spacecraft. Here the data are displayed in the GSM coordinate system since the boundary normal coordinate system is not well known due to the indentation of the cusp and some ambiguities in the result of minimum variance analysis. (The minimum variance analysis of the boundary crossing does not result in a useful boundary normal coordinate system since some low-frequency variations are embedded in the transition layer and the method works best with a thin, well-defined current layer.) Within the high-latitude FTE, there is an enhancement of the core magnetic field strength. However, none of the three GSM magnetic field components exhibits a bipolar signature. These data show that the flux tube within the high-latitude FTE does not have a flux rope s helical magnetic field structure, as in the case of the low-latitude FTE seen by Polar. [17] Although we can not infer the FTE s direction of motion based on the polarity of the bipolar signature, we can determine the direction of motion by examining the relative location of the three Cluster spacecraft and the relative timing of the magnetic signatures at the three spacecraft. The leading edge of the magnetic perturbation arrives at the three Cluster spacecraft in the sequence of C3, C4, and C1 (in green, red, and black, respectively, in Figure 8), which is the same order as the relative location of the three spacecraft from south to north along the Z-direction (see Figure 1). Thus the flux tube of the high-latitude FTE moves mainly northward, in the opposite direction of the motion of the low-latitude FTE flux rope. We can infer that Polar and Cluster likely see two branches of the reconnected flux tubes moving in opposite directions and the reconnection site is somewhere north of the equator between the two spacecraft. [18] Similar to the low-latitude FTE case at the Polar spacecraft, the high-latitude FTE also exhibits much enhanced high-frequency electric field fluctuations up to 10 mv/m. Within the bandwidth of the magnetic field data (up to 33 Hz), we do not see any evidence of enhanced high-frequency magnetic field fluctuations. Thus the electric field fluctuations are likely to be electrostatic, similar to the Polar observations. Figure 9 shows the electric field power spectra for two 20 s intervals, one within the FTE (0243: :59 UT) and the other in the magnetosheath near the cusp (0243: :21 UT). They are calculated from the highest resolution electric field data. The power enhancement of the electric field fluctuations within the FTE is seen for the entire bandwidth but most noticeably in the frequency band from 1 Hz up to the Nyquist frequency of 225 Hz. [19] We now present the plasma energy spectrogram and distribution function data for the high-latitude FTE to demonstrate its open flux tube configuration despite its lack of classical bipolar signature. To show the plasma data more clearly for the high-latitude FTE, Figure 10 is an expanded view of the Cluster 1 CIS HIA spectrogram and moment data for the 30 min interval from 0230 to 0300 UT. Throughout this interval, the plasma is of typical magnetosheath flow except within the high-latitude FTE where an additional plasma population is present. The ions within the FTE flux tube appear to be a mixture of typical magnetosheath ions with energy from a few hundreds ev to a few kev and hot tenuous ions above 10 kev. The intensity of the magnetosheath ion flux within the FTE is somewhat lower 7of13

8 Figure 8. High-resolution magnetic and electric field data within the high-latitude FTE observed by Cluster, showing (top) the magnetic field data from C1 (black), C3 (green) and C4 (red) at s time resolution (66.2 samples/s) and (bottom) the electric field data from C3 (green) and C4 (red) at s time resolution (450 samples/s). The data are displayed in the GSM coordinate system. than that in the magnetosheath, indicative of some loss of the magnetosheath ions. The mixing of the two ion populations results in a higher temperature and a lower density within the FTE in the moment data. The bulk velocity within the FTE is similar to that of the magnetosheath flow, indicating the FTE flux tube is moving with the magnetosheath flow. [20] Figure 11 shows the Cluster 1 CIS-HIA ion distribution functions observed inside and outside the highlatitude FTE. Each distribution function covers a 4 s time interval, or the spacecraft s spin period, centered at the time labeled above Figure 11 (top). The time intervals for the five distribution functions in Figure 11 are also marked by the five bars in the bottom of Figure 8. The distribution functions b and c are for ions within the high-latitude FTE. As a reference, distribution functions a, d, and c are for typical magnetosheath ions in the nearby magnetosheath. Inside the high-latitude FTE, the ion signatures we have observed from Figure 10 become more apparent: two separate ion beams are present. In addition to the cold magnetosheath plasma, a hot magnetospheric ion beam is streaming in the same direction as the magnetosheath flow. It is clear that the flux tube is open allowing mixing of the magnetosheath and magnetospheric ions. When we project the direction of the background magnetic field in the XZplane, in which the magnetosheath flow is largely contained, we can see that the hot magnetospheric plasma is streaming in the direction antiparallel to the magnetic field. This configuration applies to the open flux tube which is connected to the northern ionosphere where escaping magnetospheric plasma flows out in the direction antiparallel to the magnetic field. Thus the high-latitude FTE observed by Cluster is likely to be the northward moving branch of the FTE flux tube, which is formed at the lowlatitude reconnection site somewhere between the Polar and Cluster locations. 4. Discussion [21] The observations presented above suggest that the signatures seen at Polar and Cluster are very likely to be related although their magnetic field signatures are not exactly the same. Even though both Polar and Cluster have spent more than 1 h in the magnetosheath, the fact that only one FTE is detected at both the spacecraft indicates that the FTE only occurs once during this interval. Polar, located 8 of 13

9 Figure 9. Power spectra of the Cluster 3 electric field fluctuations within the high-latitude FTE (0243: :59 UT in black) and in the neighboring magnetosheath (0243: :21 UT in gray). near the subsolar region, observed very strong and welldefined classical FTE flux rope signatures in the magnetic field data despite the fact that the plasma data are not available for this event. The Polar data indicate the passing of a southward moving FTE flux rope through the spacecraft (based on the polarity of the bipolar magnetic field component in the boundary normal direction). Cluster s magnetic signatures at high latitudes, on the other hand, are not as conclusive due to a lack of classical bipolar variations. If time-varying reconnection had taken place somewhere near Polar but between the Cluster and Polar locations producing a pair of flux tubes moving in opposite directions, then the Cluster spacecraft, present at the right time and at the right location, were very likely to observe the passing of the northward moving flux tube. The Cluster plasma observations support this scenario. The Cluster plasma data clearly show that the high-latitude FTE flux tube contains a mixture of cold magnetosheath ions and hot tenuous magnetospheric ions. The magnetospheric ions flowed in the direction antiparallel to the magnetic field, in agreement with the north branch of the open flux tubes whose magnetic field is connected to the northern ionosphere, allowing magnetospheric ions to escape out of the magnetosphere in the antiparallel direction. [22] One could otherwise argue that the mixture of magnetosheath and magnetospheric ions could also be seen due to Cluster going into the magnetopause boundary layer if the oscillation of the magnetopause caused such an encounter. However, this is unlikely to be the case because the Cluster spacecraft are located near the cusp at high latitudes. In Figure 3, it is apparent that the Cluster spacecraft are in the tail lobe (open field lines above the northern cusp) before 0020 UT. From 0020 to 0210 UT, the Cluster spacecraft are in the cusp. After 0210 UT, the spacecraft go into the magnetosheath. The magnetospheric Figure 10. Cluster 1 CIS HIA spectrogram and moment data for the 30 min interval from 0230 to 0300 UT. The shaded interval is the high-latitude FTE. 9of13

10 Figure 11. Cluster 1 CIS-HIA ion distribution functions near the high-latitude FTE with the times marked in Figure 8. ion beams observed in CIS-HIA data (in both the spectrogram and the distribution functions) within 0243: :00 UT are typical ring current ions that are present on the closed field lines equatorward from the cusp. The Cluster spacecraft would have to go into the boundary layer adjacent to the dayside closed field lines to see them. Given the spacecraft location, when the magnetopause position oscillates, the Cluster spacecraft would most likely encounter the cusp or the tail lobe above the cusp but would be unlikely to encounter the boundary layer on closed field lines. [23] This observation and other studies cited in the introduction show that the magnetic signatures of FTEs are quite variable at high latitudes. The classical bipolar signatures, commonly used to identify FTEs are not always preserved at high latitudes. The lack of bipolar magnetic field signature in some FTEs implies that the magnetic field lines have lost their helical structure and become more straightened within the flux tube as the FTE moves toward the cusp. As the magnetic field lines within the FTE become more aligned with the neighboring flux tube in the cusp, the FTE flux tube is being added to the cusp and will eventually become a flux tube in the cusp. We note that the disintegration of FTE s core field in the cusp has been seen a study of FTEs interaction with the cusp using 2.5-dimensional hybrid simulations (N. Omidi and D. G. Sibeck, Flux transfer events in the cusp, submitted to Journal of Geophysical Research, 2008). Raeder [2003, 2006] has developed a global numerical model of the interaction of the solar wind and the interplanetary magnetic field with Earth s magnetosphere to study the formation process of FTEs and the motion of the flux tubes. Using this global model, Wang et al. [2004] have studied the transient magnetic signatures of high-latitude FTEs by placing a series of pseudo-spacecraft at different locations near the polar cusp as an FTE passed by. Unlike the low-latitude FTEs, they find that the magnetic signature of an FTE is quite variable depending on the relative locations of the pseudo-satellite. The boundary normal component of the magnetic field varies from bipolar, to unbalanced bipolar, and to unipolar. Thus high-latitude FTEs are not limited to those with bipolar magnetic field variations. Characterizing FTEs magnetic structures at high latitudes requires physical understanding of the evolution of FTEs and how they interact with the cusp. [24] Flux tube addition to the polar cusp provides a direct solar wind plasma entry mechanism. When the FTE flux tube is added to the polar cusp at high latitudes, the magnetosheath ions within the flux tube gain direct access to the ionosphere. The magnetosheath ions in the cusp can charge exchange with Earth s hydrogen exosphere producing fast neutrals [e.g., Collier et al., 2005]. This may have explained the loss of the magnetosheath ions in the high-latitude FTE flux tube as evident by the decreased density in Figure 10. The generation of fast neutrals by charge exchange between the magnetosheath ions and the exospheric hydrogen also provides a constant free energy source for plasma instabilities in the cusp at various altitudes because these neutrals can become ring beam ions when ionized by the EUV of the Sun. Locally generated ion cyclotron waves have been 10 of 13

11 frequently observed within low-, middle- and high-altitude cusp [e.g., Gurnett and Frank, 1972; Saito et al., 1987; Russell et al., 1971; Scarf et al., 1972; Le et al., 2001; Nykyri et al., 2006]. The interaction between the newly ionized ring beam ions and the cusp plasma provides a possible mechanism for the generation of these waves and should be a subject for future investigations. [25] The common features observed within both the lowand high-latitude FTEs are the enhancement of broadband electric field fluctuations. In our observations the electric field fluctuations are enhanced over the entire instrument bandwidth, up to 225 Hz. These electric field fluctuations also appear to be electrostatic for both the low- and highlatitude FTEs. These observations are very interesting as they relate FTEs with microinstabilities in the reconnection region. It has been a long-standing problem in reconnection theory that the reconnection rate predicted with a classical resistivity [Sweet, 1958; Parker, 1957] is much smaller than that observed in space plasma, such as in solar flares and substorms. This has led to widespread use of anomalous resistivity, and it has been proposed that the development of electrostatic fluctuations around the reconnection X-line can increase the anomalous resistivity. The primary candidate for the electrostatic instabilities is the lower-hybrid drift instability produced by the relative streaming of electrons and ions in the diffusion region [e.g., Davidson and Gladd, 1975; Davidson et al., 1977; Huba et al., 1977, 1978; Silin and Buchner, 2005]. Such instability has been observed near the reconnection sites at the magnetopause [e.g., Labelle and Treumann, 1988; Cattell et al., 1995], in the magnetotail [e.g., Cattell and Mozer, 1987], and in laboratory plasmas [e.g., Yamada et al., 1997; Carter et al., 2002]. Our observations of enhanced electrostatic fluctuations within FTEs provide further evidence that reconnection is responsible for the formation of FTEs. Our study is very preliminary in this aspect, but it points out that microinstabilities will be a very important topic for future investigations in understanding the role of the reconnection in generating FTEs. [26] The coordinated Polar and Cluster observations presented in this study demonstrate the importance of global observations in understanding FTE mechanisms. Despite the fact that all the existing models predict the similar characteristic bipolar magnetic field signatures in the direction normal to the magnetopause, different reconnection mechanisms result in significantly different magnetic topology, magnetic connectivity, and motion of FTE flux tubes [Sonnerup, 1987; Scholer, 1988]. Differentiating these properties requires global measurements both in the magnetosphere (e.g., largely spaced spacecraft) and ionosphere (e.g., ground-based observatory array). Our observations from largely spaced spacecraft place important constraints to the existing FTE models. The flux tube formed in the original patchy reconnection model crosses the magnetopause over the localized region as a slanted hole. Within the flux tube, the magnetic field is twisted around the tube axis, produced by a field-aligned current. However, it is not a flux rope structure. Same is true for the flux tube formed in extended single X-line bursty reconnection models since field lines do not wrap completely around the tube axis. Neither the patchy reconnection nor the bursty reconnection models can explain the flux rope structure with helical magnetic field within FTEs. On the other hand, the flux rope structure is an inherent property for flux tubes produced by multiple X-line reconnection models. In this study, the strong helical structures observed by Polar at low latitudes seem to favor multiple X-line reconnection mechanism. However, in the multiple X-line reconnection process characterized by repeated formation and convection of magnetic islands [Ding et al., 1992], it would be more likely to see multiple FTEs occurring periodically. In addition, the multiple X-line mechanism requires the presence of an eastwest IMF B y component in the formation of magnetic islands by the development of tearing instability at the magnetopause [Lee and Fu, 1985]. The subsequent motion of the two FTE flux tubes would be opposite in the eastwest direction. The two branches of FTE flux tubes would be more likely to be seen at two different local times by two largely separated spacecraft and would be seen away from the local noon at high latitudes. This is not consistent with our observations. In this study the high-latitude FTE at Cluster is observed near the local noon, similar to the longitudinal location of the low-latitude FTE at Polar, a feature more likely to be observed in the patchy or bursty reconnection mechanisms. In the patchy reconnection or bursty reconnection models, the two oppositely moving flux tubes can be observed at similar local times either due to a predominately southward IMF or an extended X-line over a large longitudinal extent. The observations presented in this study show that none of the existing FTE models can accommodate all the features in global observations. Since one cannot distinguish different reconnection mechanisms based on local spacecraft observations alone, coordinated measurements from largely spaced multiple spacecraft are crucial to the development and refinement of FTE models. 5. Summary [27] In this paper, we present observational evidence of a flux transfer event observed simultaneously at low latitude by Polar and at high latitude by Cluster. This event occurs on 21 March 2002, when both Cluster and Polar are located near local noon but with a large latitudinal separation. During the event, Cluster moves outbound from the polar cusp to the magnetosheath at high latitude, and Polar is in the magnetosheath near the equatorial magnetopause. The observations show that a flux transfer event is likely to be initialized at low latitude near Polar but between Polar and Cluster. Polar and Cluster each observe a branch of the FTE s two open flux tubes: Polar sees the southward moving flux tube near the equator, and Cluster sees the northward moving flux tube at high latitude. The lowlatitude FTE appears to be a flux rope with helical magnetic field lines as it has a strong core field and the magnetic field component in the boundary normal direction exhibits a strong bipolar variation. Unlike the low-latitude FTE, the high-latitude FTE does not exhibit the characteristic bipolar perturbation in the magnetic field. Nevertheless, plasma data reveal that the high-latitude FTE contains a mixture of cold magnetosheath plasma and hot tenuous magnetospheric plasma which is commonly seen on closed magnetospheric field lines. The plasma observations within the high-latitude FTE clearly show an open flux tube configuration of magnetic field lines. These observations indicate 11 of 13

12 that the magnetic field lines inside the FTE have straightened and become more aligned to the neighboring flux tubes as it moves to the cusp. Enhanced electric field fluctuations are observed within the FTE core, both at low and high latitudes. These fluctuations are broadband and appear to be electrostatic. The observed electric field fluctuations relate FTEs with plasma microinstabilities in the reconnection region and should be an important topic for future investigations. This event demonstrates the complex nature of the evolution of the FTE flux tube as it moves away from the subsolar magnetopause and provides us a unique opportunity to understand high-latitude FTE signatures and the nature of time-varying reconnection. It shows that existing FTE models cannot accommodate all the features in global observations and coordinated measurements from largely spaced multiple spacecraft place important constraints which are crucial to the development and refinement of FTE models. [28] Acknowledgments. Y. Zheng was a NRC Postdoctoral Research Fellow at NASA GSFC when the work was performed. The work at GSFC was supported by NASA Geospace Science Program. The work at UCLAwas supported by NASA grant NNG05GC87G. The work at UC Berkeley was supported by NASA grant NNG05GC72G. The work at Lockheed Martin was supported by NASA contract NNG05GE15G. [29] Zuyin Pu thanks James Wild and another reviewer for their assistance in evaluating this paper. References Asikainen, T., and K. Mursula (2006), Reconnection and energetic particles at the edge of the exterior cusp, Ann. Geophys., 24, Balogh, A., et al. (2001), The Cluster magnetic field investigation: Overview of in-flight performance and initial results, Ann. Geophys., 19, 1207, SRef-ID: /ag/ Berchem, J., and C. T. Russell (1984), Flux transfer events on the magnetopause: Spatial distribution and controlling factors, J. Geophys. Res., 89, Carter, T. A., H. Ji, F. Trintchouk, M. Yamada, and R. M. Kulsrud (2002), Measurements f lower-hybrid drift turbulence in a reconnecting current sheet, Phys. Rev. Lett., 88, doi: /physrevlett Cattell, C. A., and F. S. Mozer (1987), Substorm-associated lower hybrid waves in the plasma sheet observed by ISEE-1, in Magnetotail Physics, edited by A. T. Y. Lui, p. 119, Johns Hopkins Univ. Press, Baltimore, Md. Cattell, C. A., et al. (1995), ISEE 1 and Geotail observations of low-frequency waves at the magnetopause, J. Geophys. Res., 100, 11,823. Collier, M. R., T. E. Moore, M.-C. Fok, B. Pilkerton, S. Boardsen, and H. Khan (2005), Low-energy neutral atom signatures of magnetopause motion in response to southward Bz, J. Geophys. Res., 110, A02102, doi: /2004ja Daly, P. W., M. A. Saunders, R. P. Rijnbeek, N. Sckopke, and C. T. Russell (1984), The distribution of reconnection geometry in flux transfer events using energetic ion, plasma and magnetic data, J. Geophys. Res., 89, Davidson, R. C., and N. T. Gladd (1975), Anomalous transport properties associated with the lower-hybrid-drift instability, Phys. Fluids, 18, Davidson, R. C., N. T. Gladd, and C. S. Wu (1977), Effects of finite plasma beta on the lower-hybrid-drift-instability, Phys. Fluids, 20, 301. Ding, D. Q., L. C. Lee, and D. W. Swift (1992), Particle simulations of driven collisionless magnetic reconnection at the dayside magnetopause, J. Geophys. Res., 97, Dunlop, M. W., et al. (2005), Coordinated Cluster/Double Star observations of dayside reconnection signatures, Ann. Geophys., 23, Elphic, R. C., and D. J. Southwood (1987), Simultaneous measurements of the magnetopause and flux transfer events at widely separated sites by AMPTE UKS and ISEE 1 and 2, J. Geophys. Res., 92, 13,666. Farrugia, C. J., et al. (1988), A multi-instrument study of flux transfer event structure, J. Geophys. Res., 93, 14,465. Fear, R. C., A. N. Fazakerley, C. J. Owen, and E. A. Lucek (2005), A survey of flux transfer events observed by Cluster during strongly northward IMF, Geophys. Res. Lett., 32, L18105, doi: /2005gl Fu, Z. F., and L. C. Lee (1985), Simulation of multiple X line reconnection at the dayside magnetopause, Geophys. Res. Lett., 12, 291. Gurnett, D. A., and L. A. Frank (1972), ELF noise bands associated with auroral electron precipitation, J. Geophys. Res., 77, Harvey, P., et al. (1995), The Electric Field Instrument on the Polar satellite, Space Sci. Rev., 71, 583. Huba, J. D., N. T. Gladd, and K. Papandopoulos (1977), The lower-hybrid drift instability as a source of anomalous resistivity for magnetic field line reconnection, Geophys. Res. Lett., 4, 125. Huba, J. D., et al. (1978), Lower-hybrid-drift instability in field reserved plasmas, Phys. Plasmas, 23, 552. Kawano, H., and C. T. Russell (1997), Survey of flux transfer events observed with the ISEE 1 spacecraft: dependence on the interplanetary magnetic field, J. Geophys. Res., 102, Ku, H. C., and D. G. Sibeck (1997), Flux transfer events produced by bursty merging at a single X line, J. Geophys. Res., 103, 14,965. Ku, H. C., and D. G. Sibeck (2000), Flux transfer events produced by the onset of merging at multiple X lines, J. Geophys. Res., 105, Kuo, H., C. T. Russell, and G. Le (1995), Statistical studies of flux transfer events, J. Geophys. Res., 100, Labelle, J. A., and R. Treumann (1988), Plasma waves at the dayside magnetopause, Space Sci. Rev., 47, 175. Le, G., J. T. Gosling, C. T. Russell, R. C. Elphic, M. F. Thomsen, and J. A. Newbury (1999), The magnetic and plasma structure of flux transfer event, J. Geophys. Res., 104, 233. Le, G., X. Blanco-Cano, C. T. Russell, X.-W. Zhou, F. Mozer, K. J. Trattner, S. A. Fuselier, and B. J. Anderson (2001), Electromagnetic ion cyclotron waves in the high altitude cusp region: Polar observations, J. Geophys. Res., 106, 19,067. Lee, L. C., and Z. F. Fu (1985), A theory of magnetic flux transfer at the Earth s magnetopause, Geophys. Res. Lett., 12, 105. Lockwood, M., et al. (2001), Coordinated Cluster and ground-based instrument observations of transient changes in the magnetopause boundary layer during an interval of predominantly northward IMF: Relation to reconnection pulses and FTE signatures, Ann. Geophys., 19, Nykyri, K., B. Grison, P. J. Cargill, B. Lavraud, E. Lucek, I. Dandouras, A. Balogh, N. Cornilleau-Wehrlin, and H. Reme (2006), Origin of the turbulent spectra in the high-altitude cusp: Cluster spacecraft observations, Ann. Geophys., 24, Owen, C. J., et al. (2001), Cluster PEACE observations of electrons during magnetospheric flux transfer events, Ann. Geophys., 19, Parker, E. N. (1957), Sweet s mechanism for merging magnetic fields in conducting fluids, J. Geophys. Res., 62, 509. Petrinec, S. M., and C. T. Russell (1993), External and internal influences on the size of the dayside terrestrial magnetosphere, Geophys. Res. Lett., 20, 339. Raeder, J. (2003), Global modeling of flux transfer events: generation mechanism and spacecraft signatures, presented at 2003 Joint EGU/ AGU Meeting, Nice, France, Geophys. Res. Abstr., 5, Raeder, J. (2006), Flux transfer events: I. Generation mechanism for nearly southward interplanetary magnetic field, Ann. Geophys., 24, 381, SRef-ID: /ag/ Rème, H., et al. (2001), First multispacecraft ion measurements in and near the Earth s magnetosphere with the identical Cluster Ion Spectrometry (CIS) experiment, Ann. Geophys., 19, Rijnbeek, R. P., S. W. H. Cowley, D. J. Southwood, and C. T. Russell (1984), A survey of dayside flux transfer events, observed by the ISEE-1 and -2 magnetometers, J. Geophys. Res., 89, 786. Russell, C. T., and R. C. Elphic (1978), Initial ISEE magnetometer results: Magnetopause observations, Space Sci. Rev., 22, 681. Russell, C. T., C. R. Chappell, M. D. Montgomery, M. Neugebauer, and F. L. Scarf (1971), Ogo 5 observations of the polar cusp on November 1, 1968, J. Geophys. Res., 76, Russell, C. T., R. C. Snare, J. D. Means, D. Pierce, D. Dearborn, M. Larson, G. Barr, and G. Le (1995), The GGS/Polar magnetic fields investigation, Space Sci. Rev., 71, 563. Saito, H., T. Yoshino, and N. Sato (1987), Narrow-banded ELF emissions over the southern polar region, Planet. Space Sci., 35, 745. Scarf, F. L., R. W. Fredricks, I. M. Green, and C. T. Russell (1972), Plasma waves in the dayside polar cusp, 1. Magnetospheric observations, J. Geophys. Res., 77, Scholer, M. (1988), Magnetic flux transfer at the magnetopause based on single X line bursty reconnection, Geophys. Res. Lett, 15, 291. Silin, I., and J. Buchner (2005), Small-scale reconnection due to lowerhybrid drift instability in current sheets with sheared fields, Phys. Plasmas, 12, doi: / Sonnerup, B. U. O. (1987), On the stress balance in flux transfer events, J. Geophys. Res., 92, Southwood, D. J., C. J. Farrugia, and M. A. Saunders (1988), What are flux transfer events?, Planet. Space Sci., 36, 503. Sweet, P. A. (1958), The neutral point theory of solar flares, in Electromagnetic Phenomena in Cosmical Physics, edited by B. Lehnert, pp , Cambridge Univ. Press, New York. 12 of 13