Recently, Angelopoulos et al [2008] studied timing and the causal relationship in the substorm that occurred at ~04:55 UT on February 26, 2008. They concluded that the substorm was likely initiated by mid-tail reconnection. On the other hand, Lui [2009] provided evidence that near-Earth current disruption, occurring before the conventional tail reconnection signature (i.e., the convection flow), triggered the onset.

There were two substorms on February 26, 2008 with expansion phase onsets at about 04:05 UT and 04:55 UT, respectively. This paper presents further studies of the two substorms. The measurements of bi-directional anisotropy of energetic electrons indicate that midtail reconnection clearly occurred ahead of auroral signatures of both substorms. The pseudobreakups appeared about 1-2 min immediately after the reconnection onsets, while the expansion phase onsets occurred within about one min when earthward flows arrived at the near-Earth tail.

Observations--Overview

Figure 1 shows a 3-D view of THEMIS spacecraft positions at ~04:00 UT with the projection onto the X-Z plane in the GSM coordinates. Figure 2 shows the overview of the events. It is seen that two sets of substorm activities occurred separately with dipolarization/expansions at ~04:05 and ~04:55 UT, respectively. Figure 3(a) shows the overview of THEMIS measurements. Figure 3(b) shows the measurements of P1, P2, P3 and P4 during the 10 minute interval from 03:57:00 UT to 04:07:00 UT.

Figure 1. Positions of THEMIS probes in the GSM coordinates at 04:00 UT. with representative field lines based on the T-96 magnetic field mode. Click here to enlarge the image.

Figure 2. Two hour (03:30-05:30UT) observations of two substorms on February 26, 2008. From top to bottom are: the IMF in front of the dayside bow shock measured by Cluster 3; the simultaneously measured X-component of the solar wind velocity Vx and the solar wind number density N; the AE index obtained based on the magnetic field measurements of 21 THEMIS fluxgate magnetometers (GMAGs); Pi2 pulsation at the high-latitude GILL station (66.18°N, 332.78°E); Pi2 pulsations at the mid-latitude FYTS (55.765°N,35.281°W) station; the polar cap magnetic flux Ψ created by dayside reconnection calculated based on magnetic field measurements from 109 ground-based stations at magnetic latitudes Ψ> 40°; the westward electrojet current Jw obtained based on the magnetic field measurements from twenty-seven stations around the northern hemisphere auroral zone; the three geomagnetic components at the SNAQ station;the magnetic field and flow velocity observed by THEMIS/P1. Click here to enlarge the image.

Figure 3(a). Magnetic field and flow measurements of four THEMIS probes during the time period of (03:30-05:30UT). From top to bottom plotted are plots of the geomagnetic H-component at the RANK station and the magnetic field and plasma velocity measured by P1, P2, P3 and P4, respectively. Click here to enlarge the image.

Figure 3(b). Magnetic field and flow measurements of four THEMIS probes during the 10 minutes time interval of (03:57 UT-04:07 UT). The first and second panel shows the THEMIS AE index and geomagnetic H-component at the RANK station, respectively. The figure format is the same as (a). Click here to enlarge the image.

A localized auroral arc started to intensify at 04:00:21 UT. A second, stronger auroral intensification began at 04:03:30 UT, followed by rapid poleward expansion at 04:04:30 UT. Figure 4 illustrates the auroral arcs at SNKQ and GILL at five selected times: 04:00:09 UT (before the initial intensification), 04:00:33 UT (after the initial intensification), 04:01:42 UT (after the first/local expansion), 04:03:42 UT (after the major intensification), and 04:05:15 UT (after the major expansion). Meanwhile, in the near-Earth tail, earthward flow was first detected by P4 at X = -9.8 RE at 04:04:16 UT, followed by magnetic dipolarization at 04:04:48 UT. P3 at X= -10.6 R_E observed the flow arrival at 04:04:48 UT and dipolarization at 04:04:51 UT. Note that both the near-Earth onsets of dipolarization and earthward flows occurred nearly simultaneously with the major auroral intensification/expansion within about one minute. Furthermore, in the mid-tail, a tailward flow of ~300 km/s with negative Bz started at 03:58:27 UT at P1 location (X= -21.9 R_E). Beginning at 04:05:16 UT, the flow became earthward with positive Bz. During the reversal of Vx and Bz, Bx remained negative, while By changed from positive to negative. These observations are in agreement with the fact that the spacecraft was encountering the Hall-quadrupole magnetic field and crossing the reconnection region from the tailward side to the earthward side.

Figure 4. Auroral brightening of the ~04:05 UT substorm at SNKQ (right) and GILL at five selected times (left to right): 04:00:09 UT, 04:00:33 UT, 04:01:42 UT, 04:03:42 UT and 04:05:15 UT.

When a spacecraft encounters the reconnection region, it sees field-aligned accelerated electrons flowing out near the separatrix [Nagai et al., 2001]. The associated electron bidirectional anisotropy has been regarded as a good indicator for in situ reconnection observations. Figure 5(a) shows the first observation of a weak electron bidirectional anisotropy averaging from 03:58:34 to 03:58:37 UT. Figure 5(b) and 5(c) show more pronounced anisotropies as P1 was in the tailward outflow region not far from the separatrix and when P1 had moved into the earthward outflow region near the separatrix, respectively. It is seen in Figure 5(b) that higher-energy (200-4000 eV) electrons were streaming mostly tailward (PA~0o) with a higher flux, while lower-energy (10-100 eV) electrons were flowing mostly earthward (PA~180o) with a lower flux. On the other hand, the situation in Figure 5(c) is just opposite. These measurements strongly suggest that P1 did observe the tailward retreating reconnection region and that reconnection occurred before 03:58:34 UT, which is about two minute ahead of the first auroral signature (pseudobreakup).

Figure 5. Bidirectional anisotropies of energetic electrons associated with Hall-effect observed by P1. (a) The first P1 observation of electron bidirectional anisotropy averaging from 03:58:34 UT to 03:58:37 UT when P1 was on the tailward of reconnection site. (b) A typical electron bidirectional anisotropy averaging from 04:01:16 UT to 04:01:19 UT when P1 was in the tailward outflow region close to the separatrix. (c) A typical electron bidirectional anisotropy averaging from 04:05:34 UT to 04:05:37 UT when P1 had passed through the separatrix moving into the earthward outflow region.

It is also of interest to note that the polar cap flux Ψ started to rapidly decrease at ~04:07 UT, indicating that tail reconnection had already progressed from closed plasma sheet field lines to the lobe open field lines before that time.

The Second Substorm

The electron bidirectional anisotropy of collisionless reconnection was observed by P1 at 04:50:28 UT; later, P2 also observed this reconnection signature at 04:50:38 UT. The first/localized auroral intensification occurred at 04:51:39 UT at GILL. P3 saw earthward flow at 04:52:27 UT and dipolarization at 04:53:05 UT. The all-sky imager at SNKQ observed the second stronger intensification at 04:53:03 UT and subsequent extensive expansion at 04:53:48 UT. Ψ started to decrease at ~04:55 UT. Figure 6 shows the auroral activities at SNKQ and GILL at five selected times: 04:51:24 UT (before the initial intensification), 04:51:51 UT (after the initial intensification), 04:52:42 UT (after the first/limited expansion), 04:53:15 UT (after the major intensification) and 04:54:54 UT (after the major expansion). Apparently, the timing feature of the second substorm is similar to the first one.

Figure 6. Auroral activities of the ~04:55 UT substorm at SNKQ (right) and GILL (left) at five selected times: 04:51:24 UT, 04:51:51 UT, 04:52:42 UT, 04:53:15 UT and 04:54:54UT.

Click each image to enlarge.