Magnetic dipolarization is a key element of the substorm phenomenon, and occurs often in near-geosynchronous region as well as in the farther out tail region (e.g., Lee et al. 2011, Kim et al. 2012). It is well known that magnetic dipolarization in the near-tail region is sometimes accompanied by fast plasma flows. For example, Lee et al. (2012) reported that ~50% of the dipolarization in the near-tail region is accompanied by plasma flows of a significant magnitude (> 100 km/s). However, the precise relationship between dipolarization and plasma flows remains unresolved (e.g., Kim et al. 2010, 2012, Lee et al. 2012, McPherron et al. 2011, Ohtani et al. 2002, 2006, 2009). In particular, how deep the earthward flows can penetrate remains a subject of intense research (e.g., Kim et al. 2012, Dubyagin et al. 2011, Ohtani et al. 2006).

Currently, there are two different scenarios proposed to explain the occurrence of dipolarization close to the Earth, often very close to geosynchronous altitude. In the first scenario, the high-speed earthward flows originating from a farther tail region are stopped at a transition boundary between the dipolar field and the more tail-like field (Shiokawa et al. 1997). This results in a magnetic flux pileup at the transition region and is supposed to cause near- Earth dipolarization there. In the second scenario, magnetic dipolarization is attributed to some kind of local instability (e.g., Lui 2004). Various instabilities, such as currentdriven instabilities and ballooning instabilities, have been proposed (e.g., Lui 2004, Park et al. 2010). The question of which scenario is the right one remains a subject of active debate in the community, and we do not attempt to resolve this in the present work. Rather, we aim to understand the basic properties of magnetic dipolarization, associated plasma flows, and bubbles (described below) very near geosynchronous altitude.

Some researchers interpret plasma flows in the plasma sheet in terms of the entropy parameter, pV^5/3, where p is the plasma pressure and V the volume of the unit magnetic flux tube. In this interpretation, plasma flows are regarded as a bubble if they have a lower pV^5/3 than that of background. The plasma bubble is usually intended to mean “flow channels” or “bursty bulk flows” (BBFs), and often but not necessarily exclusively occurs during substorm expansions. The most likely mechanism for producing a bubble is a patch of reconnection within the plasma sheet, thus the near- Earth X-line model of substorm is in line with the formation of an earthward moving bubble. Interpretations based on pV^5/3 have proven very useful in many previous works. For example, the bubble would move earthward through the plasma sheet until its entropy parameter value equals that of the background plasma (e.g., Pontius & Wolf 1990, Chen & Wolf 1999). There are, in fact, previous studies that have examined the question of how deep bubbles can penetrate earthward (Takada et al. 2006, Ohtani et al. 2006). Dubyagin et al. (2011) also suggested that plasma tube entropy pV^5/3 demonstrates better prediction efficiency for earthward penetration than any other parameters. They report that neither initial flow velocity nor magnetic field strength can compete with the entropy parameter in predicting penetration depth. By testing the relationship between “tail bubble” in near-tail region and geosynchronous disturbance, Kim et al. (2012) found a statistical trend that shows geosynchronous disturbance is more likely to occur when associated with an earthward moving tail bubble with a more-depleted pV^5/3.

The main purpose of this paper is to understand the plasma flows and associated bubble properties around the time of the magnetic dipolarization in the near-geosynchronous region, r < ~7 R_E. We examined a total of 16 dipolarization events that occurred in 2007 to 2010, using the magnetic field and plasma/ particle observations by THEMIS spacecraft when they were situated in the near-geosynchronous region. We attempted to compare the near-geosynchronous events with the previously reported events in the near-tail region. We present the identified magnetic dipolarization events in Section 2, the accompanied plasma flows and bubble parameters in Section 3, and conclusions in Section 4.

To identify and examine dipolarization events in the near-geosynchronous region, we used the magnetic field data from Fluxgate Magnetometers, the plasma data from Electrostatic Analyzers and the energetic particle data from Solid State Telescope onboard the THEMIS satellites (Angelopoulos et al. 2008). To identify dipolarization events

in space close to geosynchronous altitude, we selected dipolarization events observed in the equatorial region of r < ~7 R_E. A well-defined dipolarization was identified based on the requirement that average B_Z over 5 min after onset be larger by 5 nT than that averaged over 5 min before the onset. Using the criteria, a total of 16 dipolarization events were identified for the period from 2007 to 2010 (See Table 1 for a list of the 16 events). The B_Z components of the events are presented in Fig. 1a, in which the red curve refers to an average, and their equatorial locations are shown in Fig. 1b. The average curve exhibits an increase of ~8 nT. It is clearly seen that the magnetic B_Z component remains increased for 30 min or longer after the onset of dipolarization.

To examine the plasma bubble properties associated with the identified magnetic dipolarization events above, we used the entropy parameter, pV^5/3. We calculated p using both the plasma data from Electrostatic Analyzer and the energetic particle data from the Solid State Telescope of THEMIS, and V using the formula by Wolf et al. (2006). Specifically, it is given as follows.

where the units for V are R_E/nT, the magnetic field is in nT, distances are in R_E, the equatorial pressure P_E is in units of nPa, in which case μ_o = 400π(nT)²/nPa, and the calibration factors C, D, and F are 0.7368, 0.7634, and -0.3059, respectively. Also, the equatorial magnetic field and pressure, respectively, are given by the following expressions.

and

where

and G = 0.0107.

Fig. 2 shows an example event obtained from the THEMIS A observations on 12 May 2010. The magnetic field data, radial component of the perpendicular flow and pV^5/3 are

shown for 1 hr interval around the time of the dipolarization onset, which is set to Time = 0, in the bottom panels. The magnetic B_Z component and pV^5/3 varied by ~ +7 nT and ~ -0.02 nPa (R_E/nT)^5/3, respectively, after the dipolarization onset. The maximum and minimum flow speeds are identified as ~50 km/s and ~ -52 km/s, respectively, within the 3 min interval after the onset.

Here, we attempt to examine plasma flows and bubble properties based on the entropy parameter, pV^5/3. Fig. 3 presents the perpendicular flows and pV^5/3 for the time interval from -30 min to +30 min around the onset time of the 16 dipolarization events. The red curves refer to an average for the 16 events. Figs. 3a-c show VperpX, VperpY, and VperpR, the X, Y and radial components of the perpendicular flows, respectively. The earthward flows are identified when either VperpX > 0 or VperpR < 0, and the

signs for the identification of the tailward flows are opposite. The duskward /dawnward directed flows are based on the sign of VperpY, i.e., duskward for VperpY > 0 and dawnward for VperpY < 0. We find that flow peaks occurred for some of the events near the onset times. However, the flow peaks for such events are < ~200 km/s in any direction. When averaged over all 16 events, radially earthward and tailward flow peaks are ~13.8 km/s and ~25.9 km/s within 3 min after the onset, respectively. At other times, the flow speed is mostly within 50 km/s for all of the events.

Lee et al. (2012) studied the plasma flows’ characteristics associated with dipolarization in the near-tail of r ~ 7 - 12 R_E. They report that the peak earthward flow is rarely above 350 km/s, and the peak tailward flow is mostly less than 250 km/s in the near-tail region. In addition, they found that the relative occurrence rate of fast plasma flow overall decreases earthward. In line with this previous report by Lee et al. (2012), we confirm in Fig. 3 that plasma flow speed drops significantly in the near-geosynchronous region.

In Fig. 3d, one can see that the entropy parameter, pV^5/3, dropped after the onset of dipolarization. Specifically, based on the average curve, the decrease is ~0.01 nPa (R_E/nT)^5/3 for ±5 min around the onset. The decrease in the entropy parameter is due to the geometrically rounder shape of the post-dipolarization magnetic field, while the plasma pressure usually increases after the dipolarization.

Kim et al. (2012) studied how deep a “tail bubble” can penetrate earthward by checking geosynchronous disturbance in response to tail bubbles identified at r ~ 7 - 12 R_E. They found a statistical trend that geosynchronous disturbance is more likely to occur when associated with an earthward moving tail bubble with a more-depleted pV^5/3, and that the probability of a bubble penetration effect on geosynchronous disturbance is indeed higher for tail bubbles with a lower pV^5/3. Let us compare our neargeosynchronous results with those obtained for the tail bubbles at r ~ 7 - 12 R_E in Kim et al. (2012).

The comparison is demonstrated in Fig. 4, which shows the average magnetic field B_Z components and pV^5/3 profiles for the events in this study (solid lines) and for those in Kim et al. (2012) (dotted lines). Fig. 4a shows that the magnetic field B_Z is stronger for our events than those of Kim et al. (2012), since our dipolarization events were identified near geosynchronous orbit, while their events were identified father out at r ~ 7 - 12 R_E.

Fig. 4b shows that the average pV^5/3 is always lower for our near-geosynchronous events than for the farther tail events of Kim et al. (2012) throughout the entire time interval. This is an indication that the overall background pV^5/3 decreases earthward, which is consistent with the already well-known fact (e.g., Kim et al. 2012). More interestingly, the pV^5/3 value indicates a smaller drop after the onset by a factor of ~5 for our near-geosynchronous events than for the neartail events of Kim et al. (2012). Therefore, we suggest that magnetic dipolarization in the near-geosynchronous region generates (or is associated with) only very weak bubbles. This means that such bubbles in the near-geosynchronous region would propagate earthward by only a small distance before they stop at a position of equilibrium or drift around the Earth.

In this paper we have reported some properties of the plasma flows and pV^5/3 associated with the 16 magnetic dipolarization events identified in the near-geosynchronous region by the THEMIS satellites. We have found the following important features. (i) The average flow speed at near-geosynchronous altitude is mostly very low, lying within a few tens of km/s on average, although the peak flows can rise up to ~200 km/s within a few minutes

around the dipolarization onset of some limited events. This is in line with the prediction by Lee et al. (2012) that the occurrence rate of fast flow events largely decreases earthward. (ii) The associated bubble parameter, pV^5/3, decreases on average by ~0.01 nPa (R_E/nT)^5/3 after the onset. This is a much weaker decrease by a factor of ~5 as compared to that for the bubbles identified at farther out tail region examined by Kim et al. (2012). Thus this implies that the near-geosynchronous dipolarization is associated with only a weak bubble structure.

Our findings do not answer the question of whether the near-geosynchronous dipolarization is triggered by an internal instability or by the arrival of earthward penetrating tail bubbles. However, our findings do imply that, in any case, the near-geosynchronous substorm dipolarization is subject to only a weak bubble with a weak flow speed. They would propagate only by a small distance before stopping at an equilibrium position slightly inside geosynchronous altitude, or drifting around the Earth. We emphasize that this result is not because the dipolarization events used here are of weak intensity. In fact, the dipolarization events used here are of major intensity at near-geosynchronous orbit, but they create or are associated only with a weak plasma bubble. This is a new feature that has not been noticed before, and demands further investigation.