themis observations of an earthward propagatingrunov et al.: dipolarization front x - 5 33 northward...

GEOPHYSICAL RESEARCH LETTERS, VOL. ???, XXXX, DOI:10.1029/,

THEMIS observations of an earthward propagating1

dipolarization front2

A. Runov,1

V. Angelopoulos,1

M. I. Sitnov,2

V. A. Sergeev,3

J. Bonnell,4

J. P. McFadden,4

D. Larson,4

K.-H. Glassmeier,5,6

and U. Auster,5

1Institute of Geophysics and Planetary

D R A F T April 29, 2009, 5:49pm D R A F T

X - 2 RUNOV ET AL.: DIPOLARIZATION FRONT

This paper presents an analysis of THEMIS observations of a dipolariza-3

tion front (sharp large-amplitude increase of the Z-component of the mag-4

netic field). The front was detected by four of the five probes, first by the5

outermost probe (P1) at X=-20.1RE then by P2 at X=-16.7RE, and finally6

by the P3/P4 pair at X=-11.0RE. Thus the dipolarization front, born tail-7

ward of -20RE, propagated earthward at a velocity of 300 km/s. The front8

thickness was found to be comparable to the ion inertial length. Compar-9

ison between observations and simulations allows us to interpret the front10

Physics, University of California, Los

Angeles, CA, USA

2Applied Physics Laboratory, Johns

Hopkins University, Laurel, MD, USA

3St. Petersburg State University, St.

Petersburg, 198504, Russia

4Space Science Laboratory, University of

California, Berkeley, CA, USA

5Institute of Geophysics and

Extraterrestrial Physics, Technical

University Braunschweig, Germany

6Max Planck Institute for Solar System

Reseach, Katlenburg-Lindau, Germany


RUNOV ET AL.: DIPOLARIZATION FRONT X - 3

as the leading edge of a flow burst formed by an impulse of magnetic recon-11

nection in the midtail.12



1. Introduction

Spatially and temporally localized dipolarizations (increases of the magnetic field ele-13

vation angle) are often observed in the near-Earth as well as in the mid-tail plasma sheet14

during bursty bulk flow (BBF) events. Recent superposed epoch analyses based on Geotail15

data reveal similarity in earthward BBF-related variations of the north-south magnetic16

field component (Bz), ion density and temperature observed at -31< X


northward magnetic field and fast propagation of that front structure over a long distance33

(several RE), may indeed be explained by transient reconnection in the magnetotail.34

In this paper we report on observations of a DF by five THEMIS spacecraft (probes),35

situated in the near-equatorial plasma sheet at distances of -20< X


Between 0751 and 0754 UT similar front-like variations of the Bz field were detected54

consecutively by P1, P2, and P3/P4 probes (Figure 1 c) with a more gradual buildup of55

Bz at P5, which was located 1RE southward of P3/P4. Assuming planar front, the timing56

of the DF at P1 (0751:26UT) and P2 (0752:35UT) suggests earthward propagation at a57

velocity of 330 km/s. Timing at P2 and P4 (0754:10UT) results in 360 km/s. Thus, the58

DF propagates from -20 to -11RE without deceleration.59

Figure 2 summarizes observations at P1 and P2 during 2.5 min around the DF (0751:0060

- 0752:30UT and 0752:00 - 0753:30UT, respectively). P1 was in the northern half of the61

plasma sheet close to the neutral sheet, while P2 was mainly in the southern half and in62

the neutral sheet. Between 0751:15 and 0751:30UT, P1/FGM detected bipolar (negative63

to positive) variations of Bx (blue trace) and Bz components (red trace), accompanied by64

a positive variations of By. Since the burst-mode survey was triggered by the variation of65

Bz, the FGM data with the highest time resolution (FGH, 128 vectors/s) are available.66

Using FGH data, the precise duration of the front passage from negative Bz peak (-5 nT)67

to the first local positive peak (20 nT) was 1.35 s, (Fig. 2 c), i.e., the DF thickness was of68

400 km. Variations of the electric field components Ey and Ex, shown in DSL (close to69

GSE) coordinates, registered by P1/EFI are consistent with the PIC simulations [Sitnov70

et al., 2009]. The time-energy spectrograms, combining high- (SST) and low-energy (ESA)71

ion and electron spectra, show rapid changes indicating ion and electron energization.72

Plasma moments reveal a drop of the plasma density and pressure accompanied by a73

gradual growth of the earthward bulk flow speed up to 1000 km/s and an increase of the74



magnetic pressure. These are the distinctive features of BBFs [Angelopoulos et al., 1992]75

supported by the model of plasma bubbles [e.g., Birn et al., 2004, and references therein].76

P2, located 3.4RE earthward of P1, detected the bipolar (small-amplitude negative then77

sharp high-amplitude positive) variations in |Bx| and Bz between 0751:25 and 0752:35UT.78

They were associated with a positive By variation, increase of Ex and Ey, rapid changes79

in particle ET spectra, a drop of the ion density and increase of the earthward plasma80

flow. Pp increased while Pm decreased during the negative Bz variation (this effect is more81

pronounced at P2 than at P1). Pp rapidly decreased while Pm quickly rose on the DF.82

The peak-to-peak (1.5 and 31 nT) front propagation time was estimated using FGH data83

(Fig. 2 d) to be 1.70 s, i.e., the DF was 500 km thick.84

Figure 3 shows DF observations at P3 and P4, situated at the same X and Z and85

separated in 1RE in YGSM . At 0753:30UT, both P3 and P4 were in the southern half of86

the plasma sheet at Bx=-25 and -22 nT, respectively, detecting Bz=3nT. At 0753:50UT,87

P3 started to detect a decrease of |Bx| and |By|, without significant variations in Bz. A88

sharp decrease of Bz down to zero was observed between 0754:06.0 and 0754:06.5UT.89

Both |Bx| and |By| also reached local minima at that short time interval, which results90

|B|=2.3 nT. Bz reached the local maximum (19 nT) at 0754:08.1UT. The DF passing time91

(1.6 s, Fig. 3 c) corresponds to DF thickness of 500 km. P4 started to detect a decrease of92

Bz and an increase of By at 0753:57UT. A local minimum of Bz (-8 nT) was detected at93

0754:10.3UT, and a local Bz maximum (26 nT) was at 0754:12.0UT (1.7 s front passage94

duration, 500 km thickness). The magnetic field variations and the signatures in particle95



spectra and moments detected by P3 and P4 are similar to those, detected earlier by P196

and P2, which suggests the probes encountered the same spatial structure.97

Observations made by P5, located at the same X and Y as P4, but 1RE southward,98

between 0752:30 and 0755:00UT are summarized in Figure 4. Although P5 observed99

more gradual dipolarization (see also Fig 1), the signatures in the spectra, moments and100

magnetic pressure at P5 were similar to those at the other probes. They started earlier101

compared to P3/P4, presumably because P5 was much farther from the neutral plane102

(Bx=-36 nT at 0752:30UT). Lag time between increases of Pm at P1 and P5 (92 s) gives103

a propagation velocity of 400 km/s.104

Minimum Variance Analysis (MVA) [Sonnerup and Scheible, 1998] was applied to the105

magnetic field time series at four probes (P1 - P4) detecting the DF to determine the front106

orientation. The MVA results are summarized in Table 1. The three MVA eigenvectors:107

R1, R2, and R3, corresponding to the three eigenvalues: λ1, λ2, and λ3 define maximum,108

intermediate, and minimum variance directions in the GSM coordinates, respectively. To109

define the direction unambiguously, the X-component of the minimum variance direction110

was set to be positive (earthward) in accord with the propagation direction defined from111

timing. R3 may be interpreted as the front-normal vector. At P1 and P2, the normals112

were close to the X direction, indicating a boundary in the Y Z plane. The bulk velocity113

directions were close to the normals (Figure 4 b). At P3 and P4, the XZ projections of114

normals were consistent, while XY projections were almost orthogonal. The normal at115

P3 indicates a tilt of the front in the XZ plane. Projections of bulk velocity directions at116



P3 and P4 were similar. We did not apply MVA for P5 data in view of the rather gradual117

variations of the magnetic field there.118

3. Discussion and Summary

Multi-point observations by the five THEMIS probes during 0751 - 0755UT on Feb.119

27, 2009 for the first time provide unambiguous evidence of earthward propagation of the120

dipolarization front (DF), which was born tailward of 20RE, up to 11RE at the velocity121

of 300 km/s. The DF was found to be a thin boundary, separating plasmas with different122

properties. The DF thickness was 400 - 500 km, which is comparable with the ion inertial123

length (di ∼300 km for N=0.5 cm−3) and smaller than the 10 keV-proton gyroradius in124

Bz=15nT (1000 km). Thus, observations reveal an example of the space plasma self-125

organization, when a micro-scale structure, with thickness less than the ion gyroradius126

implying different motions of ions and electrons, remains structurally stable during its127

propagation over a macroscopic distance of about 10RE.128

P5, located 1 RE southward of P4, observed signatures of the DF about a minute129

prior to P4. It may be explained by faster propagation of disturbances at the plasma130

sheet periphery where the Alfvén speed is larger [e.g., Krauss-Varban and Karimabadi ,131

2003]. Another cause of the relatively early DF signal detection at P5 is the sharply132

curved shape of the DF flux tube [Nakamura et al., 2005], which is also seen in MHD133

simulations of plasma bubbles [Birn et al., 2004]. Figure 4 b summarizes the observation134

and interpretation scheme. The MVA normal directions suggest that the dipolarization135

region was deflected dawnward, similar to earlier observations [Nakamura et al., 2002], so136

that P4 was eventually found on its dusk-side edge.137



The observations show that the quick southward variation observed ahead of the DF138

is a spatial structure associated with the propagating dipolarization front. The observed139

spatial scales of the DF and the Bz dip (∼ di, similar to those resulting from the PIC140

simulations [Sitnov et al., 2009]) suggest their relation to the Hall-type currents due to141

ion-electron decoupling on the front, resulting in Ex observed in the simulations and in the142

data. The ∼1 s-long Bz dips down to negative values were preceded by a longer decrease of143

magnetic pressure with a simultaneous increase of plasma pressure: a diamagnetic effect144

due to plasma compression ahead of the moving dipolarization front.145

To conclude, a major conjunction of the THEMIS probes along the magnetotail provided146

the unique evidence of DFs as boundaries with a micro-scale thickness propagating over147

a macro-scale distance of 10RE at a velocity comparable to the local Alfven speed. The148

analysis of such thin boundaries requires a kinetic approach. The agreement with PIC149

simulations [Sitnov et al., 2009] suggests transient magnetic reconnection as the most150

plausible source of DFs in the magnetotail.151

Acknowledgments. We acknowledge NASA contracts NAS5-02099 and NNX08AD85G,152

the German Ministry for Economy and Technology and the German Center for Aviation153

and Space (DLR), contract 50 OC 0302. The OMNI data are available on CDAWeb. We154

thank P.L. Pritchett and L. Lyons for discussions, B. Kerr, P. Cruce, and A. Prentice for155

help with software and edition.156

References

Angelopoulos, V. (2008), The THEMIS mission, Space Sci. Rev, 141, 5–34.157



Angelopoulos, V., , et al. (1992), Bursty bulk flows in the inner central plasma sheet, J.158

Geophys. Res., 97, 4027–4039.159

Auster, H. U., et al. (2008), The THEMIS fluxgate magnetometer, Space Sci. Rev., 141,160

235–264.161

Baumjohann, W., M. Hesse, S. Kokubun, T. Mukai, T. Nagai, and A. A. Petrukovich162

(1999), Substorm dipolarization and recovery, J. Geophys. Res., 104, 24,995–25,000.163

Birn, J., J. Raeder, Y. L. Wang, R. A. Wolf, and M. Hesse (2004), On the propagation of164

bubbles in the geomagnetic tail, Ann.Geophys., 22, 1773–1786.165

Bonnell, J. W., , et al. (2008), The electric field instrument (EFI) for THEMIS, Space166

Sci. Rev., 141, 303–341.167

Hesse, M., and J. Birn (1991), On reconnection and its relation to the substorm current168

wedge, J. Geophys. Res., 96.169

Krauss-Varban, D., and H. Karimabadi (2003), Timing and localization of reconnec-170

tion signatures - Is there a substorm model problem?, Geophys. Res. Lett., 30, doi:171

10.1029/2002GL016369.172

Lui, A. T. Y., R. E. Lopez, S. M. Krimigis, R. W. McEntire, and L. J. Zanetti (1988), A173

case study of magnetotail current sheet disruption and diversion, Geophys. Res. Lett.,174

15, 721–724.175

Mann, I. R., et al. (2008), The upgraded CARISMA magnetometer array in the THEMIS176

era, Space Sci. Rev.., 141, 413–451.177

McFadden, J. P., , et al. (2008), The THEMIS ESA plasma instrument and in-flight178

calibration, Space Sci. Rev., 141, 277–302.179



Nakamura, R., et al. (2002), Motion of the dipolarization front during a flow burst event180

observed by Cluster, Geophys. Res. Lett., 29, 1942, doi:10.1029/2002GL015763.181

Nakamura, R., et al. (2005), Multi-point observation of the high-speed flows in the plasma182

sheet, Adv. Space. Res., 36, 1444–1447.183

Ohtani, S., S. Kokubun, and C. T. Russell (1992), Radial expansion of the tail current184

disruption during substorms: A new approach to the substorm onset region, J. Geophys.185

Res., 97, 3129–3136.186

Ohtani, S. I., M. A. Shay, and T. Mukai (2004), Temporal structure of the fast convective187

flow in the plasma sheet: Comparison between observations and two-fluid simulations,188

J. Geophys. Res., 109, A03210, doi:10.1029/2003JA010002.189

Russell, C. T., , et al. (2008), THEMIS ground-based magnetometers, Space Sci. Rev.,190

141, 389–412.191

Semenov, V. S., , et al. (2005), Reconstruction of the reconnection rate from cluster192

reasurements: first results, J. Geophys. Res., 110, A11217, doi:10.1029/2005JA011181.193

Sergeev, V., R. C. Elphic, F. S. Mozer, A. Saint-Marc, and J.-A. Sauvaud (1992), A two-194

satellite study of nightside flux transfer events in the plasma sheet, Planet. Space Sci.,195

40, 1551–1572.196

Shiokawa, K., W. Baumjohann, and G. Haerendel (1997), Breaking of high-speed flows in197

the near-Earth tail, Geophys. Res. Lett., 24, 1179–1182.198

Sitnov, M. I., M. Swisdak, and A. V. Divin (2009), Dipolarization fronts as a signa-199

ture of transient reconnection in the magnetotail, J. Geophys. Res., 114, A04202, doi:200

10.1029/2008JA013980.201



Slavin, J. A., et al. (2003), Geotail observations of magnetic flux ropes in the plasma202

sheet, J. Geophy. Res., 108, 1015, doi:10.1029/2002JA009557.203

Sonnerup, B. U. Ö., and M. Scheible (1998), Minimum and maximum variance analysis,204

in Analysis Methods for Multi-Spacecraft Data, edited by G. Paschmann and P. Daly,205

pp. 185–220, ESA, Noordwijk.206



Figure 1. THEMIS sc positions in XZ (a) and XZ (b) GSM planes and c) time series

of Bz (GSM) at all five probes (P1 - P5).

Figure 2. Summary of THEMIS P1 (a) and P2 (b) observations during 0750:30

- 0755:00UT. For each probe from top to bottom: X,Y ,and Z GSM components of

the magnetic field; X and Y DSL (Y component was 180◦ rotated to point duskward)

components of the electric field (spin resolution); Ion energy-time spectrogram combining

SST and ESA ions (a blank stripe indicates the energy gap between the two instrument

ranges); Electron (SST and ESA) time-energy spectrogram; ion number density X, Y ,

and Z GSM components of the ion bulk velocity calculated with ESA and SST inputs;

Magnetic (Pm) and plasma (Pp) pressures. c) and d): GSM B (128 vectors/s) during 15 s

around the dipolarization front at P1 and P2, respectively.

Figure 3. Summary of THEMIS P3 and P4 observations during 0750:30 - 0755:00UT.

The same format as in Fig. 2.



Figure 4. a) Summary of THEMIS P5 observations between 0752:30 and 0755:00UT.

The same format as in Fig. 2. b) An interpretation scheme. Black arrows are projections

of MVA normals (R1) onto XY (upper panel) and XZ (bottom panel) GSM planes; blue

arrows are projections of bulk velocity directions to the XY GSM plane. Gray segments

represent a flux tube with the enhanced magnetic flux populated by hot and tenuous

plasma. Times of DF crossing by each probe are shown.

sc UT λ1, λ2, λ3, R1 R2 R3

P1 0751:26 76.5, 6.02, 0.13 0.37,-0.16, 0.91 -0.29,-0.95,-0.05 0.88,-0.25,-0.40

P2 0752:35 107.0, 8.93, 0.17 0.35,-0.26, 0.90 -0.50,-0.86,-0.05 0.79,-0.43,-0.43

P3 0754:07 102.3, 4.64, 0.72 0.83,0,-0.08,-0.55 0.12, 0.99, 0.03 0.54,-0.09, 0.84

P4 0754:11 153.2, 1.90,0.04 0.69, 0.24,-0.69 0.69,-0.50, 0.52 0.22, 0.83, 0.51

Table 1. Minimum Variance Analysis results. UT indicates instances of the center

of the positive Bz variations. MVA was performed over a variable window around the

specified UT. Results with the best ratio of λ2 and λ3 are shown.



UT

a)

b)

P1

P2

P3

P4

P5

-4

-2

0

2

4Z

0 -5 -10 -15 -20X

Y

-4

-2

0

2

4

P3(THB)

P4(THC)

P5(THA)

P1(THB)

P2(THC)

c)

Figure 1.



B,

nT

E,

mV

/mE

NE

RG

Y,

eV

i+

e-

N,

1/c

m3

V,

km

/sP,

nP

a

Vx

Vy

Vz

Pp

Pm

B,

nT

E,

mV

/mE

NE

RG

Y,

eVN

, 1

/cm

3V

, k

m/s

P, n

Pa

i+

e-

Vx

Vy

Vz

Pp

Pm

Ex

Ey

DSL

GSM

Ex

EyDSL

GSM

GSM

GSM

TH

EM

IS P

1 (

TH

B)

TH

EM

IS P

2 (

TH

C)

B,

nT

Bx, By, Bz

GSM

Bx, By, Bz

GSM

a)

b)

c) d)

P1

P2

Figure 2.



B,

nT

E,

mV

/mE

NE

RG

Y,

eVN

, 1

/cm

3V

, k

m/s

P, n

Pa

Vx

Vy

Vz

Pp

Pm

i+

e-

B,

nT

E,

mV

/mE

NE

RG

Y,

eVN

, 1

/cm

3

Vx

Vy

Vz

Pp

Pm

V,

km

/sP,

nP

a

Ex

Ey

Ex

EyDSL

DSL

GSM

GSM

GSM

GSM

TH

EM

IS P

3 (

TH

D)

TH

EM

IS P

4 (

TH

E)

B,

nT

Bx, By, BzGSM

Bx, By, BzGSM

P3 P4

a)

b)

c) d)

Figure 3.



B,

nT

E,

mV

/mE

NE

RG

Y,

eVN

, 1

/cm

3V

, k

m/s

P, n

Pa

Vx

Vy

Vz

Pp

Pm

DSL

GSM

GSM

TH

EM

IS P

5 (

TH

A)

Ex

Ey

MVA V_bulk

0751:260752:350752:58

0754:07

0754:10

b)

a)

Figure 4.


themis observations of an earthward propagatingrunov et al.: dipolarization front x - 5 33 northward...

Documents