Ion and electron dynamics in the ionelectron decoupling region of

JOURNAL OF GEOPHYSICAL RESEARCH: SPACE PHYSICS, VOL. 118, 7703–7713, doi:10.1002/2013JA019135, 2013
Ion and electron dynamics in the ion-electron decoupling region
of magnetic reconnection with Geotail observations
T. Nagai,1 S. Zenitani,2 I. Shinohara,3 R. Nakamura,4 M. Fujimoto,3 Y. Saito,3 and T. Mukai 3
Received 17 June 2013; revised 22 October 2013; accepted 24 November 2013; published 23 December 2013.
[1] The ion-electron decoupling region where electron outflow speed differs from ion
outflow speed is formed in the magnetic reconnection site. Ion and electron dynamics in the
ion-electron decoupling region is derived with magnetic field and plasma observations by
the spacecraft Geotail in near-Earth magnetotail magnetic reconnection. The ion-electron
decoupling region has a spatial extent of approximately 11 λi (ion inertial length) along the
GSM x direction, and the dawn-dusk current sheet with main current carriers of electrons
exists over this region. An intense electron current layer with a spatial extent of 0.5–1 λi
occupies in its center around the X line. High-speed electron outflow jets are formed just
outside the central intense electron current layer. They are decelerated and become non-jet
outflows with speed slightly higher than ion outflow speed. Electrons have flattop
distribution functions indicating heating and acceleration in both the outflow jets and the
non-jet outflows; however, heating and acceleration are weak in the central intense current
layer. Inflowing ions enter the central intense electron current layer, and these ions are
accelerated up to 10 keV inside the electron outflow jet regions. Ion acceleration beyond
10 keV and thermalization operate mostly in the non-jet electron outflow regions. Electrons
show thermal distributions without any heating/acceleration signatures immediately beyond
the edge of the ion-electron decoupling region, while higher-energy ions pervade even
beyond the edge and hot MHD plasma flows are produced.
Citation: Nagai, T., S. Zenitani, I. Shinohara, R. Nakamura, M. Fujimoto, Y. Saito, and T. Mukai (2013), Ion and
electron dynamics in the ion-electron decoupling region of magnetic reconnection with Geotail observations, J. Geophys.
Res. Space Physics, 118, 7703–7713, doi:10.1002/2013JA019135.
1.
Introduction
[2] Magnetic reconnection in the near-Earth magnetotail
current sheet has a great advantage for exploring physical
processes in magnetic reconnection. Magnetic reconnection
usually takes place in the symmetric current sheet without
any strong guide fields. The Alfvén velocity, which corresponds to outflow velocity, is high (> 2000 km s 1). Ion inertial length, which probably determines spatial-scale size of
magnetic reconnection, can become approximately 1000 km.
A possible motion of the magnetic reconnection site itself is
thought to be small relative to the outflow speed. Hence, even
current spacecraft observations can resolve various important
characteristics of magnetic reconnection. Observations of
magnetic reconnection in the near-Earth magnetotail by the
1
Department of Earth and Planetary Sciences, Tokyo Institute of
Technology, Tokyo, Japan.
2
National Astronomical Observatory of Japan, Mitaka, Japan.
3
Institute of Space and Astronautical Science, Japan Aerospace
Exploration Agency, Sagamihara, Japan.
4
Space Research Institute, Austrian Academy of Sciences, Graz, Austria.
Corresponding author: T. Nagai, Department of Earth and Planetary
Sciences, Tokyo Institute of Technology, Tokyo 152-8551, Japan.
([email protected])
©2013. American Geophysical Union. All Rights Reserved.
2169-9380/13/10.1002/2013JA019135
spacecraft Geotail and Cluster are summarized by Nagai
et al. [2011], and concise reviews on recent in situ observations are given by Fuselier and Lewis [2011], Fujimoto
et al. [2011], and Paschmann et al. [2013].
[3] On the basis of in situ observations by Geotail for the
15 May 2003 event, Nagai et al. [2011] successfully identify
bidirectional high-speed electron outflow jets from the X line
and an intense cross-tail electron current layer in the center of
the ion-electron decoupling region. Zenitani et al. [2012]
confirm that strong dissipation is confined to the central
intense electron current layer. Hence, the central intense electron layer corresponds to the diffusion region of magnetic
reconnection. Furthermore, it is clearly shown that currentcarrying electron (dawnward) flow speed in the central
intense current layer exceeds electron earthward and tailward
outflow speeds. These are chief ingredients of magnetic
reconnection found even in early full-particle simulations
[e.g., Hesse et al., 1999]. Here the ion-electron decoupling
region is the region where electron bulk flow speed is different from ion bulk flow speed. This region may correspond to
the so-called “Hall physics” regime; however, we cannot
evaluate the Hall term in the generalized Ohm’s law unambiguously in the Geotail data. The “ion diffusion region”
might be used for indicating this region. We do not use this
term since any roles of ions in the diffusion process are not
clarified. The ion-electron decoupling region is simply based
on the observational results.
7703
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
[4] Geotail has been surveying the near-Earth plasma sheet
since 1996, and several statistical studies on magnetic
reconnection have been carried out [e.g., Nagai et al., 1998,
2001, 2005, 2013]. However, encounter of magnetic
reconnection near the equatorial plane of the magnetotail is
extremely rare. The best event relevant to exploring various
aspects of magnetic reconnection is the 15 May 2003 event
used by Nagai et al. [2011]. For this event, 3-D plasma distribution function and 2-D electric field data are available in
addition to the magnetic field data; however, duration of
the event is relatively short and energies of accelerated electrons are not high in this event in comparison with other
events [e.g., Nagai et al., 1998, 2001]. In situ observations
of magnetic reconnection were made on 5 May 2007 near
the equatorial plane. Although ion and electron velocity
distribution function data were available only in twodimensional form and there was no electric field measurement for this event, it is confirmed that this event shows all
important characteristics of magnetic reconnection found in
the 15 May 2003 event. Since electron acceleration appears
to be more evident in this event, dynamics of energetic ions
and electrons can be examined in this event.
[5] In this paper, we investigate the 5 May 2007 event with
Geotail magnetic field and plasma observations. We present
evolution of ion and electron velocity distribution functions
inside the ion-electron decoupling region in order to explore
underlying physical processes. In order to confirm findings
from the 5 May 2007 event, we examine the 15 May 2003
event and carry out statistical analyses. It is clarified that
heating and acceleration are essential characteristics of
electrons in the ion-electron decoupling region although they
are weak in the central intense current layer. It is also
revealed that ions can be accelerated even in electron outflow
jet regions and that further acceleration and thermalization
processes operate for ions outside the electron outflow
jet regions.
2.
Data Sets
[6] We use plasma and magnetic field data from the spacecraft Geotail. The magnetic field data are obtained with the
magnetic field experiment [Kokubun et al., 1994]. Full time
resolution of magnetic field data is 1/16 s (16 vectors s 1).
The magnetic field data are presented in the geocentric solar
magnetospheric (GSM) coordinates. The ion (assumed as
proton) and electron data are obtained with the low-energy
plasma (LEP) experiment [Mukai et al., 1994]. One velocity
distribution function is taken for each 12 s sampling of ions
and electrons. Three components of ion velocity (Vi x, Vi y,
and Vi z), ion density, and ion temperature are calculated on
the spacecraft and transmitted to the ground. Hence, the 3-D
ion velocity moment values are available for all periods.
However, electron velocity moment data are calculated from
velocity distribution function data transmitted to the ground
with correction of photoelectrons. The velocity distribution
function data are taken in the geocentric solar ecliptic (GSE)
coordinate system. Since the data for the 5 May 2007 event
are in 2-D form, we only have electron velocity data Ve x
and Ve y in GSE for this event. We have compared all simultaneous 2-D and 3-D electron velocity data when 3-D electron
data are available, and we confirm that these 2-D and 3-D velocities agree well. We have also confirmed that electron
velocity agrees well with ion velocity for fast MHD flows in
the central plasma sheet (ions show Maxwellian distribution
functions). It is concluded that electron velocity moments are
correctly calculated in our procedure, excluding those in
low-density regions (e.g., tail lobes). In this paper, we mostly
examine the data near the equatorial plane with the condition
of 10 < Bx < +10 nT, so that the limitation of the 2-D velocity data seems to be minimized [see also Nagai et al.,
2011, 2013].
3.
Observations
3.1. The 5 May 2007 Event
[7] The Geotail was located at ( 21.3, 6.9, 1.3 RE) at 0700
UT on 5 May 2007. A small substorm with a minimum AL of
125 nT started near the ground station Fort Churchill
(geomagnetic latitude was 67.85° and magnetic local time
was 00.1 h) near 0703 UT. Figure 1 presents magnetic field
and plasma data for the period from 0658 to 0708 UT.
[8] The magnetic field data are presented at 1/16 s time resolution. During the period 0701–0704 UT, By is almost zero
when Bx becomes zero. When Bx makes several zero
crossing near 0620 and 0653 UT (prior to the data period,
not shown here), By becomes almost zero. These indicate
that there is no strong guide field during this event. Highfrequency variations appear for the period from 0701 to
0705.5 UT in all magnetic field components. The magnetic
field shows field reversal from southward to northward at
0702:40.4 UT. Bz is almost southward prior to this timing
and almost northward after this timing. In the 0704–0706
UT northward Bz period, Bx is negative and By is negative.
In the premidnight region of the magnetotail, By should be
slightly positive (negative) in the Southern (Northern)
Hemisphere where Bx is negative (positive) in the GSM
coordinate system. This negative By deflection is consistent
with the quadruple structure of the Hall magnetic field
[Sonnerup, 1979]. Indeed, Hall electrons flowing into the
magnetic reconnection site [e.g., Nagai et al., 2001, 2003]
are seen in electron velocity distribution functions (not
shown here).
[9] Plasma moment data are plotted with time resolution of
12 s. Here timing used in this paper is start time of a 12 s
plasma sampling. Velocity moment values are plotted in
GSE. Since the rotation angle from GSE to GSM is 11.5°
and the maximum difference between Vi y in GSM and Vi
y in GSE is 18 km s 1, choice of the coordinate system does
not affect any results in this discussion. Flow reversal from
tailward to earthward takes place at 0702:41 UT, in association with the northward turning of Bz. Tailward ion flows
have an almost constant speed of 480 km s 1 for the southward Bz period, while earthward ion flows have an almost
constant speed of +500 km s 1 for the northward Bz period.
Electron bidirectional outflow jets are seen near the ion flow
reversal, and electrons show a high-speed dawnward flows at
the timing of the flow reversal at 0702:41 UT. Electron outflows with speed higher than ion flow speed are observed
for the period from 0701 to 0705.5 UT when the magnetic
field shows the high-frequency variations.
[10] There are two positive spikes in Ve y at 0702:17 and
0703:17 UT. The electron flow is almost field aligned at
0702:17 UT, so that Ve ⊥ y becomes zero (see Figure 3).
For the electron flow at 0703:17 UT, it is possible that the
7704
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
10
Geotail
05 May 2007
(a) Bx (nT)
0
-10
10
(b) By (nT)
0
10
(c) Bz (nT)
-10
(d) Bt (nT)
20
0
-10
10
4000
0
(e) Vx (km/s)
Ve
2000
Vi
0
-2000
-4000
4000
(f) Vy (km/s)
2000
Vi
0
-2000
-4000
Ve
0.24
-6000
-8000
(g) Density (/cm3 )
0.12
0.00
(h) Electron PSD ratio
1.00
5 keV / 1 keV
ePSD
Ion PSD > 20 keV
1.0E-15 (m-6 s 3)
UT 0658
0659
0700
0701
0702
0703
iPSD
0704
0705
0.10
0706
0707
0.01
0708
Figure 1. Magnetic field and plasma data from Geotail for the period from 0658 to 0708 UT on 5 May
2007. (a–d) The magnetic fields Bx, By, Bz, and Bt with time resolution of 1/16 s. (e–f) Electron flow
velocities Ve x and Ve y (thick lines) and ion flow velocities Vi x and Vi y (dashed lines). (g) Ion number
density. (h) Ratio of electron phase space density (PSD) at 5 keV to that at 1 keV (thick line) and > 20 keV
ion phase space density (PSD) (dashed line). Time resolution of plasma data is 12 s. The time interval for
the ion-electron decoupling region is indicated by vertical dashed lines.
flow direction may be disturbed by a strong positive spike in
Bx during this 12 s plasma sampling. Plasma (ion) density
becomes low during the high-speed electron flow period of
0701–705:05 UT. These variations in the magnetic field
and plasma moment data for this event have close similarities
to those in the 15 May 2003 magnetic reconnection event analyzed in detail by Nagai et al. [2011]. For the 5 May 2007,
the ion-electron decoupling region, where electron outflow
speed exceeds ion outflow speed, can be defined as the time
interval from 0701:17 to 0705:29 UT with examinations of
velocity distribution functions described later. The electron
current layer (Ve y < 0) exists during this time interval.
Near the outer edge of the ion-electron decoupling region, a
negative bump appears in Bz near 0701:20 UT while a positive change appears in Bz near 0705:30 UT.
[11] We investigate dynamics of ions and electrons in this
magnetic reconnection site. Figure 2 presents ion and electron energy-time diagrams for the period from 0658 to
0708 UT on 5 May 2007. Figures 2a–2d show ion counts/
sample, while Figures 2e–2h show electron counts/sample.
Ions with energies of ~10 keV show tailward flow
(Figure 2a) for the first half period ending at 0702:41 UT,
and then they show earthward flows (Figure 2c) for the second half period ending at 0707 UT. Duskward flowing ions
(Figure 2b) are seen mostly in the energy range above 10
keV for the period of 0701–0705 UT, indicating the normal
flow reversal from tailward, through duskward, to earthward
[Nagai et al., 2013]. Dawnward flowing ions (Figure 2d)
with maximum counts near 6 keV are ions flowing into the
magnetic reconnection site [Nagai et al., 2011, 2013].
[12] A striking feature in the electron diagrams
(Figures 2e–2h) is enhancement in counts in the energy range
near 5 keV in the period from 0701:17 to 0705:29 UT in
which electron outflow speed is higher than ion outflow
speed. It is also noted that reduction in < 1 keV electron
counts occurs in all directions for this period. When electrons
has thermal distribution, counts in the energy range less than
1 keV are dominant, as seen in the 0658–0701 UT period and
the 0707–0708 UT period. Furthermore, electron flow direction can be recognized in these diagrams. Excess in tailward
7705
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
Geotail/LEP
(a) tailward keV
05 May 2007
10.
ion
1.0
0.1
(b) duskward
ion
10.
1.0
0.1
(c) earthward
ion
10.
1.0
0.1
(d) dawnward 10.
counts/sample
ion
102.4
1.0
0.1
(e) tailward
electron
10.
1.0
0.1
10
0.1
(f) duskward
electron
10.
1.0
0.1
(g) earthward 10.
electron
1.0
0.1
(h) dawnward 10.
electron
1.0
0.1
UT
0658
0659
0700
0701
0702
0703
0704
0705
0706
0707
0708
Figure 2. Ion and electron energy-time spectrograms for the period from 0658 to 0708 UT on 5 May
2007. Tailward, duskward, earthward, and dawnward ions and electrons (counts/sample) are color coded
according to the logarithmic color bar at the right-hand side.
flowing electrons (Figure 2e) is seen until 0702:29 UT, while
excess in earthward flowing electrons (Figure 2g) is seen
from 0703:29 UT. During the period from 0702:29 to
0703:17 UT corresponding to the flow reversal, dawnward
flowing electrons (Figure 2h) show marked excess, relative
to duskward flowing electrons (Figure 2f). This is the most
evident for the data at 0702:41 UT, where color presenting
counts/sample becomes red in Figure 2h while color is green
in Figure 2f. These dawnward flowing electrons produce a
largely negative Ve y (Figure 1), indicating the central
intense electron current layer. Even in other intervals, counts
of dawnward flowing electrons are generally higher than
those of duskward flowing electrons.
[13] In order to further study ion and electron dynamics
inside the ion-electron decoupling region, we examine ion
and electron velocity distribution functions. Characteristics
in the distribution functions can be divided into four classes,
as presented in Figure 3. Here distribution function values,
hereafter called phase space densities (PSDs), are summed
in all directions, and they are presented by a function of
energy. Figure 3 also shows ion and electron flow velocities
perpendicular to the local magnetic field (Vi ⊥ x, Vi ⊥ y, Ve
⊥ x, and Ve ⊥ y) for the period from 0700 to 0707 UT.
Electrons are probably magnetized outside the central small
diffusion region, so that Ve ⊥ is a good measure for the electron
outflow jets. Plasmas are MHD, and both ions and electrons
show thermal distributions outside the ion-electron decoupling
region. The ion and electron data before 0701:17 UT and after
0705:29 UT are presented in the leftmost and rightmost boxes
of Figure 3 as class M (from MHD). Eleven PSD profiles are
stacked in the leftmost box, while 13 PSD profiles are stacked
in the rightmost box. In the central electron current layer at
0702:41 UT (one PSD profile), both ion and electron PSDs
show monotone decrease without any “shoulder” as energy
increases, as presented in the central box of Figure 3 as class
C (from current layer). Both ion and electron PSDs in class C
have excess in the high-energy part and decrease in the lowenergy part, in comparison to those in class M.
[14] In the course of the ion-electron decoupling region,
ion PSDs for < 5 keV are significantly depressed (see
Figure 2). In class C (the central intense current layer), ion
PSDs for < 5 keV show a local maximum while ion PSDs for
10 keV show a local minimum relative to nearby PSDs just
before and after 0702:41 UT. This indicates that inflowing
ions exist in the central intense current layer and that ion
acceleration processes are not effectively operating here.
Ion distribution functions between class C and class M can
be divided into two classes: class A and class F. A clear
7706
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
Geotail
UT 0700
(a)
4500
05 May 2007
0701
0702
0703
0704
V x (km/s)
0706
0707
Ve
0
0
0705
Vi
Vi
V y (km/s)
Ve
-4500
M
F
A
C’ C
C’ A
F
M
-9000
Ion
(b)
M
10-13
F
A
C’
C
C’
A
F
M
10-14
10-15
10-16
(m-6s3 )
0.3 1 3 10
Electron
(c)
10-17
M
0.3 1 3 10
F
0.3 1 3 10
A
0.3 1 3 10
0.3 1 3 10
C’
C
0.3 1 3 10
C’
0.3 1 3 10
A
0.3 1 3 10
F
0.3 1 3 10
keV
M
10-18
10-19
10-20
(m-6s3 )
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
keV
Figure 3. (a) Ion and electron flow velocities perpendicular to the local magnetic field (Vi ⊥ x, Vi ⊥ y, Ve
⊥ x, and Ve ⊥ y) for the period from 0700 to 0707 UT on 5 May 2007. (b) Ion distribution functions are
presented as a function of energy. (c) Electron distribution functions are presented as a function of energy.
Phase space densities (PSDs) are summed for all directions. One count level from each detector is indicated
by dotted line. Distribution functions are classified as classes M, F, A, C′, and C as described in the text.
characteristic for class A (from acceleration) is a peak of ion
PSDs around 10 keV for 0701:41–0702:17 UT (four PSD
profiles) and 0703:29–0703:41 UT (two PSD profiles). This
ion component forms tailward and earthward outflows (see
Figures 2a and 2c). Ion PSDs for > 10 keV increase in
comparison to those in class C, while PSDs for 1–5 keV
significantly decrease in comparison to those in class C.
It is likely that ion acceleration processes up to energy of
10 keV operate efficiently here. Just before class C
(at 0702:29 UT) and after class C (for the 0702:29–0703:17
UT time interval), the ion distribution functions show transitions from class C to class A, so that they are classified as
class C′. In class F (from flattop electron distributions, two
PSD profiles in the left box and eight PSD profiles in the right
box), between class A and class M, ion PSDs for < 10 keV
increase and ion PSDs for 10–20 keV are unchanged, so that
the distributions approach to thermal distributions. These features can be seen as ions with various energies in all directions
in the energy-time diagrams of Figure 2. It is noted that ion
PSDs for energies > 20 keV become high in class F and class
M (see Figure 1h), indicating that further acceleration processes operate for ions near the edge of the ion-electron
decoupling region.
[15] Electrons show a flattop distribution with a shoulder
near 5 keV in class F and class A. As indicated, electron
PSDs for < 1 keV decrease while electron PSDs for > 5 keV
increase. Since electron counts for < 0.3 keV become almost
zero in class F and class A as seen in Figure 2, all electrons
are accelerated in comparison to electrons in thermal distributions (class M). No significant difference between class
F and class A is found in the “shape” of electron distribution
functions. Ve ⊥ appears to show a difference between class
F and class A. Electron speed for class A is generally higher
than that for class F (see also Figure 1), suggesting that the
leading edge of the electron outflow jet defines the boundary
between class A and class F. The electron distributions in class
C′ have transition features between class A and class C.
Figure 1h presents ratio of electron PSD at 5 keV and that at
1 keV. This ratio mostly exceeds 0.1 in the ion-electron
decoupling region where the electron distribution functions
show a flattop shape. This ratio shows a local minimum for
class C (and class C′). Hence, this ratio can be a good indicator
for the electron flattop distribution.
3.2. The 15 May 2003 Event
[16] In order to confirm the dynamics of ions and electrons
in the ion-electron decoupling region from the 5 May 2007
event, the data from the 15 May 2003 are examined.
Geotail was located at ( 27.8, +3.4, +3.5 RE) at 1050 UT.
Figure 4 presents magnetic field and plasma data presented
in GSM for the period from 1051 to 1101 UT for an overview. Figure 5 presents ion and electron velocity distribution
functions (the 3-D data are available for this event). The central intense current layer is seen at 1055:44 UT, as indicated
by a largely negative Ve y. Tailward ion flow has an almost
constant velocity of 820 km s 1 before the flow reversal
and earthward ion flow has an almost constant velocity of
730 km s 1 after the flow reversal in the ion-electron
decoupling region. Electron flow speed shows spiky variations. Large negative Ve x during the earthward ion flow
7707
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
15
Geotail
(a) Bx (nT)
15 May 2003
0
-15
15
(b) By (nT)
0
15
(c) Bz (nT)
-15
(d) Bt (nT)
30
0
-15
15
3000
0
(e) Vx (km/s)
1500
Vi
0
-1500
Ve
-3000
-4500
(f) Vy (km/s)
1500
Vi
0
-1500
-3000
Ve
0.20
-4500
-6000
(g) Density (/cm3 )
0.10
0.00
(h) Electron PSD ratio
5 keV / 1 keV
UT 1051
1052
1053
1.00
iPSD
ePSD
Ion PSD > 20 keV
1.0E-15 (m-6 s 3)
0.10
1054
1055
1056
1057
1058
1059
1100
0.01
1101
Figure 4. Magnetic field and plasma data from Geotail for the period from 1051 to 1101 UT on 15 May
2003. (a–d) The magnetic fields Bx, By, Bz, and Bt with time resolution of 1/16 s. (e–f) Electron flow
velocities Ve x and Ve y (thick lines) and ion flow velocities Vi x and Vi y (dashed lines). (g) Ion number
density. (h) Ratio of electron phase space density (PSD) at 5 keV to that at 1 keV (thick line) and > 20
keV ion PSD (dashed line). Time resolution of plasma data is 12 s. The time interval for the ion-electron
decoupling region is indicated by vertical dashed lines.
period for 1055:56–1057:21 UT are caused by Hall electrons
flowing into the magnetic reconnection site [Nagai et al.,
2001, 2003]. Indeed, By is positive and Bx is positive during
this earthward flow and northward Bz period, so that the By
deflection is consistent with the quadruple structure of the
Hall magnetic field [Sonnerup, 1979]. It is noted that there
is a positive spike in Bz near 1057 UT at the boundary of
the ion-electron decoupling region. The magnitude of negative Bz near the boundary at 1054:43 UT is larger than that
inside the electron jet region near 1055:30 UT.
[17] Ion and electron distributions are classified according
to the analyses for the 5 May 2007 event, as described in
Figure 5. Electrons show a flattop distribution for the period
from 1054:43 to 1056:57 UT, although the electron velocity
distribution function at 1055:44 is classified as class C. The
ratio of electron PSD at 5 keV and that at 1 keV is presented
in Figure 4h. The ratio exceeds 0.1 for class F and class A,
and it shows a local minimum for class C. The ion velocity
distribution function at 1055:44 UT can be classified as class
C. Ion PSDs at 1055:19–1055:31 UT and 1055:56–1056:08
UT show a peak near 10 keV so that these are classified as
class A. The corresponding ions form tailward flows or earthward flows (the energy-time diagrams are presented in Nagai
et al. [2011, Figure 2]). Ion PSDs with energies of > 20 keV
show increases for class F and class M (see Figure 4h).
Furthermore, the electron outflow jets are seen for class A,
while the electron outflow speed is relatively slow, but still
higher than ion outflow speed, for class F. Hence, the classification on the basis of the 5 May 2007 event is applicable to
the 15 May 2003 event, and the dynamics of ions and electrons in 15 May 2003 event is essentially the same as that
in the 5 May 2007 event.
3.3. Statistical Studies
[18] Dynamics of ions and electrons in the ion-electron
decoupling region is examined in the 5 May 2007 and 15
May 2003 events. There exists an intense electron current
layer around the flow reversal and simultaneous Bz reversal.
Electron outflow jets start from the central intense electron
current layer. The electron outflows keep high speed in the
7708
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
Geotail
UT 1053
3000
V x (km/s)
(a)
15 May 2003
1054
1055
1057
1058
1059
Ve
Vi
0
0
1056
Vi
V y (km/s)
Ve
-3000
M
F
A C A
F
M
-6000
Ion
(b)
10-13
10-14
10-15
10-16
(m-6s3 )
Electron
(c)
M
0.3 1 3 10
10-17
M
F
0.3 1 3 10
F
A
C
A
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
A
C
A
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
F
0.3 1 3 10
F
M
0.3 1 3 10
keV
M
10-18
10-19
10-20
(m-6s3 )
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
0.3 1 3 10
keV
Figure 5. (a) Ion and electron flow velocities perpendicular to the local magnetic field (Vi ⊥ x, Vi ⊥ y, Ve
⊥ x, and Ve ⊥ y) for the period from 1053 to 1059 UT on 15 May 2003. (b) Ion distribution functions are
presented as a function of energy. (c) Electron distribution functions are presented as a function of energy.
Phase space densities (PSDs) are summed for all directions. One count level from each detector is indicated
by dotted line. Distribution functions are classified as classes M, F, A, and C.
entire current layer. Somewhat unexpected feature is that ion
flows become high-speed flows in the immediate vicinity of
the flow reversal and keep its high speed inside the ion-electron decoupling region, although ion flow speed is much
smaller than electron flow speed. Electrons have a flattop distribution in the ion-electron decoupling region. In order to
confirm these findings, we examine average variations in
magnetic field and plasma flows near the equatorial plane
of the ion-electron decoupling region.
[19] In order to see ion and electron characteristics in the
ion-electron decoupling region, we use 30 magnetic
reconnection events in 1996–2012 used by Nagai et al.
[2013]. These 30 events are selected with the criterion that
a flow reversal (with an intense electron current layer) occurs
under the condition of 10 < Bx < +10 nT. Only eight
events have 3-D distribution functions. Average variations
in the magnetic field and plasma data for the period from
20 to +20 min are presented with standard deviations in
Nagai et al. [2013, Figure 5]. Figure 6 shows 12 s averaged
ion and electron flow data (Vx and Vy) for the period from
10 to +10 min with the zero epoch defined by the flow reversal. Ve y shows a large negative spike at the zero epoch
( 0.5 to +0.5 min) because of the event selection method
(Ve y should exceed 1000 km s 1). As seen in Figure 6a,
electron outflows immediately have high speed just before
and just after the flow reversal, as seen in Ve x. Ion outflows
have almost constant velocity of 710 km s 1 before the
flow reversal and almost constant velocity of +510 km s 1
after the flow reversal. The electron outflows are decelerated,
and their speed becomes the same as that of the ion outflows.
These timings define the outer boundaries of the ion-electron
decoupling region, as indicated by vertical dashed lines in
Figure 6. The leading edge of electron outflow jet appears to
be obscured in averaging processes. Electron PSDs at 1 keV
decrease while electron PSDs at 5 keV increase for this period
(Figure 6c). Hence, the electron PSD ratio (PSD at 5 keV/PSD
at 1 keV) becomes high for this period (Figure 6d), indicating
a flattop shape in distribution functions. As seen in Figure 6c,
ion PSDs at 1 keV are reduced in the ion-electron decoupling
region; however, it shows a broad local maximum near the
zero epoch. Ion PSDs at 28 keV have a local minimum near
the center of this region. Furthermore, > 10 keV ion PSDs
increase toward the edge of the electron high-speed region.
These characteristics support the findings in the 5 May 2007
and 15 May 2003 events.
4.
Discussion
[20] It has been well recognized that ion dynamics is
decoupled from electron dynamics in the magnetic
reconnection site [e.g., Sonnerup, 1979; Birn and Priest,
2007]. It might be envisaged as follows. Electrons can be accelerated strongly inside the small electron diffusion region
including the magnetic field X line structure and electrons
form high-speed outflow jets. Although the electron flows
are decelerated, their speed is relatively high in the outflow
regions. Ions cannot approach the small diffusion region,
and they are not accelerated efficiently. Hence, ion outflow
speed is much smaller than electron outflow speed, and the
ion-electron decoupling region is formed. Ions are gradually
accelerated, and their flow speed catches the decelerated
electron flow speed at the place which is defined as the edge
7709
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
1000
500
0
-500
-1000
Geotail
(a) Vx (km/s)
0
-500
-1000
-1500
-2000
Vi
(b) Vy (km/s)
Vi
Ve
(c) PSD
10 -13
Ve
(m-6 s3 )
Ion 1 keV x 3
Ion 10 keV
10 -14
Ion 28 keV x 3
10 -15
Ele 1 keV x 100
10 -16
Ele 5 keV x 1000
10 -17
-10
0.10
(d) 5 keV/1 keV Ratio
-5
0
5
0.01
10
minutes
Figure 6. Superposed epoch analyses with 30 magnetic
reconnection events for the period from 10 to +10 min
centered at the flow reversal. (a–b) Electron flow velocities
Ve x and Ve y (thick lines) and ion flow velocities Vi x and
Vi y (dashed lines). (c) Ion and electron PSDs for selected
energy ranges. (d) Ratio of electron PSD at 5 keV to that at
1 keV. The time interval for the ion-electron decoupling
region is indicated by vertical dashed lines.
of the ion-electron decoupling region. One of the main findings in the present study is that ions can be efficiently accelerated even just outside the electron diffusion region near the
center of the ion-electron decoupling region.
[21] It is important to determine spatial scales of the observed structure in the events examined here. In the 15 May
2003 event [Nagai et al., 2011], the tailward retreat speed
is estimated to be 100 km s 1 from MHD outflow speed
and inflow speed, and we can estimate the spatial scales for
the magnetic reconnection structure. Unfortunately, it is difficult to determine tailward retreat speed of the magnetic
reconnection site unambiguously for the 5 May 2007 event.
There are no relevant data periods for examining MHD outflows for the 5 May 2007 event. However, inflow speed can
be obtained. Low-energy (< 1 keV) inflowing ions have Vi
x = 475 km s 1 and +362 km s 1 just before and just after
the flow reversal. The tailward retreat speed of 56 km s 1
can make symmetric inflow speed of 418 km s 1. If we adopt
this value and ion inertial length (1310 km for 0.03 cm 3), the
total extent in the GSM x direction of the ion-electron
decoupling region is ~11 λi and the central intense current
layer appears to be confined into ~0.5 λi. Nagai et al. [2011]
reported that these values are 8 λi and 1 λi for the 15 May
2003 event. In their study, the ion-electron decoupling region
was determined as the time interval from 1055:07 to 1056:44
UT with inspection of ion and electron velocity differences.
However, when we use ion and electron distribution functions
for determining the boundaries as done for the 5 May 2007
event, the time interval from 1054:43 to 1056:57 UT is more
appropriate (three data points are added) and the total extent
of the ion-electron decoupling region becomes 11 λi. Hence,
the magnetic reconnection site has the similar scale sizes for
its structure in the 5 May 2007 and 15 May 2003 events, when
the ambiguity for determining the boundaries and the tailward
retreat speed is taken into account.
[22] Figure 7 presents a two-dimensional (GSM x-z) picture of the ion-electron decoupling region derived from the
present analyses. Electron and ion distribution functions are
constructed with averaging data from the 5 May 2007 and
15 May 2003 events. The ion-electron decoupling region
has the full extent of approximately 11 λi in the x direction
(the Earth-tail axis of the magnetotail), and the electron
current layer is formed in the entire region. Although the full
extent (11 λi) of the ion-electron decoupling region can be
derived from only two events, this is consistent with occurrence of magnetic reconnection in the Geotail observations,
as discussed by Nagai et al. [2011]. The magnetic field has
the X line in the center and has piled-up structure at each
edge. There is a marked asymmetry for the extent of the
ion-electron decoupling region (the earthward side is longer
than the tailward side) in the 5 May 2007 and 15 May 2003
events. This is probably caused by temporal evolution of
magnetic reconnection, rather than by the Earth-tail asymmetry in the background magnetic field and plasma structure.
However, there is a possibility that a tailward retreat speed
of the magnetic reconnection site is slowed down. This point
should be tested with future multipoint observations. Near
the X line, an intense electron current layer (ICL in Figure 7)
with the full width of 0.5–1 λi is formed. A simulation study
[Nagai et al., 2011] suggests that the full thickness (in the z
direction) of the central intense current layer is less than 0.2
λi. Nakamura et al. [2006] reported on the basis of Cluster
multispacecraft observations that the full-scale size of the
reconnection current sheet is ~1 λi in the z direction. In this
2-D picture, the structure is drawn with elongation with approximately a factor of 2 in the z direction.
[23] The electron outflow speed becomes significantly high
in the vicinity of the X line. The electron outflow speed is
slowed down, although it is slightly higher that the ion outflow speed. Electron flattop distributions, which indicate
heating and acceleration of electrons, are seen in the entire
ion-electron decoupling region. It is noteworthy that electrons
in the central current layer are less accelerated (class C),
relative to those inside the electron outflows (class A and
class F). In the central intense current layer (class C), lowenergy inflowing ions make dawnward flows. Ions with energies of > 10 keV can exist even in the central intense
current layer, and they form the duskward flowing component. Ions are efficiently accelerated earthward and tailward
just outside the central intense current layer (class A).
High-energy (> 20 keV) ions are present near and beyond
the ion-electron decoupling region (class F and class M).
7710
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
Ion and Electron Dynamics in Magnetic Reconnection
11 λi
current layer
0.5−1 λi
ICL
Earth
Tail
electron jet
electron jet
electron flat-top
10 keV ion peak
> 20 keV ions
MHD
ion-electron decoupling region
MHD
Ion PSDs
10 -13
10 -14
10 -15
10 -16
-6
(m s 3)
1
10
Electron PSDs
10 -18
10 -19
10 -20
10 -21
(m-6s 3)
F
M
1
10
1
1
10
10
1
10
1
10
1
10
F
A
1
C
A
F
M
1
10
C
A
10
1
A
1
10
10
M
1
F
1
10
10
keV
M
1
10
keV
Figure 7. Structure of the ion-electron decoupling region of magnetic reconnection in the near-Earth
magnetotail current layer. The magnetic field has the X line in the central intense current sheet (ICL) and
the flux piled-up occurs at each edge. Electron outflow jets pervade the central parts of this region.
Electrons show a flattop shape in distribution functions here (class F and class A), except near the
central intense electron current layer (class C). Ion PSDs have a peak at 10 keV in class A, while high-energy
(> 20 keV) ions appear near and beyond the edge. Both ions and electrons show thermal distributions for
class M in the MHD regions. The scales are derived from the two events examined here.
Since ions with energies of ~ 10 keV are dominant in the
ion-electron decoupling region (class A and class F), the
ion bulk outflow speed is relatively constant there.
[24] The electron outflow speed (as seen in Ve ⊥, since
Ve ⊥ is more relevant for representing convection flows than
Ve) for class F is slow, relative to that for class A. It is likely
that the boundary between class A and class F corresponds to
the leading edge of the electron outflow jet (the point where
the electron speed is slowed down). The shoulder of the
electron distribution function may be slightly evident in class
F, relative to that in class A, in average shapes, as seen in
Figure 7. Ions for class A mainly show signature of acceleration (up to ~10 keV), while ions in class F show acceleration
and thermalization. It is possible that this difference in ion
distribution characteristics is caused by convection speed of
the background magnetic field flux. Inside the electron
outflow jet regions, the magnetic field lines that are just
reconnected at the X point are transported with high speed.
Ions in these field lines (class A) are mostly those accelerated
highly near the X point. In the region where the electron outflows are decelerated, low-energy ions flow into this region
along the magnetic field lines so that low-energy ions and
highly accelerated ions can coexist, resulting in the class F
distribution. This point should be verified with future high
time resolution measurements. Outside the ion-electron
decoupling region, plasmas become MHD and ion and electron distributions have thermal distributions (class M).
[25] Heating and acceleration of electrons are commonly
observed in the magnetic reconnection site, as reported in
previous studies [e.g., Nagai et al., 1998, 2001, 2003;
Shinohara et al., 1998; Nakamura et al., 2006; Asano
et al., 2008; Teh et al., 2012]. Electrons with energy
of > 5 keV are enhanced, while electrons with energies of
1 keV are highly depressed. Furthermore, electrons with
energies of < 0.1 keV almost disappear. Electron velocity
distributions have the flattop shape. Asano et al. [2008] examine flattop distributions in Cluster electron measurements
with higher time resolution. They show that flattop distributions are seen in the ion-electron decoupling region near the
equatorial plane. The energy for a shoulder of the flattop
shape in the Cluster observations is consistent with that observed in the present studies by Geotail. Hence, the “flattop”
shape is not caused by any smoothing effect in low time resolution electron measurements. In the flattop distributions in
the events examined here, the “flatness” of the low-energy
part in distribution functions is almost equal for any directions. The high-frequency variations in the magnetic field
are well associated with the electron flattop distributions. It
is possible that wave activities make electron distributions
isotropic, as suggested by Shinohara and Hoshino [1999].
It should be noted that there is almost no asymmetry in
the field direction in the flattop distributions except the
flowing-in field-aligned Hall electron component off the
equatorial plane with larger magnetic field magnitudes (not
7711
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
shown here, see Nagai et al. [1998, 2001]). The amount of the
observed field-aligned beamlike electrons is small so that they
cannot supply the total amount of electrons for the flattop distribution. The flattop distributions can be reproduced in the
recent full-particle simulations for magnetic reconnection
[Zenitani et al., 2013]. However, any generation mechanism
for the flattop distributions in the magnetic reconnection site
is not fully understood.
[26] Suprathermal electrons (with energies of > 10 keV)
are not observed in the ion-electron decoupling region with
the plasma instrument LEP on Geotail in the 5 May 2007
and 15 May 2003 events, although electrons with energies of
10 keV are commonly observed in the hot stationary plasma
sheet of the near-Earth tail and the inner magnetosphere with
the same instrument. Geotail has the energetic particles and
ion composition (EPIC) instrument which is capable of
observing electrons with energies of > 38 keV. The EPIC instrument does not observe any energetic electrons in the ionelectron decoupling region in these two events. However,
both the LEP and EPIC instruments observed high fluxes of
electrons with energies of > 10 keV in the MHD region
after the passage of the ion-electron decoupling region.
Suprathermal electrons are seen after 0707 UT on 5 May
2007 (Figure 2). The tailward and earthward asymmetry
in > 10 keV electron PSDs are seen in the averaged
profiles in Figure 7 (tailward class M versus earthward
class M). The same situations are reported in the Cluster
PEACE (low-energy electrons) and RAPID (high-energy
electrons) observations by Asano et al. [2008]. There
are heated and accelerated electrons with energies
of < 30 keV (observed by PEACE), and flux levels for
energies of > 30 keV (observed by RAPID) are low in
the ion-electron decoupling region. Higher fluxes of electrons with energies of > 30 keV are observed later (after
the event) with high-temperature thermal electrons [see
also Nakamura et al., 2006]. Hence, the present study
cannot help to understand mechanisms for creating
suprathermal electrons in the plasma sheet.
[27] Ions are efficiently accelerated in the ion-electron
decoupling region. As seen in the energy-time diagram for
the 5 May 2007 event in Figure 2 (also see Nagai et al.
[2011, Figure 2] for the 15 May 2003 event), > 10 keV ions
are observed throughout the event. Even in the central intense
current layer, which corresponds to the electron diffusion region, dawnward inflowing (< 1 keV) ions and duskward
high-energy (> 10keV) ion coexist. This may not mean that
high-energy ions continuously stay inside the small diffusion
region; however, it is likely that inflowing ions can be accelerated by the reconnection electric field in the vicinity of the
diffusion region. Ions are accelerated more efficiently in the
outer part of the ion-electron decoupling region, and acceleration processes in this region can supply hot MHD plasmas
in the plasma sheet.
5.
Conclusions
[28] This study reveals ion and electron dynamics in the
ion-electron decoupling region of magnetic reconnection
with magnetic field and plasma observations by Geotail in
the near-Earth magnetotail. The two events studies give scale
sizes in the magnetic reconnection site. The ion-electron
decoupling region extends in the full spatial extent of
approximately 11 λi (ion inertial length) in the GSM x direction, and the dawn-to-dusk electron current layer exists over
this region. An intense electron current layer with the full
spatial extent of 0.5–1 λi is present in its center around the
X line. Electron outflow jets start from this current layer
and occupy the central parts of the ion-electron decoupling
region. Electrons show signatures of heating and acceleration
over the ion-electron decoupling region; however, heating
and acceleration are weak in the central intense current layer.
Any evident difference in electron dynamics is not found between the jet-like outflow region and the non-jet-like outflow
region, except flow speed; however, ion dynamics is probably controlled by the electron outflow speed through the
magnetic field flux transport. Inflowing ions exist even in
the central intense current layer, and these ions can be accelerated up to ~10 keV even in the vicinity of the central
intense current layer inside the electron outflow jet regions.
Ion acceleration processes appear to operate more efficiently
near and beyond the edge of the ion-electron decoupling
region where the electron outflow speed is decelerated.
Acceleration processes for ions are likely major processes
for hot MHD plasmas in the plasma sheet. Although present
electron measurements with a single spacecraft cannot
resolve electron dynamics in magnetic reconnection fully,
clear characteristics can be identified even in observed
electron distribution functions. Main ion dynamics is likely
derived in the present observations. This 2-D picture is probably applicable to the full dawn-dusk extent of magnetotail
magnetic reconnection except the duskside edge [Nagai
et al., 2013]. Hence, the findings of this study provide a good
guide for future researches.
[29] Acknowledgments. T.N. thanks V. Angelopoulos for the informative discussion. The work of T.N. was supported by the JSPS Grant-in-Aid for
Scientific Research 25400476. The research of R.N. was supported by
Austrian Science Fund (FWF) I429-N16 and P23862-N16. All Geotail data
are from the Institute of Space and Astronautical Science/Japan Aerospace
Exploration Agency (ISAS/JAXA). The magnetic field and ion moment data
are available from the Data Archives and Transmission System (DARTS) of
ISAS. Ion and electron energy-time plots are open in the DARTS.
[30] Philippa Browning thanks Alessandro Retinò and an anonymous
reviewer for their assistance in evaluating this paper.
References
Asano, Y., et al. (2008), Electron flat-top distributions around the magnetic
reconnection region, J. Geophys. Res., 113, A01207, doi:10.1029/
2007JA012461.
Birn, J., and E. Priest (Eds) (2007), Reconnection of Magnetic Fields,
Magnetohydrodynamics and Collisionless Theory and Observations,
pp. 1–342 , Cambridge Univ. Press, Cambridge, U. K.
Fujimoto, M., I. Shinohara, and M. Kojima (2011), Reconnection and
waves: A review with a perspective, Space Sci. Rev., 160, 123–143.
Fuselier, S. A., and W. S. Lewis (2011), Properties of near-Earth magnetic
reconnection from in-situ observations, Space Sci. Rev., 160, 95–121.
Hesse, M., K. Schindler, J. Birn, and M. Kuznetsova (1999), The diffusion region in collisionless magnetic reconnection, Phys. Plasmas,
6, 1781–1795.
Kokubun, S., T. Yamamoto, M. H. Acuña, K. Hayashi, K. Shiokawa, and
H. Kawano (1994), The GEOTAIL magnetic field experiment,
J. Geomagn. Geoelectr., 46, 7–21.
Mukai, T., S. Machida, Y. Saito, M. Hirahara, T. Terasawa, N. Kaya,
T. Obara, M. Ejiri, and A. Nishida (1994), The low energy particle
(LEP) experiment onboard the GEOTAIL satellite, J. Geomagn.
Geoelectr., 46, 669–692.
Nagai, T., M. Fujimoto, Y. Saito, S. Machida, T. Terasawa, R. Nakamura,
T. Yamamoto, T. Mukai, A. Nishida, and S. Kokubun (1998), Structure
and dynamics of magnetic reconnection for substorm onsets with Geotail
observations, J. Geophys. Res., 103(A03), 4419–4440, doi:10.1029/
97JA02190.
7712
NAGAI ET AL.: PLASMA DYNAMICS IN MAGNETIC RECONNECTION
Nagai, T., I. Shinohara, M. Fujimoto, M. Hoshino, Y. Saito, S. Machida, and
T. Mukai (2001), Geotail observations of the Hall current system:
Evidence of magnetic reconnection in the magnetotail, J. Geophys. Res.,
106(A11), 25,929–25,949, doi:10.1029/2001JA900038.
Nagai, T., I. Shinohara, M. Fujimoto, S. Machida, R. Nakamura, Y. Saito,
and T. Mukai (2003), Structure of the Hall current system in the vicinity
of the magnetic reconnection site, J. Geophys. Res., 108(A10), 1357,
doi:10.1029/2003JA009900.
Nagai, T., M. Fujimoto, R. Nakamura, W. Baumjohann, A. Ieda,
I. Shinohara, S. Machida, Y. Saito, and T. Mukai (2005), Solar wind control of the radial distance of the magnetic reconnection site in the
magnetotail, J. Geophys. Res., 110, A09208, doi:10.1029/2005JA011207.
Nagai, T., I. Shinohara, M. Fujimoto, A. Matsuoka, Y. Saito, and T. Mukai
(2011), Construction of magnetic reconnection in the near-Earth
magnetotail with Geotail, J. Geophys. Res., 116, A04222, doi:10.1029/
2010JA016283.
Nagai, T., I. Shinohara, S. Zenitani, R. Nakamura, T. K. M. Nakamura,
M. Fujimoto, Y. Saito, and T. Mukai (2013), Three-dimensional structure
of magnetic reconnection in the magnetotail from Geotail observations,
J. Geophys. Res. Space Physics, 118, 1667–1678, doi:10.1002/jgra.50247.
Nakamura, R., W. Baumjohann, Y. Asano, A. Runov, A. Balogh, C. J. Owen,
A. N. Fazakerley, M. Fujimoto, B. Klecker, and H. Rème (2006), Dynamics
of thin current sheets associated with magnetotail reconnection, J. Geophys.
Res., 111, A11206, doi:10.1029/2006JA011706.
Paschmann, G., M. Øieroset, and T. Phan (2013), In-situ observations
of reconnection in space, Space Sci. Rev., doi:10.1007/s11214-0129957-2.
Shinohara, I., and M. Hoshino (1999), Electron heating process of the lower
hybrid drift instability, Adv. Space Res., 24, 43–46, doi:10.1016/S02731177(99)00420-2.
Shinohara, I., T. Nagai, M. Fujimoto, T. Terasawa, T. Mukai, K. Tsuruda,
and T. Yamamoto (1998), Low-frequency electromagnetic turbulence observed near the substorm onset site, J. Geophys. Res., 103, 20,365–20,388,
doi:10.1029/98JA01104.
Sonnerup, B. U. Ö. (1979), Magnetic field reconnection, in Solar System
Plasma Physics, Solar System Plasma Processes, vol. 3, edited by L. T.
Lanzerotti, C. F. Kennel, and E. N. Parker, pp. 45–108, North-Holland,
New York.
Teh, W.-L., R. Nakamura, M. Fujimoto, E. A. Kronberg, A. N. Fazakerley,
P. W. Daly, and W. Baumjohann (2012), Electron dynamics in the
reconnection ion diffusion region, J. Geophys. Res., 117, A12225,
doi:10.1029/2012JA017896.
Zenitani, S., I. Shinohara, and T. Nagai (2012), Evidence for the dissipation
region in magnetotail reconnection, Geophys. Res. Lett., 39, L11102,
doi:10.1029/2012GL051938.
Zenitani, S., I. Shinohara, T. Nagai, and T. Wada (2013), Kinetic aspects of
the ion current layer in a reconnection outflow exhaust, Phys. Plasmas, 20,
092120, doi:10.1063/1.4821963.
7713