The Kinetic Physics of
Magnetotail Dynamics:
Reconnection, Ballooning,
and Dipolarization Fronts
P.L. Pritchett
UCLA
Collaborators: F.V. Coroniti
Y. Nishimura
Outline
• Introduction to magnetotail and PIC simulations
• Effect of Bz in stabilizing spontaneous tearing instability
• Observational guidance
• Possible role of ballooning/interchange instability
• Generates fronts with ion kinetic scale thickness and ~ RE east-west extent
• Generates intense off-axis waves ahead of fronts
• Leads to structuring of auroral streamers
• Properties of jet (dipolarization) fronts associated with localized reconnection
• Summary/Questions
Essential complication in the magnetotail (as opposed to magnetopause)
is the presence of a quasi-permanent Bz component of the magnetic
field. Field lines generally are never strictly anti-parallel. Bz ~ 0.1 B0.
Characteristic Parameters
• Proton Cyclotron Frequency in 20 nT field : Ωi0 = 1.9 s-1 (Ωi0-1 = 0.52 s)
• Proton Inertial Length at density of 0.3 cm-3 : di = c/ωpi = 416 km = RE/15
• Proton (5 keV) Gyroradius in 20 nT field: ρi0 = VTi/Ωi0 = 520 km
• Stochasticity Parameter (Büchner & Zelenyi): κj = (Bn/B0) (L/ρj0)1/2
~ 0.5 for protons (L = 2 RE)
~ 5 for electrons (Te = Ti/5)
• Curvature Radius Rc = 1/kc, kc = n ⋅ [(b ⋅ ∇)b] = (1/Bn)(dBx/dz) on axis
Characteristics of (Most) PIC Simulations
PIC simulations follow full dynamics of ions and electrons; must resolve all spatial and
temporal scales.
Cost of simulation: In 2D
∝ (M /m ) ,
i
e
2
For Harris current sheet equilibrium:
In 3D
∝ (M /m )
i
e
5/2
c/VTe = (1 + Ti/Te)(1/2) ωpe/Ωce
Magnetotail: Ti/Te ~ 7, ωpe/Ωce ~ 8 - 10, c/VTe ~ 25
Typical PIC Simulation: Ti/Te = 4, ωpe/Ωce = 2, c/VTe = 4.5
∝ (ω
Cost in 3D ∝ (ω
Cost in 2D
pe/Ωce)
3
pe/Ωce)
4
Spatial Scales: given in units of c/ωpi (normally)
Velocity: given in units of VA (or sometimes VTi), typically ~ 1000 km/s
Reconnection Onset in Presence of Bz
Harris Sheet
• Usual Reconnection Configuration: Harris Current sheet;
not directly applicable to magnetotail due to Bz.
5
z
−5
−5
0
B
0
10
x
Bz/B0 = 0.05
z
5
0
−5
0
10
20
30
40
x
kρen = kL (meTe/MiTi)1/2 (ρi0/L)(B0/Bn) < 1
0.5
0.010
1 - 0.10
Bn/B0 > 0.0005 - 0.005
Bn > 0.01 - 0.1 nT
0
0.2
−10
• Ion Tearing Instability?
• Unless half width comparable to di, growth rate too
small to overcome ion magnetization.
• Electron stabilization effect: Either electron adiabaticity
(Lembège & Pellat, 1982) or simply conservation of Py
in 2D system (Pellat et al., 1991) ensures that tearing
mode EM field produces strong compression of
electron density. Energy associated with this compression exceeds free energy in reversed B configuration.
• Condition for electron stabilization: kρen < 1.
5
0.4
0
z
• Electron Tearing Instability: not viable since cyclotron
motion removes electron Landau resonance.
B
Thus spontaneous tearing
instability unlikely to occur
in magnetotail
2D PIC Simulations
kρen = 3.7
Ti = Te
Ti = 4 Te
0
312
625
0.04
kρen = 1.8
0
395
790
B /B
0
0.03
Bz/B0
0.03
0.04
0.01
0
0.02
z
0.02
0.01
0
10
20
−3
10
30
40
x/di
0
50
0
10
−4
6
7
8
−4
10
10
6
7
8
−5
10
30
40
x/di
50
Mode Intensities
−5
10
−6
−6
0
200
t
i0
400
5
600
10
−0.05
0
0
200
400
600
t
i0
0
5
0
z/di
z/di
20
−3
Mode Intensities
10
10
−0.05
0
−0.1
−0.1
−5
−0.15
0
20
x/di
40
−5
−0.15
0
20
x/di
40
2D x,z simulations: Bz = 0.02 B0, half-width L = ρi0, mi/me = 64, nb = 0
Since Reconnection Clearly Occurs in Magnetotail, What are the Alternatives?
• External Driving: Enhanced convection Ey field arising from southward IMF can force
thinning of current sheet and force local reversal of Bz profile [Pritchett and
Coroniti, 1995, 1997; Hesse and Schindler, 2001].
• Direct day-night coupling [Nishimura et al., 2014]: dayside flow channel and PMAF
followed by polar cap airglow patch that crosses the polar cap followed by
nightside PBIs and plasma sheet flow bursts.
• Turbulence in the Tail: local reversals of Bz might arise randomly from tail variability.
• Sitnov & Schindler (2010): May be able to circumvent electron stabilization for
equilibria with more than 2 characteristic spatial scales (“hump” configuration).
• Non-Reconnection processes: Role of bubbles in the magnetotail [Pontius and Wolf,
1990; Chen and Wolf, 1993]. Once produced, bubbles should propagate earthward
due to buoyancy effect, perhaps explaining formation of fronts in near-Earth plasma
sheet.
Plasma Sheet
• Curved magnetic geometry (finite
B normal) can have significant effects
on particle dynamics for both
electrons and ions: bounce and
drift resonant interactions.
• Must use (at least) 2D equilibrium
to describe tail current sheet.
• Mid-tail minima in Bz are observed
during periods of extended magnetospheric convection (Sergeev et al.,
1994) and expected from MHD
models (Hau et al., 1989).
• Bz minimum also observed by Saito et
al. [2010] (THEMIS) at ~ 10 RE: persists
for 20 minutes.
Sergeev et al., 1994
Plasma Sheet Onset Conditions
Saito et al. [2010]: THEMIS observations on
April 8 & 12, 2009.
• Infer Bz minimum at 10 RE, persists for Figure
20 2. Magnetic field B and B (averaged for 3 min) as a function of location for 8 April 2009 event. The three
umns represent features of (left)-1
1 hr,
3 min prior to dipolarization.
hr(middle) 20 min and (right)
-20 m
-3 mThe latitudinal profile o
min prior to breakup
are categorized into two types: (bottom left) flat profile and (bottom middle and right) increasing function of ∣Z
∣
3.1. Model Assumptions
which is applicable for thin current sheet geometry and
• Equatorial Bz as small as 1 nT
Bz being closely located in r‐! plane.
[ ] In local Cartesian coordinates with x and y all or two spacecraft
z
r
GSM
11
corresponding to r and !, respectively,
t (min) magnetic field
measured at a point (r, !, z) is B(t, r, !, z). Simultaneous
two‐point observation gives:
!Bz ¼
-6
@Bz
@Bz
@Bz
!r þ
!! þ
!z;
@r
@!
@z
ð1Þ
Machida et al. [2009]: Statistical study of temporal
where D means difference between two points. From the
above analyses we may assume:
and spatial development of near-Earth magnetotail
@B
@B
around substorm onset using Geotail data.
j%j
j
ð2Þ
j
@z
@r
z
z
that r and z are distances from the Earth and the mag
equator (Br = 0), respectively. In this model mag
field lines are confined in a plane (local two‐dimensi
or 2D geometry) and a direction of r is defined fro
direction of Br.
3.2. Gradient Analysis for 2D Magnetic Field Struc
[12] For 2D geometry B(r, z) = (Br(r, z), Bz(r, z)) in
local Cartesian coordinates, r · B = 0 is
@Br ðr; zÞ @Bz ðr; zÞ
þ
¼ 0:
@r
@z
and
Observe enhanced Bz around -20RE with a
minimum earthward at ~-18RE; tailward of max Bz
decreases, suggesting a possible hump structure.
@Bz
’ 0;
@!
ð3Þ
[13] In the northern hemisphere, Bz > 0 and Br < 0. T
signs may hold for the magnetotail configuration o
macroscale for most of the time. In order to obtain r
3 of 5
0
4
|z|
x
Simulation Model
(a)
0.4
• 3D Particle-in-Cell
1
2
3
1A
Bz/B0
0.3
0.2
Bz
0.1
• 2D Current sheet equilibrium in x,z
with minimum in equatorial Bz field
0
0
100
200
• mi/me = 64, L/ρi0 = 1.6, ρi0 = 16 Δ,
512 X 1024 X 256 grid
Density
• Boundary conditions:
x: closed, δEy = 0, particles reflected
into opposite half z plane
y: periodic
z: conducting
(b)
0
n/n
1
2
3
1A
−2
10
0
100
200
300
400
500
x
6
B0 ≈12 nT;
(c)
1
2
3
1A
Bx lobe
0
4
x∞
equatorial β ~ 600
500
0
10
B /B
• Lobe field = 2.4B0 at x = 192
400
x
2
10
• Electrons are adiabatic in right half of
tailward gradient Bz region
300
2
0
0
100
200
300
x
400
500
Entropy
Bubbles and the Entropy Profile
The behavior of bubbles has been shown to be closely
connected to the variation of the entropy as a function
of distance down the tail (Birn et al., 2004).
1
1A
6
300
Constant B
S
It has long been speculated that small, isolated density
depletions (“bubbles”) could form in the mid- or distant
tail and propagate earthward due to buoyancy (Pontius
and Wolf, 1990; Chen and Wolf, 1993).
400
z
200
Tailward Grad
100
Tailward Grad
0
200
300
400
500
x
MHD Interchange Instability:
• Hurricane et al. (1996) showed that the Lembege-Pellat (constant Bz) profile is unstable.
• Maltsev & Mingalev (2000) showed criterion is satisfied for an increasing Bz if dV/dr < 0,
where V = ∫dl/B is the flux tube volume.
• More generally, tail-like equilibria are stable (unstable) when the entropy S increases
(decreases) with distance down the tail (Schindler & Birn, 2004).
The present equilibria have a decreasing entropy profile in the region of the tailward gradient
in Bz and thus are likely to be unstable.
2D Reconnection Bz
(a)
(c)
1
Bz/B0
Bz/B0
2
1.5
1
0.5
0
−50
−40
−30
−20
−10
0
10
0
Density
(b)
(d)
1
2.5
10
2
1.5
1
0.5
0
−50
n/n0
2D simulation of a Harris sheet driven by a
finite duration Ey driving field peaked at x = 0
produces an X line with fronts propagating
both earthward and tailward.
180
200
220
0.5
n/n0
Alternative interpretation: The tailward
gradient region in Bz could also represent the
leading edge of a reconnection front viewed
in the rest frame of the front.
2.5
0
10
180
200
220
−1
−40
−30
−20
x/di
−10
0
10
10
(a)
Bz
Potential
500
0.15
400
0.1
512.
512.
256.
256.
0.
0.
0.05
y
300
0
200
−0.05
100
−0.1
0
0
(b)
100
200
x
Ey
500
0.1
0.
128.
256.
0.
128.
256.
400
0.05
y
300
0
200
−0.05
100
−0.1
0
0
100
200
x
Mode structures develop in right half of tailward gradient
region where electrons are adiabatic. Equatorial plane
structure has character of an interchange mode.
Ue||
Ey
(a)
y = 464
(a)
0.12
100
y = 450
0.6
100
0.4
0.1
50
0.08
0.06
0
z
z
50
0.2
0
0
0.04
0
0
100
200
x
(b)
y = 496
0
100
z
50
0
0
100
200
x
Mode structures also
extend along field lines
earthward toward
ionosphere.
Unlike MHD interchange
modes, there is large
perturbation resulting
from the finite E|| field.
0
100
200
x
(b)
y = 474
0.5
100
50
z
0.02
0
0
0
100
200
x
Mode Intensities
−2
10
Initial case: kyρin ~ 6
3.9
4.7
5.5
6.3
7.1
7.8
8.6
9.4
−4
10
−6
10
ky
in
−8
10
0
10
20
t
i0
40
50
60
Mode Intensities
−2
10
Second case: kyρin ~ 1
30
0.5
1.0
1.6
2.1
2.6
3.1
3.7
4.2
−4
10
−6
10
ky
in
−8
10
0
20
40
60
t
i0
80
100
120
Mode Identification
• Consider linear kinetic analysis where we model δφ and δB|| along equilibrium field lines
and compute the local ion and electron density perturbations at the midplane.
• Ions: Motion is stochastic (destroys gyromotion) and mode localized in x,z to ρin scale
straight line orbits, respond only to electrostatic perturbation
δni/nL = [Z′(s)/2] eδφ/T = -[1 + sZ(s)] eδφ/T,
s = (ω - kyvdi)/kyvTi
δni/nL = -[0.532 + 0.577i] eδφ/T
WRONG PHASE RELATION
Electron orbits average interaction with wave fields over flux tube volume; flux-tube
averaged perturbed ion density becomes
<δni/nL> = -[0.75 +sZ(s)/2] = -[0.515 + 0.289i]eδφ0/T
AGREES WITH SIMULATION
• Electrons: Electron orbits are adiabatic (κe >> 1) but wave and drift frequencies are not
small compared with the bounce frequency. Thus full kinetic response including bounce
and drift resonance interactions must be retained. Find that perturbed electron and ion
flux tube averaged densities are in agreement.
• Perturbed densities yield dispersion relation consistent with the simulations. The dominant polarizations δφ and δB|| are similar to polarization of lower hybrid drift instability
in straight magnetic field geometry; the present ballooning/interchange mode appears to
be related to the low frequency limit of the LHDI.
Effect of Background Density on BICI
(a)
t = 56
i0
nb/n0 = 0
1000
(b)
t = 69
i0
1000
0.45
0.4
800
(c)
t = 103
i0
n /n = 0.08
b 0
1000
1.2
800
1
t = 122
i0
1000
0.8
0.3
800
0.7
800
0.25
0.35
600
(d)
0.8 600
0.3 600
0.6
600
0.5
y
0.2
0.25
400
0.2
400
0.6
0.4
0.15200
200
0.4
400
200
400
0.15
0.3
200
0.2
0.1
0.2
0.1
0.1
0
300
400
x
500
0
200 300 400 500
x
0
200 300 400 500
x
0
0
200
400
x
• Initial case (nb = 0): linear BICI spectrum peaks at kyρin ~ 6 (wavelength ~ ρin); during
earthward propagation, heads narrow and peak Bz increases, leaving several dominant heads
with width ~ 600 - 900 km.
• Second case (nb/n0 = 0.08): peak wavenumber is much smaller (kyρin ~ 1.0 - 1.5). Nonlinear
evolution produces single dominant head with width ~ 5000 - 6000 km. This is now
compatible with observed scale of DFs.
Equatorial Plane DF Structure
Profile at y = 630
1
(a)
Bz
Bz/B0
• Transition thickness of Bz (from 0.25 to 0.80) occurs
over distance of 11 Δ = 1.1 di.
0.5
• Density drops by factor of 3 across the front.
0
100
8
• Ey strength increases with Bz to a peak of ~ 11 mV/m
150
200
250
(b)
Density
n/n0
6
4
as opposed to leaving
amplification of front.
0
100
0.2
0
−0.2
−0.4
−0.6
−0.8
100
0
150
200
250
(c)
cEy/vTiB0
• Electron bulk flow exceeds the ion behind the front
but is smaller ahead of front. Magnetic flux frozen in
to electron flow net increase in flux entering front
2
E
y
150
200
250
(d)
−0.5
Ux/vTi
• Ion bulk flow speed increases only slowly across front
and reaches max. behind the front [“growing” DF Fu et al., 2011]
−1
−1.5
100
U
ix
Uex
150
200
x
250
Ωi0t = 127
Breakup of DF Head
and FACs
Bz, Jx,y
B
z
700
800
750
1
690
700
1.2
y
• As DF propagates earthward,
Bz increases; breakup occurs
when local ρin < DF thickness.
0.5
650
680
600
• Multiple sub-heads comparable
to local gyroradius.
550
50
800
• Pair of Region 1 sense FACs
associated with each subhead.
• FAC structuring of the DF is
manifested in auroral streamer.
A number of such structured
streamers have been observed
by THEMIS.
J||
10
750
0.8
660
5
y
700
650
0
600
−5
550
50
670
100 150 200 250
y
• Cross-tail current Jy diverted
around each of the heads.
1
0
100 150 200 250
x
650
0.6
640
0.4
630
620
0.2
610
0
600
100
110
120
x
130
THEMIS ASI OBSERVATION
Structured Auroral Streamer
70°
100 km
Streamer
65°
FSIM
Streamer breaking into substructures
FSMI
Size at 110 km:
Map to plasma sheet:
Simulation:
substructure ~ 30 km
~ 1000 km width
subhead ~ 800 km
poleward portion ~ 100 km
~ 3000 km
upward FAC region ~ 4000 km
x = 100, y = 620, z = 20
• Away from the center of the plasma sheet, the DF
is strongly modulated in time in association with
intense wave activity.
0
115
120
125
130
120
125
130
125
130
125
130
2.8
2.6
(b)
|B |
x
2.4
2.2
115
(c)
Density
n/n0
2
1
0
cEy/vTiB0
• cEy/VTiB0 = 2 - 3 or 30 - 50 mV/m.
• EM waves with δE/δB ~ 4 VA.
Bz
0.2
• This activity is a precursor to the actual arrival of
the DF.
• The modulation is most apparent in Ey, Bx, and
density, with less pronounced variation in Bz.
(a)
B/B0
0.4
B/B0
Off-Axis Properties of DF
at the edge of the density gradient
0.6
2
115
(d)
120
E
y
0
−2
• Similar intense waves much closer to the axis
have been reported by M. Zhou et al. (2014),
Bz >> Bx.
Front position:
115
x=
208
120
t
i0
138
70
Observation of large-amplitude magnetosonic waves at dipolarization fronts
M. Zhou et al. 2014
Journal of Geophysical Research: Space Physics
2014JA019796, 2 JUN 2014 DOI: 10.1002/2014JA019796
http://onlinelibrary.wiley.com/doi/10.1002/2014JA019796/full#jgra51055-fig-0003
Intense EM Waves
2
20
1
t
125
i0
−20
0
−2
120
−40
500
(b) E
600
y
y
−1
700
−2
y = 628
115
2
−3
z
20
500
0
0
(c) Ey
50
100
x
150
200
3
y = 636
2
125
20
−0.5
z
4
130
0
0
1
0
120
−1
−1
−2
−20
−1.5
−40
700
Ey y = 622, z = 20
t
40
650
i0
−40
600
y
(b)
• Frequency = 0.24 ωlh
= 0.96 Ωi,loc
550
1
0
−20
• Standing wave structure
along field lines
2
0
0
40
• Wavelength = 0.7 x (local di)
= 700 km
3
130
1
−1
• Waves propagate duskward
at 0.51 VTi
Ey x = 100, z = 20
(a)
Ey x = 112
40
z
• Wave activity concentrated
in region of DF and extends
somewhat duskward.
(a)
0
50
100
x
150
200
115
−3
0
100
200
x
y = 618
Particle Distributions in
regions of intense waves
x = 110, z = -22
Large net parallel drift for
electrons that reverses
sign depending on wave
phase; net ion - electron
parallel drift ~ 3.5 VTi.
Likely to drive electromagnetic ion cyclotron
instability [Perraut et al.,
2000].
4.
4.
0.
0.
-4.
Also strong ∂fe/∂v⊥ > 0
slope electron cyclotron
harmonic waves.
4.
16.
16.
0.
0.
-16.
Zhou et al. (2014) attribute
waves to ∂fi/∂v⊥ > 0 and
identify it as ion Bernstein.
y = 626
16.
-4.
4.
-16.
16.
E J y = 620
Dissipation
E J y = 620
40
40
20
20
2
0
2
−20
−2
−2
0
50
100
150
E J y = 628
−40
200
40
2
−2
−20
−4
0
50
100
150
150
200
E J y = 644
2
0
0
−40
200
−2
0
50
100
150
E J y = 644
200
40
20
10
0
5
−20
0
0
50
100
x
150
200
6
20
z
z
100
E J y = 628
−20
40
−40
50
20
z
0
0
−40
0
40
20
• Strong dissipation is
localized (on slightly
sub-di scale) at the sub
heads of the DF,
E′ . J ~ 2 nW/m3.
0
0
0
−20
−40
z
z
4
z
• In wave region alternating regions of generator
(E . J < 0) and load
(E . J > 0); very little
dissipation (E′. J > 0),
where E′ = E + U x B.
4
6
4
0
2
−20
0
−40
−2
0
50
100
x
150
200
Poynting Vector Components
Sx y = 628
Sy y = 628
40
Sz y = 628
40
40
0.4
0.2
20
20
0
−0.4
0
0
50
100
150
200
−0.8
0
0
−2
−2
−20
−20
−0.6
−40
2
z
z
z
−0.2
−20
4
20
0
0
6
2
−4
−4
−40
0
50
Sx y = 636
100
150
−6
−40
200
0
50
Sy y = 636
40
150
200
Sz y = 636
40
40
4
0.5
20
100
5
20
2
20
0
0
0
−0.5
−20
−40
−1
0
50
100
x
150
200
z
0
z
z
0
−20
−2
−20
−40
−4
−40
0
50
100
150
200
0
−5
0
x
Intense Poynting Vector Fluxes:
• Circulating cell patterns (north-dawn-south-dusk) ~ 1 mW/m2
• Alternating components to and from ionosphere ~ 0.1 mW/m2
50
100
x
150
200
Properties of Reconnection Fronts
• Multi-point studies suggest that full width of high-speed flow channels in plasma sheet
are of order 1 - 3 RE (Sergeev et al., 1996; R. Nakamura et al., 2004).
• Can such finite cross-tail flows be produced directly by reconnection?
• Hall MHD studies indicate that expansion of a finite reconnection region proceeds at a
rate given by the relative electron - ion flow speed (Huba and Rudakov, 2002; Shay et al.,
2003; TKM Nakamura et al., 2012).
netohydrodynamic effects for three-dimensional magnetic reconnection
with finite
Hall magnetohydrodynamic
effects for three-dimensional magnetic reconnection wit
width along the direction of the current
width along the direction of the current
Ions carry initial current
sical Research: Space Physics
3, A03220, 16 MAR 2012 DOI: 10.1029/2011JA017006
ley.com/doi/10.1029/2011JA017006/full#jgra21592-fig-0001
Electrons carry initial current
Journal of Geophysical Research: Space Physics
Volume 117, Issue A3, A03220, 16 MAR 2012 DOI: 10.1029/2011JA017006
http://onlinelibrary.wiley.com/doi/10.1029/2011JA017006/full#jgra21592-fig-0002
• Use 3D PIC simulation in which localized region is initiated by periodically blocking the
Jy current density in the center of the current sheet (Pritchett & Coroniti, 2002). This
represents the upper limit to the effects that can be produced by the onset of an
anomalous resistance within a growth phase thin near-Earth current sheet. The blocking
leads to an enhancement of the Ey and to the formation of a localized reconnection
region.
• Initial blocking region: 0 < x < 3di, 30di < y < 34di
• System size: -64di < x < 64di, 0 < y < 64di
J
J
iy
36
0
y
36
0.1
34
34
y/di
−0.2
32
0
32
0
2
x/di
4
6
8
28
−4 −2
36
0
34
−0.2
−0.1 32
−0.4
−0.4 30
30
28
−4 −2
E/ t
ey
−0.2 30
0
2
x/di
4
6
8
−0.3 28
−4 −2
−0.6
0
2
x/di
4
6
8
Bz
Add 0.02 normal field and
vary width of blocking
region.
50
0.5
y/di
40
30
0
20
−0.5
−50
Bz
0
t = 75
i0
50
0.5
30
0
20
−0.5
−50
0
0
50
t = 75
i0
x/di
1
30
0
20
−1
−50
Bz w = 8di
0
50
x/di
Density w = 8di
1.5
1
y/di
20
0.5
0
10
4
30
3
20
2
10
1
−0.5
0
• Appears likely to have a
minimum width of front.
Uix
40
10
50
30
• Breakup less pronounced
on tailward propagating
front.
−50
50
40
10
0.5
20
10
50
1
30
y/di
10
1.5
40
y/di
50
t = 75
i0
y/di
• Extended front appears to
break up on scales of a few
di.
Density
y/di
• Max Bz follows electrons
slightly dawnward, but the
jet front is blown duskward
by the ion flow; elongated
front on scale of tens of di
>> initial perturbation.
t = 47
i0
−20
0
x/di
20
0
−20
0
x/di
20
Summary/Questions
• Key feature of plasma sheet is presence of a finite Bz component. This necessitates a
kinetic treatment of plasma sheet dynamics.
• Spontaneous tearing instability appears to be unlikely due to magnetization of the electrons.
How can reconnection be initiated?
• Jet (dipolarization) fronts are common feature of the tail; possess kinetic scale thickness.
What determines the cross-tail extent of ~ 1 - 3 RE?
• Ballooning/interchange mode reproduces the key features of the fronts and predicts the
structuring of auroral steamers and the presence of intense off-axis EM waves near Ωci.
Recent THEMIS observations offer confirmatory examples of these features.
• But how common are the conditions of decreasing entropy with radial distance down the
tail that appear to be necessary for excitation of the modes?
• Can localized perturbations (perhaps from the dayside) produce reconnection jets that
mimic the properties of the BICI mode?
• What happens when these fronts impact a growth-phase thinned near-Earth plasma sheet
system?