a

r
X
i
v
:
1
4
0
7
.
3
2
9
8
v
1


[
a
s
t
r
o
-
p
h
.
S
R
]


1
1

J
u
l

2
0
1
4
3D MHD simulation of post–flare supra–arcade downflows
in a turbulent current sheet medium
M. C´ ecere
1,2
, E. Zurbriggen
1,2
, A. Costa
1,2,3
, M. Schneiter
1,2,3
ABSTRACT
Supra–arcade downflows (SADs) are sunward, generally dark, plasma density depletions originated
above posteruption flare arcades. In this paper using 3D MHD simulations we investigate if the SAD
cavities can be produced by a direct combination of the tearing mode and Kelvin–Helmholtz instabilities
leading to a turbulent current sheet (CS) medium or if the current sheet is merely the background where
SADs are produced triggered by an impulsive deposition of energy. We find that to give account of the
observational dark lane structures an addition of local energy provided by a reconnection event is required.
This local reconnection can trigger a nonlinear internal wave dynamic, generated by the bouncing and
interfering of shocks and expansion waves that compose relatively stable voids.
Subject headings:
1. Introduction
Due to the very large value of the coronal magnetic
Reynolds or Lundquist number, the magnetic energy
can only be released in confined regions where the
magnetic field forms small scales and steep gradients
leading to the formation of current sheets (CS). These
regions, where the field line topology significantly de-
viates from a potential configuration have free mag-
netic energy and can eventually become unstable, e.g.,
when a large Parker angle is reached or when the gra-
dients exceed a critical value (Vlahos and Georgoulis
2004). The formation of smaller scale structures, al-
lowing larger diffusion and faster reconnection rates, is
a consequence of the nonlinear evolution that eventu-
ally ends in a turbulent regime of many spatial scales,
where the smaller ones decrease with increasing dis-
tance from the CS. Associated with the fragmenta-
tion of large CSs into plasmoids and smaller CSs,
(Guo et al. 2013) the coalescence of magnetic islands
is generally observed. These authors, based on obser-
vational data and using a 2D simulation estimated a
1
Instituto de Investigaciones en Astronom´ıa Te´ orica y Experi-
mental IATE, C´ ordoba, Argentina.
2
Consejo Nacional de Investigaciones Cient´ıficas y T´ ecnicas
(CONICET), Argentina.
3
Facultad de Ciencias Exactas, F´ısicas y Naturales, Universidad
Nacional de C´ ordoba (UNC), C´ ordoba, Argentina
CS width of 3Mm. However, Ciaravella and Raymond
(2008) reported CS widths ranging within ∼ [28 −
56]Mm. This discrepancy could be due to the impos-
sibility to observationally distinguish between the CS
and a sheath of hot plasma surrounding the CS as sug-
gested by Seaton and Forbes (2009).
Recently McKenzie (2013) analyzed high–resolution
observations in a sheet–like structure above a post–
CME flare arcade where the turbulent dynamic of a
complex flow is described. He found the plasma β
(the ratio of gas to magnetic pressure) is of the or-
der of unity, and described the flow variability in the
hot plasma (T > 10MK) identifying small vortices
moving towards the arcade as probable supra–arcade
downflows (SADs).
SAD features are known to be dark moving trails
originated [40 − 60]Mm above posteruption flare ar-
cades with decelerating speeds in the range ∼ [50 −
500]km s
−1
(McKenzie 2000; McKenzie and Savage
2009; Savage and McKenzie 2011). They were first
detected with the Yohkoh Soft X–ray Telescope (SXT)
(McKenzie and Hudson 1999). Since then, they have
been extensively reported using other instruments
such as TRACE (Innes et al. 2003a,b), SOHO/SUMER
(Innes et al. 2003b) and SDO/AIA(Savage et al. 2012).
There is a consensus that due to the lack of X–ray and
extreme–ultraviolet (EUV) signatures in images and
spectra, these SAD structures are voided flows gener-
1
ated by reconnection processes in a CS above the flare
arcade.
Several scenarios were proposed to give account
of the observations. SADs have been interpreted
as cross–sections of thin and empty flux tubes re-
tracting from a reconnection site high in the corona
(McKenzie and Savage 2009; Savage and McKenzie
2011). However, new AIA observations with high
resolution and cadence, allowed a reinterpretation of
SADs (Savage and McKenzie 2011) as wakes created
by the retraction of thin loops (renamed by the au-
thors as SADLs). SADLs ought to be features of sizes
[∼ 0.9 − 1.3]Mm observed during the early phase of
the eruptive event and SADs should be features of
sizes ∼ 9Mm that become apparent afterwards. These
authors proposed that deceleration is expected due to
the buildup of downstream magnetic pressure and/or
drag mechanisms.
Linton et al. (2009) proposed that the dynamic of
retracting magnetic fields is triggered by a localized
reconnection event that produces up and down flow-
ing reconnected flux tubes, which are slowed down by
underlying magnetic arcade loops. A drawback with
this scenario is that the observed SADspeeds are lower
than expected for reconnection outflows in regions of
typical Alfv´ en speeds of 1000km s
−1
. Linton et al.
(2009) suggested that drag forces could work against
the reconnection outflow.
Verwichte et al. (2005) analyzed TRACE SAD os-
cillations transverse to the magnetic field. They found
that the initial speeds and the displacement amplitudes,
of kink–like type in the observational dark lanes of
variable sizes [∼ 2 − 9]Mm, decrease as they propa-
gate downwards, while the period remains fairly con-
stant with height.
In Costa et al. (2009), Schulz et al. (2010), Maglione et al.
(2011) and C´ ecere et al. (2012) we reproduced the dy-
namics of multiple decelerating downflows through
the assumption that the dark tracks are confined
voided cavities –of high β and T > 20 MK– colli-
mated in the direction of the ambient magnetic field
and generated by the bouncing and interfering of
shocks and expansion waves upstream of the initial
localized deposition of energy provided by recon-
nection events. In this scenario, the different ob-
servational sizes distinguished as SADLs and SADs
by Savage and McKenzie (2011) could be interpreted
as reconnection events that are triggered in a back-
ground homogeneous media (SADLs) or as recon-
nection events triggered in a background media with
previous SAD formations, respectively. We found that
the observed wavy character (Verwichte et al. 2005)
can be interpreted as an indication of interaction be-
tween SADs. This interaction is significative when the
bursts that trigger the phenomenon act on the wakes
left by previous SADs. This wavy character is en-
hanced with the strength of the reconnection and/or
with lower values of the magnetic field intensity.
Motivated by the description provided by McKenzie
(2013) we explore a new scenario. We consider a tur-
bulent CS as the media where SAD features are ob-
served. Our intention is to numerically analyze the
capability of a CS to generate and sustain the structure
and dynamic of SAD cavities with the observational
characteristics reported in literature.
2. Numerical code and initial conditions
The software used in this work was in part devel-
oped by the DOE NNSA-ASC OASCR Flash Cen-
ter at the University of Chicago. We performed the
3D MHD simulations with the extensively validated
FLASH4 code ( f lash.uchicago.edu/site/publications).
We used the adaptive mesh refinement procedure
(AMR) with the Powell’s 8-wave scheme (Powell et al.
1999) to solve the MHD equations.
To generate a turbulent CS we start from a mono-
lithic CS configuration with Lundquist value S 10
10
,
greater than the critical value (S
c
∼ 10
4
) at which the
Sweet–Parker CS becomes unstable. In this way, a
turbulent regime is generated giving account of a hi-
erarchical configuration of overdense plasma features
connected by secondary CSs as described in the litera-
ture (see references in Priest and Forbes (2000)). The
ideal regime is set with an initial numerical diffusivity
of η ≈ 1m
2
s
−1
, which is a the classical Spitzer coro-
nal diffusivity for a T ∼ 1MK corona (η ∼ 10
9
T
−3/2
).
As we choose a typical CS background temperature
of T = 13MK the Spitzer diffusivity is almost 50
times larger than the classical coronal one, yet lower
than the estimated anomalous diffusivity (Bemporad
2008). We set the background numerical density to be
n = 4 × 10
9
cm
−3
. A cartesian grid with 4 levels of
refinement was employed leading to a maximum grid
refinement of (128, 256, 64). The physical domain was
set to (50, 100, 25)Mm, with the y coordinate pointing
sunwards and the x and z coordinates perpendicular
to the initial magnetic field direction. The basic CS
device is initializated assuming a corona in pressure
equilibrium with a uniform random velocity in the x
2
direction given by
pert = v
0
sin(2ωy) ×random ω =

y
max
− y
min
(1)
where v
0
= 2km s
−1
, random is the uniform random
distribution (varying between [0−1]) and y
max
−y
min
is
the domain size in the y direction. The magnetic field
configuration is:
B
y
=

B
0
if x < 0
−B
0
if x ≥ 0
(2)
where B
0
depends on the model (see Table 1). No
boundary effects are expected in the development of
turbulence, thus, periodic boundary conditions are im-
posed in the y and z directions.
3. Results and Discussion
A turbulent picture
From the initial conditions of the coronal plasma
parameters (model M0) we obtain a numerical turbu-
lent configuration. Density slices of the z = 0 plane
are shown in Figure 1 for times: 17.6min and 35.2min.
As shown in the figure, and extensively described in
literature, the tearing mode instability leads to a turbu-
lent regime composed of dynamic and coalescent plas-
moids where the desired subdense structures are only
obtained as secondary linear CSs (they connect neigh-
bour plasmoids). Thus, the usual tear–drop–shaped
SAD features are not easy to obtain from models of
the M0 type, in times comparable with their observa-
tions.
However, if a shear in the flow speed to the sides of
the CS is introduced (models M1, M2 and M3 of Ta-
ble 1), the turbulent features change markedly allow-
ing the appearance of subdense cavities. As stated by
Soler et al. (2012), a collisionless plasma becomes un-
stable when super–Alfv´ enic shears are present. In this
conditions, the shear initiates Kelvin–Helmholtz per-
turbations that, combined with the tearing instability,
modify the overall dynamic. Accordingly, McKenzie
(2013) reported strong coronal velocity shears (up to
≃ [250 − 350]km s
−1
). Also, as initially large down-
flow values (v
f low
≃ v
A
≃ [400 − 500]km s
−1
) are
expected coming for a reconnection site, the inho-
mogeneities of the flaring media could lead to strong
shears in the flow speed (see e.g. the flow speed com-
ing from the right side into the fan structure in slices
between 11 : 58 : 09 and 12 : 02 : 09 of movie 2 by
Savage et al. (2012)).
We thus performed several runs, with initial strong
shear velocities (see Table 1), to gain insight into the
turbulent features. Figure 2 shows density slices of
the z = 0 plane for M1, and for the same times than
in Figure 1. The model configuration is the same as
in M0, adding a random speed with a shear in the y
direction. Comparing Figure 2 and Figure 1 we note
that models with shear can generate subdense cavities,
lasting for times of the order of decades of minutes, as
the main features of the turbulent regime.
We also performed runs for models M2 and M3
with a shear in the z direction. Left panel of Fig-
ure 3 shows the density slice for M2, at t = 7.2min,
of a) plane z = 0 and b) plane x = 0. The right
panel of the figure shows the density slice for M3, at
4min, of c) plane z = 0 and d) plane x = 0. As in
Figure 2, Figure 3a and Figure 3c show evolving sub-
dense structures. The viewing orientation of Figure 3b
and Figure 3d is perpendicular to the CS plane. These
x = 0 plane descriptions resemble observational in-
homogeneities as seen, for example, in movies 1 and
2, or figure 1 by Savage et al. (2012). Nevertheless,
these numerical sunward subdense features are not sta-
ble during times comparables with the observation of
that inhomogeneities, they cannot sustain their shape
for more than 1min.
The emission measure
To analyze if the subdense features seen in Fig-
ures 2 and 3 are compatible with a SAD descrip-
tion we first evaluate the emission measure (EM)
(Aschwanden 2005) as:
EM =

n
2
dz (3)
Figure 4 shows the EMof the CS viewfor M1 and M2.
The integration along the line of sight direction (z) was
performed considering a CS width of 4Mm (see esti-
mations in Guo et al. (2013)). The temperature range
obtained is [12.5 − 13.8]MK. The weak contrast ob-
tained for the EM, where the subdense cavities have
EMs which are less than ∼ 1.3 the background fea-
tures, is lower than the reported in Savage et al. (2012)
for the [10 − 13]MK range, where the values of EM in
SADs were factor of four with the surroundings (see
Figure 4 of the paper). Thus, the turbulent picture by
itself is not sufficient to fully give account of dark ob-
servational regions.
3
Model B
0
[G] v
y
[km sec
−1
](x < 0|x > 0) v
z
[km sec
−1
](x < 0|x > 0) (∆P/P)
M0 3 0|0 0|0 0
M1 3 (0| − 414) + pert 0|0 0
M2 0.3 0|0 (0| − 207) + pert 0
M3 3 0|0 (0| − 414) + pert 0
M4 3 (0| − 414) + pert 0|0 4
Table 1: Simulated models: B
0
the background magnetic field, v
i
the velocity in the i direction, x = 0 the position of
the CS, pert = v
0
sin(2ωy) ×randomthe perturbation in the i direction (same functional dependence for all directions).
Numerical density [10
9
cm
-3
]
x [Mm]
0.0 -10.0 -20.0 20.0 10.0
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
3.995 4.001 4.997 4.013 4.019
t = 17.6min t = 35.2min
a) )
Fig. 1.— Simulation of M0, density slices of the plane
z = 0 at a) t = 17.6min and b) t = 35.2min.
2.870 3.287 3.704 4.122 4.539 3.396 3.596 3.796 3.996 4.196
Numerical density [10
9
cm
-3
]
t = 17.6min t = 35.2min
x [Mm]
0.0 -10.0 -20.0 20.0 10.0
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
a !
Fig. 2.— Simulation of M1, density slices of the plane
z = 0 at a) t = 17.6min and b) t = 35.2min.
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
x [Mm]
0.0 -10.0 -20.0 20.0 10.0
z [Mm]
-2.5 -7.5 7.5 2.5
Numerical density [10
9
cm
-3
]
3.604
3.732
3.860
3.988
4.117
z = 0 x = 0
a !
3.017
3.369
3.721
4.074
4.426
x [Mm]
0.0 -10.0 -20.0 20.0 10.0
z [Mm]
-2.5 -7.5 7.5 2.5
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
Numerical density [10
9
cm
-3
]
z = 0 x = 0
c) d)
Fig. 3.— Simulation of model M2 at 7.2min, density
slice: a) CS plane (x = 0) and b) z = 0 plane. Simu-
lation of model M3 at t = 4min, density slice: c) CS
plane and d) z = 0 plane.
4
0.71
0.73
0.74
0.76
0.77
0.64
0.68
0.73
0.78
0.83
z [Mm]
-2.5 -7.5 7.5 2.5
Emission Measure [10
28
cm
-5
]
z [Mm]
-2.5 -7.5 7.5 2.5
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
a) b)
Fig. 4.— CS view (x = 0) of EM for models a) M1 at
17.6min and b) M2 at 7.2min.
2.01
2.55
3.08
3.62
4.15
13.1
16.0
18.9
21.8
24.6
n [10
9
cm
-3
] T [MK]
x [Mm]
0.0 -10.0 10.0
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
x [Mm]
0.0 -10.0 10.0
a) b)
Fig. 5.— Slices of the z = 0 plane for Model M4 at
t = 44min: a) numerical density, n and b) temperature,
T.
Later energy depostions
We imposed a pressure pulse to emulate a new re-
connection event occurring high in the corona, where
reconnection is prone to occur triggered by a local
change in the magnetic field line linkage and the mag-
netic topology, e.g., null points, separatrices or 3D
quasi–separatrix layers, (see Demoulin et al. (1996)).
Such localized processes that lead to bursts of im-
pulsive deposition have been proposed in Scott et al.
(2013); C´ ecere et al. (2012); Maglione et al. (2011);
Costa et al. (2009); Cassak et al. (2013). Figure 5a–b
shows one of these events (model M4) resulting from
a pressure pulse with P = 4P
0
. The figure shows the
z = 0 slice for the numerical density and the tempera-
ture at t = 44min. We obtained a tear–drop SAD trav-
eling sunwards with a speed of ∼ 155kms
−1
, leaving a
persistent voided region along a distance of ∼ 38Mm.
The SADs numerical density is at least half that of the
background, and its temperature is 23MK, whereas the
numerical density of the subdense cavities, obtained
through turbulence, was only ∼ 1.13 times less than
the background (see Figure 2) and the temperature was
13MK. Moreover, the temperature peak occurs sur-
rounding the subdense cavities. In accordance with the
observations (McKenzie 2013) we find that the turbu-
lent background has almost uniformβ ≥ 1 values, with
no appreciable change at the location of the SAD. As
in C´ ecere et al. (2012), we also find an almost uniform
value of the total pressure in the simulation domain.
Thus, this suggests that the SADs can sustain their
structure compensating the density depletion with the
higher inside temperatures. Also, following the tempo-
ral sequence we can see a wavy evolution of the SAD
tail.
Wavy behavior
To analyze the wavy behavior of the subdense re-
gions we show in Figure 6a–b the z = 0 slice of the
velocity (arrows), superimposed to the numerical den-
sity, for the turbulent vortice (M1) of Figure 2a and the
SAD structure (M4) of Figure 5a, respectively. As in
McKenzie (2013) we obtain a variable spectrum of ve-
locities and vortice–like frames which correlate with
the motion of density depletions. Figure 6c–d is as
6a–b but the arrows are indicating the magnetic field
vector. The better correlation between the magnetic
field and the plasma density suggests a greater effi-
ciency of the magnetic field to outline the subdense
cavities. Despite the almost uniform values of the to-
5
tal pressure and the β parameter, there is an important
difference. As pointed out by McKenzie and Savage
(2009); Savage and McKenzie (2011), Figure 6c sug-
gests that the large inside magnetic pressure could be
the reason that the voids are able to resist being filled in
immediately by the surrounding plasma. Moreover, we
find that the center of the subdense region has values
of β slightly less than one. On the contrary Figure 6d
shows that the magnetic field is larger outside, with
appropiate larger values of the inside plasma pressure,
tracing the motion of the SAD (C´ ecere et al. 2012).
In both cases, the eddy–like features of sizes ∼
20Mm have an average speed of 50km s
−1
. Un-
like the interpretation done in C´ ecere et al. (2012),
where the wavy behavior is due to the interaction of
SADs between each other and with the inhomoge-
neous medium, Figure 6b,d shows that the wavy tail
can be produced only by the interaction of a SAD with
the turbulent background.
SADs emission measure
Finally, figure 7a–b shows the CS view of the
emission measure for t = 40min and t = 44min for
model M4. The EM was calculated along the line
of sight (CS-direction) considering the whole temper-
ature range and assuming a thickness of the CS of
∼ 4Mm (Guo et al. 2013). As seen in the observations
(e.g., see the high cadence movies of Savage et al.
(2012)), at the initial times the SADs appear as rapidly
increasing spherical features (see Figure 7a). Later
(Figure 7b), they acquire a tailed tear–drop shape
while they are elongated and collimated by the tur-
bulent background plasma. As stated in Savage et al.
(2012), during times comparable with the observa-
tions, the SAD emission measure is approximately 3
times lower than its background value.
4. Conclusions
In this paper, motivated by recent CS observations
(McKenzie 2013; Savage et al. 2012), we numerically
simulate a turbulent CS generated by a combination of
the tearing instability and the Kelvin–Helmholtz one
that developed generating subdense cavities with tran-
sient and variable vortical motions. While comparing
these features with SAD observations we found that
the EM contrast and the characteristic times were not
enough to match the observations. However, imposing
a pressure pulse to this turbulent background -in order
to emulate a local deposition of energy produced by an
y

[
M
m
]
0
.
0
-
1
0
.
0
-
2
0
.
0
2
0
.
0
1
0
.
0
y

[
M
m
]
2
0
.
0
1
0
.
0
0
.
0
4
0
.
0
3
0
.
0
5
0
.
0
x [Mm]
0.0 -5.0 5.0 10.0 15.0 -10.0 -15.0
x [Mm]
0.0 -5.0 5.0 10.0 15.0 -10.0 -15.0
v = 350km s
-1
v = 350km s
-1
B = 2.4G B = 1.6G
a) b)
c) d)
Fig. 6.— Velocity of the z = 0 view for models a) M1
at t = 17.6min and b) M4 at t = 43min.
0.33
0.43
0.54
0.64
0.75
z [Mm]
-2.5 -7.5 7.5 2.5
0.29
0.44
0.58
0.72
0.86
z [Mm]
-2.5 -7.5 7.5 2.5
y

[
M
m
]
0
.
0
-
2
0
.
0
-
4
0
.
0
4
0
.
0
2
0
.
0
Emission Measure [10
28
cm
-5
]
a) b)
Fig. 7.— CS view of the EM for model M4, at a) t =
40min and b) t = 44min.
6
impulsive reconnection event-, we find that the EM,
the characteristic times and the other plasma parame-
ters are capable to justify the observed data.
We want to emphasize that the downflowing SADs
seem to require more than an internal thermodynamic
configuration that allows to sustain their structure for
times comparable with the observations, i.e., due to a
constant total pressure of the phenomenon. While seen
the numerous movies that captured the phenomenon, it
appears as if energetic internal SADprocesses were re-
quired to support their dynamic sunward motion. We
think that a nonlinear internal wave structure, gener-
ated by the bouncing and interfering of shocks and
expansion waves allowing a dynamic composition of
local voids, is at the heart of the explanation of the
phenomenon (C´ ecere et al. 2012). If this scenario is
accurate -as a difference with the unstable subdense
background vortices-, the wave internal configuration
of the SADs provides a unified structure acting as a dy-
namical pressure which resists to disaggregate, how-
ever subject to a variable background that can make
them acquire a wavy tail.
REFERENCES
Aschwanden, M. J. (2005). Physics of the Solar
Corona. An Introduction with Problems and Solu-
tions (2nd edition).
Bemporad, A. (2008). Spectroscopic Detection of
Turbulence in Post-CME Current Sheets. ApJ,
689:572–584.
Cassak, P. A., Drake, J. F., Gosling, J. T., Phan, T.-D.,
Shay, M. A., and Shepherd, L. S. (2013). On the
Cause of Supra-arcade Downflows in Solar Flares.
ApJ, 775:L14.
C´ ecere, M., Schneiter, M., Costa, A., Elaskar, S., and
Maglione, S. (2012). Simulation of Descending
Multiple Supra-arcade Reconnection Outflows in
Solar Flares. ApJ, 759:79.
Ciaravella, A. and Raymond, J. C. (2008). The Current
Sheet Associated with the 2003 November 4 Coro-
nal Mass Ejection: Density, Temperature, Thick-
ness, and Line Width. ApJ, 686:1372–1382.
Costa, A., Elaskar, S., Fern´ andez, C. A., and Mart´ınez,
G. (2009). Simulation of dark lanes in post-flare
supra-arcade. MNRAS, 400:L85–L89.
Demoulin, P., Henoux, J. C., Priest, E. R., and Man-
drini, C. H. (1996). Quasi-Separatrix layers in solar
flares. I. Method. A&A, 308:643–655.
Guo, L.-J., Bhattacharjee, A., and Huang, Y.-M.
(2013). Distribution of Plasmoids in Post-coronal
Mass Ejection Current Sheets. ApJ, 771:L14.
Innes, D. E., McKenzie, D. E., and Wang, T. (2003a).
Observations of 1000 km s
−1
Doppler shifts in 10
7
K solar flare supra-arcade. Sol. Phys., 217:267–
279.
Innes, D. E., McKenzie, D. E., and Wang, T. (2003b).
SUMER spectral observations of post-flare supra-
arcade inflows. Sol. Phys., 217:247–265.
Linton, M. G., Devore, C. R., and Longcope, D. W.
(2009). Patchy reconnection in a Y-type current
sheet. Earth, Planets, and Space, 61:573–576.
Maglione, L. S., Schneiter, E. M., Costa, A., and
Elaskar, S. (2011). Simulation of dark lanes in post-
flare supra-arcades. III. A 2D simulation. A&A,
527:L5.
McKenzie, D. E. (2000). Supra-arcade Downflows
in Long-Duration Solar Flare Events. Sol. Phys.,
195:381–399.
McKenzie, D. E. (2013). Turbulent Dynamics in Solar
Flare Sheet Structures Measured with Local Corre-
lation Tracking. ApJ, 766:39.
McKenzie, D. E. and Hudson, H. S. (1999). X-Ray
Observations of Motions and Structure above a So-
lar Flare Arcade. ApJ, 519:L93–L96.
McKenzie, D. E. and Savage, S. L. (2009). Quanti-
tative Examination of Supra-arcade Downflows in
Eruptive Solar Flares. ApJ, 697:1569–1577.
Powell, K. G., Roe, P. L., Linde, T. J., Gombosi, T. I.,
and De Zeeuw, D. L. (1999). A Solution-Adaptive
Upwind Scheme for Ideal Magnetohydrodynamics.
Journal of Computational Physics, 154:284–309.
Priest, E. and Forbes, T. (2000). Magnetic Reconnec-
tion.
Savage, S. L. and McKenzie, D. E. (2011). Quan-
titative Examination of a Large Sample of Supra-
arcade Downflows in Eruptive Solar Flares. ApJ,
730:98.
7
Savage, S. L., McKenzie, D. E., and Reeves, K. K.
(2012). Re-interpretation of Supra-arcade Down-
flows in Solar Flares. ApJ, 747:L40.
Schulz, W., Costa, A., Elaskar, S., and Cid, G. (2010).
Simulation of dark lanes in post-flare supra-arcades
- II. A contribution to the remote sensing of the
coronal magnetic field. MNRAS, 407:L89–L93.
Scott, R. B., Longcope, D. W., and McKenzie, D. E.
(2013). Peristaltic Pumping near Post-coronal Mass
Ejection Supra-arcade Current Sheets. ApJ, 776:54.
Seaton, D. B. and Forbes, T. G. (2009). An Analyti-
cal Model for Reconnection Outflow Jets Including
Thermal Conduction. ApJ, 701:348–359.
Soler, R., D´ıaz, A. J., Ballester, J. L., and Goossens,
M. (2012). Kelvin-Helmholtz Instability in Par-
tially Ionized Compressible Plasmas. ApJ, 749:163.
Verwichte, E., Nakariakov, V. M., and Cooper, F. C.
(2005). Transverse waves in a post-flare supra-
arcade. A&A, 430:L65–L68.
Vlahos, L. and Georgoulis, M. K. (2004). On the Self-
Similarity of Unstable Magnetic Discontinuities in
Solar Active Regions. ApJ, 603:L61–L64.
This 2-column preprint was prepared with the AAS L
A
T
E
X macros
v5.2.
8

Sign up to vote on this title
UsefulNot useful