Original citation:

Osman, K. T., Matthaeus, W. H., Gosling, J. T., Greco, A., Servidio, S., Hnat, B.,
Chapman, Sandra C. and Phan, T. D.. (2014) Magnetic reconnection and intermittent
turbulence in the solar wind. Physical Review Letters, Volume 112 . Article number
215002.

http://dx.doi.org/10.1103/PhysRevLett.112.215002

Permanent WRAP url:

http://wrap.warwick.ac.uk/61932

Copyright and reuse:

The Warwick Research Archive Portal (WRAP) makes this work by researchers of the
University of Warwick available open access under the following conditions. Copyright ©
and all moral rights to the version of the paper presented here belong to the individual
author(s) and/or other copyright owners. To the extent reasonable and practicable the
material made available in WRAP has been checked for eligibility before being made
available.

Copies of full items can be used for personal research or study, educational, or not-for-
profit purposes without prior permission or charge. Provided that the authors, title and
full bibliographic details are credited, a hyperlink and/or URL is given for the original
metadata page and the content is not changed in any way.

Publisher’s statement:
© 2014 American Physical Society

A note on versions:
The version presented in WRAP is the published version or, version of record, and may
be cited as it appears here.For more information, please contact the WRAP Team at:
publications@warwick.ac.uk

http://wrap.warwick.ac.uk/
 week ending
PRL 112, 215002 (2014) PHYSICAL REVIEW LETTERS 30 MAY 2014

Magnetic Reconnection and Intermittent Turbulence in the Solar Wind

K. T. Osman,1,* W. H. Matthaeus,2 J. T. Gosling,3 A. Greco,4 S. Servidio,4 B. Hnat,1 S. C. Chapman,1,5,6 and T. D. Phan7
1
Centre for Fusion, Space and Astrophysics; University of Warwick, Coventry CV4 7AL, United Kingdom
2
Bartol Research Institute, Department of Physics and Astronomy, University of Delaware, Delaware 19716, USA
3
Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado 80303, USA
4
Dipartimento di Fisica, Università della Calabria, 87036 Rende (CS), Italy
5
Department of Mathematics and Statistics, University of Tromsø, N-9037 Tromsø, Norway
6
Max Planck Institute for the Physics of Complex Systems, 01187 Dresden, Germany
7
Space Sciences Laboratory, University of California, Berkeley, California 94720, USA
(Received 18 March 2014; published 29 May 2014)
A statistical relationship between magnetic reconnection, current sheets, and intermittent turbulence in
the solar wind is reported for the first time using in situ measurements from the Wind spacecraft at 1 AU.
We identify intermittency as non-Gaussian fluctuations in increments of the magnetic field vector B that are
spatially and temporally nonuniform. The reconnection events and current sheets are found to be
concentrated in intervals of intermittent turbulence, identified using the partial variance of increments
method: within the most non-Gaussian 1% of fluctuations in B, we find 87%–92% of reconnection
exhausts and ∼9% of current sheets. Also, the likelihood that an identified current sheet will also
correspond to a reconnection exhaust increases dramatically as the least intermittent fluctuations are
removed from the data set. Hence, the turbulent solar wind contains a hierarchy of intermittent magnetic
field structures that are increasingly linked to current sheets, which in turn are progressively more likely to
correspond to sites of magnetic reconnection. These results could have far reaching implications for
laboratory and astrophysical plasmas where turbulence and magnetic reconnection are ubiquitous.

DOI: 10.1103/PhysRevLett.112.215002 PACS numbers: 94.05.Lk, 52.35.Vd, 96.50.Bh, 96.50.Ci

Introduction.—Turbulence is ubiquitous in space plasmas enhanced turbulent dissipation in kinetic collisionless

and leads to the emergence of coherent structures (see Ref. [1] plasma simulations [21,22] and nonuniform heating in
for a review). These are organized and concentrated structures the solar wind [23–25]. These results are consistent with
such as current and vorticity sheets that are phase correlated the Kolmogorov refined similarity hypothesis [26] in neutral
over their spatial extent, and are characterized by relatively fluid turbulence theory, which relates local concentrations of
long lifetimes. The solar wind is an ideal laboratory for the the dissipation rate to large intermittent fluctuations. There is
in situ study of coherent structures, which have traditionally evidence to suggest solar wind proton temperature anisot-
been described as ideal magnetohydrodynamic (MHD) dis- ropies are linked to coherent structures, which have been
continuities [2]. However, these structures also display sig- preferentially found in regions unstable with respect to
natures of intermittency in the form of rare large amplitude plasma microinstabilities [27,28]. Recent work has also
magnetic and velocity field fluctuations that produce highly indicated these structures may contribute to the acceleration
non-Gaussian heavy tailed probability distribution functions and transport of suprathermal particles [29]. There is some
[3], and have properties consistent with dynamical generation observational support for the presence of coherent structures
by strong plasma turbulence [4–6]. Indeed, turbulence gen- at subproton kinetic scales [30,31]. These can cause devia-
erates coherent structures in hydrodynamics [7], MHD [8–10] tions from local thermal equilibrium in velocity distribution
and at kinetic scales in collisionless plasmas [11]. Therefore, functions [32–35]. Therefore, relationships must exist
coherent structures embedded in the solar wind should reflect between coherent structures, intermittent turbulence, plasma
thenonlinear dynamics that give rise to intermittency [12] such heating and broader kinetic activity.
as random magnetic reconnection between adjoining flux Magnetic reconnection is also an important element in a
tubes [13–15]. Here we ask whether some coherent structures broad range of space and laboratory plasmas [36]. These
in the solar wind might participate in magnetic reconnection, a plasmas are often turbulent [37], but the influence of
fundamental process that converts magnetic energy into heat turbulence on the reconnection process is not fully under-
and plasma kinetic energy. This question, originally raised stood. The link between intermittent plasma turbulence and
over 40 years ago [16], is the subject of this Letter. magnetic reconnection is well supported by numerical
There has been renewed interest in the role of coherent simulations [38]. However, similar statistical connections
magnetic structures in the solar wind based on the develop- have not yet been found in the solar wind. This is due, at
ment and application of several identification techniques least in part, to a lack of sufficient numbers of identified
[4,17–20]. These intermittent structures are associated with reconnection events. While magnetic reconnection in the

0031-9007=14=112(21)=215002(5) 215002-1 © 2014 American Physical Society

 week ending
PRL 112, 215002 (2014) PHYSICAL REVIEW LETTERS 30 MAY 2014

solar wind was originally thought to be relatively rare where ΔBðt; τÞ ¼ Bðt þ τÞ − BðtÞ is the magnetic field
[39–42], higher resolution measurements coupled with vector increment using a time lag τ and h


i denotes an
refined techniques have recently led to an increase in the appropriate time average. All the PVI results presented here
number of identified cases [43–45]. This has now reached a use a 30 min interval of averaging, which corresponds
point where a statistically meaningful study of the type roughly to the time scale τc reported in Table I, where solar
presented here can be conducted. This Letter presents novel wind turbulent fluctuations become uncorrelated [48]. In
observational results linking magnetic reconnection to non- order to match the data resolution used when identifying
Gaussian features in the solar wind that are associated with reconnection exhausts and current sheets, a time lag τ ¼ 3 s
intermittent turbulence. is selected. This corresponds to a plasma frame spatial
Analysis.—We use 3 s resolution magnetic field measure- separation when using Taylor’s hypothesis (r ¼ −Vτ) [49]
ments from the Magnetic Field Investigation (MFI) [46] and that is within the inertial range, slightly larger than the typical
proton moments from the 3D Plasma Analyzer (3DP) [47] transition to kinetic scales [50]. The PVI time series is
onboard the Wind spacecraft. The data intervals used in this constructed in a manner that is connected with familiar
investigation are listed in Table I, and were originally selected diagnostics of intermittency and has had previous success
randomly and then carefully scrutinized for reconnection identifying reconnection sites in direct numerical MHD [51]
exhausts. In the solar wind reconnection exhausts are and Hall MHD [52] turbulence simulations. However, the
identified as roughly Alfvénic-jetting plasma (based on focus here will be on the identification of reconnection
the antiparallel field components) that are bounded on one exhausts rather than the reconnection sites themselves.
side by correlated changes in the antiparallel components of Events are selected by imposing thresholds on the magnetic
V and B and by anticorrelated changes in those components PVI time series, which leads to a hierarchy of coherent
on the other side. The list of identified reconnection exhausts structure intensities. This threshold level λ is applied by
that we use is assembled by application of these methods. In finding the PVI value above which λ% of the observations is
addition, current sheets are identified in a separate list as a contained and then removing all the lower PVI data. Thus,
reversal in at least one geocentric solar ecliptic (GSE) the smallest values of λ correspond to the largest values of
component of the magnetic field vector. While this method PVI, and the most intermittent structures. The transition
will miss current sheets where no field component actually from Gaussian to non-Gaussian magnetic field fluctuations
reverses sign, it will not significantly affect our results since occurs around PVI ¼ 3 [23,24], and thus higher values of
these are likely to be associated only with small fluctuations PVI (λ ≤ 1) largely correspond to the non-Gaussian heavy
and not the most intermittent structures of interest. The tails of the magnetic field increment distribution.
numbers of reconnection exhausts (RE) and current sheets Results.—Figure 1 shows selected plasma and magnetic
(CS) identified in this way are listed in Table I. field data for a 5 min interval encompassing a reconnection
Here we investigate whether magnetic reconnection and event. This exhaust was swept past the spacecraft in 9 s, and
current sheets are related to the intermittent character of the had a maximum local width of 3.3 × 103 km. It was also
turbulent solar wind. Note that not all reconnecting current
sheets are thought to be associated with turbulence,
including reconnection in the heliospheric current sheet
and at the leading or trailing edges of interplanetary coronal
mass ejections. However, some fraction of reconnection
exhausts in the ambient solar wind might be linked to
intermittent plasma turbulence. To this end, the partial
variance of increments (PVI) method is used to find
coherent non-Gaussian magnetic field structures:

jΔBj
PVI ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; (1)
hjΔBj2 i

TABLE I. All the data analyzed in this study, including the

numbers of reconnection exhausts (RE) and current sheets (CS).
The mean correlation time hτc i was computed in subintervals of FIG. 1 (color online). An interval encompassing a reconnection
6 h duration, and then averaged for each interval. exhaust. From top to bottom, the parameters plotted are PVI, the
GSE components of the magnetic field and solar wind velocity, the
Interval Duration (yyyy mm dd) RE CS hτc i (min) proton number density and the proton temperature. In order to
resolve the bifurcated current sheet, 92 ms resolution magnetic field
1 2001, 01 01–02 03 UT 138 28438 30
data are used. All other plotted data have a resolution of 3 s. The x
2 2006, 03 01–04 01 UT 125 29062 26
component of the flow velocity has been shifted up by 400 km s−1
3 2007, 04 01–05 01 UT 105 28131 25
and the y component of the magnetic field has been shifted down by
4 2007, 06 01–07 01 UT 153 26639 27
5 nT. Vertical lines bracket the reconnection exhaust.

215002-2
 week ending
PRL 112, 215002 (2014) PHYSICAL REVIEW LETTERS 30 MAY 2014

TABLE II. Percentage of magnetic reconnection events (ERE ) reconnection exhausts in the solar wind persists for most
and current sheets (ECS ) identified at each PVI threshold (λ) for events in our statistical ensemble.
the intervals listed in Table 1. The average PVI value ζ that Table II lists the percentage of reconnection exhausts and
corresponds to the threshold is also included (PVI ≥ ζ). current sheets identified in the solar wind for a selection of
PVI thresholds. The application of these thresholds results
Int. 1 (%) Int. 2 (%) Int. 3 (%) Int. 4 (%)
in the exclusion of all but λ% of the original data set,
λ (%) ζ ERE ECS ERE ECS ERE ECS ERE ECS leaving only the highest PVI values remaining. As the
100 0 100 100 100 100 100 100 100 100 threshold is lowered, exhausts and current sheets associated
50 0.5 100 86.0 100 86.2 100 86.4 100 86.0 with smaller PVI values are systematically removed.
10 1.4 100 40.7 100 41.4 100 41.5 100 40.6 However, this response is not linear and some identified
5 2.0 99.3 27.0 100 27.8 100 27.7 98.0 26.2 events remain even at the lowest thresholds. For example,
1 3.5 87.7 9.4 90.4 9.6 92.4 9.7 86.9 9.0 87%–92% of all reconnection exhausts and about 9% of
0.5 4.4 72.5 5.7 76.0 5.9 81.9 5.9 79.1 5.5 all current sheets are concentrated within the highest 1%
0.1 6.6 29.0 1.6 36.8 1.6 41.9 1.7 44.4 1.5
of PVI values. Therefore, the effective concentration of
0.05 7.6 22.5 0.9 26.4 0.9 25.7 0.9 32.7 0.8
0.01 10.2 9.4 0.2 11.2 0.2 13.3 0.2 13.1 0.2
current sheets and reconnection exhausts is significantly
0.005 11.2 6.5 0.1 7.2 0.1 6.7 0.1 7.8 0.1 increased by using PVI thresholds as data acceptance
0.001 14.4 2.2 0.02 2.4 0.03 2.9 0.02 3.3 0.02 criteria. This trend exists at all PVI thresholds and for all
0.0005 15.8 1.4 0.007 0.8 0.01 1.9 0.01 2.0 0.01 four intervals used in this study.
0.0001 18.1 0.7 0.004 0.8 0.003 1.0 0.004 0.7 0.004 Figure 2 shows the percentage of reconnection exhausts and
current sheets at each PVI threshold normalized to the
percentage of data occupied by these events. This concentration
associated with a magnetic field rotation of 53° that had a
of reconnection events CRE ¼ ERE =λ and current sheets
double-step character, which is characteristic of reconnec-
CCS ¼ ECS =λ should remain around unity for every threshold
tion exhausts in the solar wind. The changes in V and B are
if there is no PVI dependence. However, it is clear that both
anticorrelated on the leading edge of the exhaust and
reconnection events and current sheets are concentrated
correlated on the trailing edge. These coupled changes
preferentially at the largest PVI values, which are related to
correspond to Alfvénic disturbances that respectively
the intermittent properties of turbulence. Note that reconnec-
propagated parallel and antiparallel to B on reconnected
tion exhausts are significantly more concentrated than current
field lines and bifurcated the original current sheet. The
sheets at all but the largest λ thresholds. While the dependence
proton density and temperature were slightly enhanced
of current sheet concentration on PVI is to be expected, the
within the exhaust and V x was slightly elevated, indicating
more intermittent character of the exhausts is a novel result.
the reconnection exhaust was directed sunward. This
The diverging behavior of magnetic reconnection and
quasistationary magnetic reconnection event coincides with
current sheet concentrations in Fig. 2 suggests that the
a significant increase in PVI, which shows a clear signature
relationship between these events may depend upon PVI
within the exhaust. Note that we cannot describe this
threshold. Figure 3 shows the percentage of current sheets
interval as typical since there is considerable variation
that correspond to reconnection exhausts at each threshold,
across all 521 reconnection events used in this study.
ðRE=CSÞ × ðERE =ECS Þ × 100. For the highest thresholds
However, the connection between elevated PVI and

FIG. 2 (color online). The concentration of magnetic recon- FIG. 3 (color online). The percentage of current sheets that
nection exhausts CRE ¼ ERE =λ and current sheets CCS ¼ ECS =λ correspond to magnetic reconnection exhausts ½ðRE=CS ×
at each PVI threshold λ. ðERE =ECS Þ × 100Š at each PVI threshold λ.

215002-3
 week ending
PRL 112, 215002 (2014) PHYSICAL REVIEW LETTERS 30 MAY 2014

(lowest PVI), the number of nonreconnecting current sheets applications. This includes, but is not limited to, burst-mode
far exceeds the number of exhausts. As the threshold triggers on spacecraft instruments.
decreases, many more current sheets are removed in Further work is required to determine whether the
comparison to reconnection events. The percentage of relationship between PVI and magnetic reconnection
reconnecting current sheets increases from around 0.5% exhausts in the solar wind is universal. The overall physical
in the entire original data set, eventually reaching 100% for nature and occurrence rate of reconnecting current sheets
the highest PVI values. Therefore, this suggests the turbu- depends on solar wind speed; high speed streams contain
lent solar wind has a hierarchy of intermittent structures that current sheets that are more Alfvénic and fewer in number
are increasingly linked to current sheets, which in turn are than those found in low speed streams [40,41]. Note that the
more likely to correspond to sites of magnetic reconnection. PVI method used here does not explicitly distinguish
Discussion.—We have presented the first direct evidence between fast and slow speed solar wind. It also ignores
that reconnection exhausts in the solar wind are statistically the effects of shear angle and proton plasma beta, combi-
associated with non-Gaussian, large amplitude magnetic nations of which can suppress magnetic reconnection even at
field fluctuations, which are thought to be connected to thin current sheets [56–58]. However, solar wind speed and
the intermittent character of MHD turbulence. This result plasma beta can also affect the nature of MHD turbulence
provides further insight into the relationship between these [59], and thus the associated PVI values could be modified
fundamental processes. However, the exact nature of this in different parameter regimes. Hence, similar studies will be
link between magnetic reconnection and plasma turbulence conducted in different solar wind streams and plasma beta
is unclear. It is known that MHD turbulence dynamically regimes with the aim of reproducing the present results.
generates current sheetlike coherent structures that are Laboratory and space plasma observations suggest fast
candidate sites for active reconnection [8,13,14]. Thus, reconnection onset occurs when the current layer approaches
the reconnection exhausts seen in the solar wind could ion Larmor scales [60,61]. Furthermore, magnetic islands are
result from reconnection sites that are dynamically generated important in the development of reconnection processes in
by plasma turbulence. Alternatively, intermittent turbulence tearing theory [62] and turbulent reconnection [63], and via
could be driven by reconnection exhausts [53]. While there plasmoid instability [64]. However, in the solar wind only one
is not yet any direct observational evidence for turbulence case of multiple island reconnection has been reported [65]
generated by reconnection exhausts in the solar wind, MHD and magnetic reconnection often does not lead directly to
simulations [13] have suggested that there is a complicated dissipation or plasma heating, particularly for asymmetric
feedback mechanism underlying the interaction between boundary conditions and modest shear angles [45]. This could
plasma turbulence and magnetic reconnection. Indeed, be because the highest resolution combined plasma and
turbulence has been observed within a magnetic reconnec- magnetic field in situ data currently available to identify
tion ion diffusion region in the magnetotail [54], and kinetic reconnection events in the solar wind are generally
simulations suggest that fluctuations generated by magnetic insufficient to access the relevant kinetic scales. The
reconnection can exhibit the hallmarks of intermittent upcoming NASA Deep Space Climate Observatory and
turbulence [55]. These results could have far reaching Magnetospheric Multiscale missions will make high-resolu-
implications in laboratory and astrophysical plasmas where tion plasma and magnetic field measurements, and thus
turbulence and magnetic reconnection are ubiquitous. should identify more reconnection events across a wide range
The present study has demonstrated that application of of spatial scales, including important processes such as
increasing PVI thresholds on solar wind data acts to increase, magnetic island formation. This will further improve our
understanding of magnetic reconnection and turbulence at
within the selected population, the concentration of current
energy dissipation scales, and could have far reaching
sheets and reconnection exhausts. It also increases the
implications for turbulent dissipation in collisionless plasmas.
probability that an identified current sheet will correspond
to a reconnection event. Therefore, the PVI method could This research is supported by the UK STFC, the EU
form the basis of an automated reconnection identification Turboplasmas project (Marie Curie FP7 PIRSES-2010-
tool, and has previously been successful in numerical 269297), the NASA Magnetospheric Multiscale Mission
simulations [51]. At the highest thresholds the PVI statistic Theory and Modeling program (NNX08AT76G) and Solar
identified all reconnection exhausts, but these are greatly Probe Plus ISIS project, NASA Grants No. NNX08AO84G,
outnumbered by other nonreconnection events (false pos- No. NNX10AF26G, and No. NNX08AO83G, and the
itives). As the threshold is lowered a greater percentage of the NSF SHINE (AGS-1156094) and Solar Terrestrial (AGS-
remaining events correspond to reconnection exhausts, but 1063439) programs, and in Italy by POR Calabria FSE 2007/
increasing numbers of exhausts are not identified. From this 2013. K. T. O. and S. S. acknowledge ISSI Team 292.
perspective the PVI method cannot supplant more detailed
exhaust identification methodologies. However, its simplicity
and exclusive reliance upon magnetic field measurements, *
k.t.osman@warwick.ac.uk
which are typically available at a higher resolution than [1] R. Bruno and V. Carbone, Living Rev. Solar Phys. 10, 2
plasma measurements, make it attractive in some practical (2013).

215002-4
 week ending
PRL 112, 215002 (2014) PHYSICAL REVIEW LETTERS 30 MAY 2014

[2] L. F. Burlaga, J. Geophys. Res. 98, 17467 (1993). [34] D. Perrone, F. Valentini, S. Servidio, S. Dalena, and P. Veltri,
[3] L. Sorriso-Valvo, V. Carbone, and P. Veltri, Geophys. Res. Astrophys. J. 762, 99 (2013).
Lett. 26, 1801 (1999). [35] C. T. Haynes, D. Burgess, and E. Camporeale, Astrophys. J.
[4] A. Greco, P. Chuychai, W. H. Matthaeus, S. Servidio, 783, 38 (2014).
and P. Dmitruk, Geophys. Res. Lett. 35, L19111 (2008). [36] M. Yamada, R. Kulsrud, and H. Ji, Rev. Mod. Phys. 82, 603
[5] A. Greco, W. H. Matthaeus, S. Servidio, P. Chuychai, (2010).
and P. Dmitruk, Astrophys. J. 691, L111 (2009). [37] D. Sundkvist, A. Retinó, A. Vaivads, and S. D. Bale, Phys.
[6] M. J. Owens, R. T. Wicks, and T. S. Horbury, Sol. Phys. Rev. Lett. 99, 025004 (2007).
269, 411 (2011). [38] W. H. Matthaeus and M. Velli, Space Sci. Rev. 160, 145
[7] F. Anselmet, Y. Gagne, E. J. Hopfinger, and R. A. Antonia, (2011).
J. Fluid Mech. 140, 63 (1984). [39] J. T. Gosling, R. M. Skoug, D. J. McComas, and
[8] W. H. Matthaeus and D. Montgomery, Ann. N.Y. Acad. Sci. C. W. Smith, J. Geophys. Res. 110, A01107 (2005).
357, 203 (1980). [40] J. T. Gosling, S. Eriksson, R. M. Skoug, D. J. McComas,
[9] V. Carbone, P. Veltri, and A. Mangeney, Phys. Fluids A 2, and R. J. Forsyth, Astrophys. J. 644, 613 (2006).
1487 (1990). [41] J. T. Gosling, S. Eriksson, and R. Schwenn, J. Geophys.
[10] D. Biskamp and W. C. Muller, Phys. Plasmas 7, 4889 (2000). Res. 111, A10102 (2006).
[11] H. Karimabadi et al., Phys. Plasmas 20, 012303 (2013). [42] J. T. Gosling, Astrophys. J. 671, L73 (2007).
[12] P. Veltri, Plasma Phys. Controlled Fusion 41, A787 (1999). [43] J. T. Gosling, T. D. Phan, R. P. Lin, and A. Szabo, Geophys.
[13] W. H. Matthaeus and S. L. Lamkin, Phys. Fluids 29, 2513 Res. Lett. 34, L15110 (2007).
(1986). [44] J. T. Gosling and A. Szabo, J. Geophys. Res. 113, A10103
[14] S. Servidio, W. H. Matthaeus, M. A. Shay, P. Dmitruk, P. A. (2008).
Cassak, and M. Wan, Phys. Plasmas 17, 032315 (2010). [45] J. T. Gosling, Space Sci. Rev. 172, 187 (2012).
[15] N. Loureiro, D. A. Uzdensky, A. A. Schekochihin, S. C. [46] R. P. Lepping et al., Space Sci. Rev. 71, 207 (1995).
Cowley, and T. A. Yousef, Mon. Not. R. Astron. Soc. 399, [47] R. P. Lin et al., Space Sci. Rev. 71, 125 (1995).
L146 (2009). [48] W. H. Matthaeus, S. Dasso, J. M. Weygand, L. J. Milano,
[16] L. F. Burlaga and N. F. Ness, Can. J. Phys. 46, S962 (1968). C. W. Smith, and M. G. Kivelson, Phys. Rev. Lett. 95,
[17] R. Bruno, V. Carbone, P. Veltri, E. Pietropaolo, 231101 (2005).
and B. Bavassano, Planet. Space Sci. 49, 1201 (2001). [49] G. I. Taylor, Proc. R. Soc. A 164, 476 (1938).
[18] T. Hada, D. Koga, and E. Yamamoto, Space Sci. Rev. 107, [50] R. J. Leamon, C. W. Smith, N. F. Ness, W. H. Matthaeus,
463 (2003). and H. K. Wong, J. Geophys. Res. 103, 4775 (1998).
[19] B. J. Vasquez, V. I. Abramenko, D. K. Haggerty, and C. W. [51] S. Servidio, A. Greco, W. H. Matthaeus, K. T. Osman, and
Smith, J. Geophys. Res. 112, A11102 (2007). P. Dmitruk, J. Geophys. Res. 116, A09102 (2011).
[20] G. Li, B. Miao, Q. Hu, and G. Qin, Phys. Rev. Lett. 106, [52] S. Donato, A. Greco, W. H. Matthaeus, S. Servidio, and
125001 (2011). P. Dmitruk, J. Geophys. Res. 118, 4033 (2013).
[21] M. Wan, W. H. Matthaeus, H. Karimabadi, V. Roytershteyn, [53] G. Lapenta, Phys. Rev. Lett. 100, 235001 (2008).
M. Shay, P. Wu, W. Daughton, B. Loring, and S. C. [54] J. P. Eastwood, T. D. Phan, S. D. Bale, and A. Tjulin, Phys.
Chapman, Phys. Rev. Lett. 109, 195001 (2012). Rev. Lett. 102, 035001 (2009).
[22] J. M. TenBarge and G. G. Howes, Astrophys. J. 771, L27 (2013). [55] E. Leonardis, S. C. Chapman, W. Daughton, V.
[23] K. T. Osman, W. H. Matthaeus, A. Greco, and S. Servidio, Roytershteyn, and H. Karimabadi, Phys. Rev. Lett. 110,
Astrophys. J. 727, L11 (2011). 205002 (2013).
[24] K. T. Osman, W. H. Matthaeus, M. Wan, and [56] M. Swisdak, M. Opher, J. F. Drake, and F. Alouani Bibi,
A. F. Rappazzo, Phys. Rev. Lett. 108, 261102 (2012). Astrophys. J. 710, 1769 (2010).
[25] X. Wang, C. Tu, J. He, E. Marsch, and L. Wang, Astrophys. [57] T. D. Phan, J. T. Gosling, G. Paschmann, C. Pasma, J. F.
J. 772, L14 (2013). Drake, M. Øieroset, D. Larson, R. P. Lin, and M. S. Davis,
[26] A. N. Kolmogorov, J. Fluid Mech. 13, 82 (1962). Astrophys. J. 719, L199 (2010).
[27] K. T. Osman, W. H. Matthaeus, B. Hnat, and [58] J. T. Gosling and T. D. Phan, Astrophys. J. 763, L39 (2013).
S. C. Chapman, Phys. Rev. Lett. 108, 261103 (2012). [59] A. Petrosyan, A. Balogh, M. L. Goldstein, J. Léorat,
[28] S. Servidio, K. T. Osman, F. Valentini, D. Perrone, E. Marsch, K. Petrovay, B. Roberts, R. Steiger, and
F. Califano, S. Chapman, W. H. Matthaeus, and P. Veltri, J. C. Vial, Space Sci. Rev. 156, 135 (2010).
Astrophys. J. 781, L27 (2014). [60] A. Viavads, Y. Khotyaintsev, M. André, A. Retinò,
[29] J. A. Tessein, W. H. Matthaeus, M. Wan, K. T. Osman, D. S. C. Buchert, B. N. Rogers, P. Décréau, G. Paschmann,
Ruffolo, and J. Giacalone, Astrophys. J. 776, L8 (2013). and T. Phan, Phys. Rev. Lett. 93, 105001 (2004).
[30] S. Perri, M. L. Goldstein, J. C. Dorelli, and F. Sahraoui, [61] J. Egedal, W. Fox, N. Katz, M. Porkolab, K. Reim,
Phys. Rev. Lett. 109, 191101 (2012). and E. Zhang, Phys. Rev. Lett. 98, 015003 (2007).
[31] P. Wu, S. Perri, K. Osman, M. Wan, W. H. Matthaeus, M. A. [62] B. Coppi, G. Laval, and R. Pellat, Phys. Rev. Lett. 16, 1207
Shay, M. L. Goldstein, H. Karimabadi, and S. Chapman, (1966).
Astrophys. J. 763, L30 (2013). [63] M. Wan, W. H. Matthaeus, S. Servidio, and S. Oughton,
[32] A. Greco, F. Valentini, S. Servidio, and W. H. Matthaeus, Phys. Plasmas 20, 042307 (2013).
Phys. Rev. E 86, 066405 (2012). [64] F. Pucci and M. Velli, Astrophys. J. 780, L19 (2014).
[33] S. Servidio, F. Valentini, F. Califano, and P. Veltri, Phys. [65] S.Eriksson, D. L. Newman, G. Lapenta, and V. Angelopoulos,
Rev. Lett. 108, 045001 (2012). Plasma Phys. Controlled Fusion 56, 064008 (2014).

215002-5