Draft version November 7, 2018Preprint typeset using LATEX style emulateapj v. 5/2/11

NATURE OF THE MHD AND KINETIC SCALE TURBULENCE IN THE MAGNETOSHEATH OF SATURN:CASSINI OBSERVATIONS

L. Hadid1, F. Sahraoui1, K. H. Kiyani1,2, A. Retino1, R. Modolo3, P. Canu1, A. Masters4, and M. K. Dougherty5

1 Laboratoire de Physique des Plasmas, CNRS-Ecole Polytechnique-UPMC, Observatoire de Saint-Maur, France2 Centre for Fusion, Space and Astrophysics; University of Warwick, Coventry, CV4 7AL, United Kingdom

3 LATMOS, CNRS-UVSQ-UPMCS, Guyancourt, France4 ISAS-JAXA, Sagamihara, Japan and

5 The Blackett Laboratory, Imperial College London, SW7 2BZ, U.K.

Draft version November 7, 2018

Abstract

Low frequency turbulence in Saturn’s magnetosheath is investigated using in-situ measurementsof the Cassini spacecraft. Focus is put on the magnetic energy spectra computed in the frequencyrange ∼ [10−4, 1]Hz. A set of 42 time intervals in the magnetosheath were analyzed and three mainresults that contrast with known features of solar wind turbulence are reported: 1) The magneticenergy spectra showed a ∼ f−1 scaling at MHD scales followed by an ∼ f−2.6 scaling at the sub-ionscales without forming the so-called inertial range; 2) The magnetic compressibility and the cross-correlation between the parallel component of the magnetic field and density fluctuations C(δn, δB||)indicates the dominance of the compressible magnetosonic slow-like modes at MHD scales rather thanthe Alfven mode; 3) Higher order statistics revealed a monofractal (resp. multifractal) behaviour ofthe turbulent flow downstream of a quasi-perpendicular (resp. quasi-parallel) shock at the sub-ionscales. Implications of these results on theoretical modeling of space plasma turbulence are discussed.Subject headings: plasmas — magnetic fields — Saturn’s Magnetosheath — turbulence — waves

1. INTRODUCTION

The solar wind is unmatched by any other astrophys-ical system in the level of details in which turbulencecan be investigated. This is due the availability of manyspacecraft missions that provide high quality field andparticle in-situ measurements. The available data al-lowed significant progress in understanding turbulenceand energy dissipation in collisionless magnetized plas-mas. One of the most common and insightful ways ofmeasuring the multiscale nature of turbulence is via thePower Spectral Density (PSD) of the turbulent fluctua-tions. From that point of view, it has been shown thatthe magnetic energy spectrum in the solar wind is gen-erally characterized by at least four different dynamicalranges of scales. First is the energy-containing range thatfollows a scaling ∼ f−1, which is observed essentially inthe fast solar wind and thought to be filled by uncor-related random-like fluctuations that may originate fromreflected waves in the solar corona (Bavassano et al. 1982;Velli et al. 1989). The second region is the so-called in-ertial range with a scaling ∼ f−5/3 or ∼ f−3/2 thoughtto originate from nonlinear interactions between counter-propagating incompressible Alfvenic wave-packets trans-ferring the energy down to shorter wavelength (Irosh-nikov 1964; Kraichnan 1965). This spectrum terminateswith a breakpoint occuring near the ion gyro-scale or in-ertial length scale which is generally followed by a steeperpower-law spectrum f−α at the sub-ion scales with abroader range of slopes, α ∈ [−2.3,−3.1], where themagnetic energy starts to dissipate into particle heat-ing (Goldstein et al. 1994; Leamon et al. 1998; Alexan-drova et al. 2008; Hamilton et al. 2008; Sahraoui et al.2009; Kiyani et al. 2009a). As the cascade approaches

lina.hadid@lpp.polytechnique.fr

the electron scale, the spectrum steepens again, which isinterpreted as due to dissipation of the remaining mag-netic energy into electron heating via Landau damping ofKinetic Alfven Wave (KAW) turbulence (Leamon et al.1998, 1999; Hollweg 1999; Howes et al. 2006; Sahraouiet al. 2009). Due to instrumental limitations, the ac-tual scaling at sub-electron scales and the fate of theenergy cascade remain open questions (see discussionsin Sahraoui et al. (2013)).

Turbulence in the terrestrial magnetosheath is morecomplex than in the solar wind as different wavesand instabilities can be generated by, e.g., temperatureanisotropy generally observed behind the bow shock.Moreover, boundaries such as the magnetopause and theshock may influence some of the turbulence properties(e.g., its spatial anisotropy) (Russell et al. 1990; Cat-taneo et al. 2000; Sahraoui et al. 2006; Yordanova et al.2008). Previous studies of magnetic energy spectra in theterrestrial magnetosheath showed some similarities withthe solar wind: the presence of the Kolmogorov spec-tral index −5/3 at MHD scales (Sundkvist et al. 2007;Alexandrova et al. 2008) and a broad range of slopes,[−2.5,−3], at sub-ion scales (Czaykowska et al. 2001;Huang et al. 2014). Some differences seem to exist re-garding the scaling at sub-electron scales (Huang et al.2014).

In planetary systems other than Earth, turbulence ismuch less explored. For turbulence studies, there is atleast one major interest in investigating planetary mag-netospheres: they offer access to a broader range of plas-mas parameters that are not available in the near-Earthspace (vonPapen et al. 2014; Tao et al. 2015). This isthe case, for example, of the Alfven Mach number, rel-evant for the physics of shocks and compressible turbu-lence, which can reach values as high as ∼ 100 near Sat-

arX

iv:1

611.

0824

5v1

[ph

ysic

s.sp

ace-

ph]

24

Nov

201

6

2

urn (Masters et al. 2013). The reason is that the mag-netic field magnitude and the density fluctuations de-crease with different scaling laws whereas the solar windspeed stays relatively constant (Masters et al. 2011). An-other interest is to understand the role that the planet’ssatellites (e.g., Io for Jupiter) may play in modifying lo-cally the turbulence properties through different plasmaprocesses and instabilities that the planet-moon couplingmay generate (Chust et al. 2005; Kivelson et al. 2004;Saur et al. 2004; Bagenal 2007). Using a list of long(several hours) crossings of the Kronian magnetosheathby the Cassini spacecraft, we investigate the propertiesof turbulence at MHD and sub-ion (kinetic) scales andcompare them to the previously reported ones in the so-lar wind and in the terrestrial magnetosheath. We try toanswer three main questions regarding turbulence in themagnetosheath: is the f−5/3 Kolmogorov inertial rangeubiquitous? What is the nature of the plasma mode(s)(e.g., Alfvenic or magnetosonic) that dominate the cas-cade at different scales; iii) Do turbulence properties de-pend on the local plasma parameters (e.g., the normalangle to the shock)?

2. RESULTS

2.1. Statistics of the spectral slopes

Figure 1 illustrates a typical magnetosheath crossingon day March 17, 2005 at 02:00 UT and 08:30 UT (7.30and 7.35 Local Times respectively) at a distance of ∼42Rs (1Rs = 60.268 km). From the field magnitude andthe density measurements we see that the spacecraft wasin the solar wind until about 0200 UT, when it encoun-tered the bow shock and entered into the magnetosheathwhere the field strength and the density increased signif-icantly. The magnetic field data, sampled at 32Hz, weremeasured by the Flux Gate Magnetometer (FGM) sen-sor from the Cassini MAG experiment (Dougherty et al.2004). The FGM is mounted halfway along the 11-mspacecraft boom to minimize the interference from thespacecraft-generated electromagnetic fields. The ion andelectron moments were measured by the Cassini PlasmaSpectrometer (CAPS) (Young et al. 2004). Cassini beinga non-spinning spacecraft and the CAPS sensor havinga limited Field-of -View (FoV), a careful handling of thethermal ion population is required because the ion ther-mal speed is smaller than the bulk flow speed. In factthe moments are not reliable when the bulk of the plasmaflow is not in the FoV of the ion instrument (Thomsenet al. 2010; Romanelli et al. 2014). However, since theelectrons have a thermal speed that is larger than thebulk fow speed downstream the bow shock, the previ-ous condition can be relaxed, and the electron can beassumed isotropic (at least on large time scales that weconsider in this work) (Lewis et al. 2008). This impliesthat electron moments, computed from the ELS instru-ment, would have less uncertainties than the ion ones.For that reason, we use the electron density measuredby ELS as the plasma density under the assumption ofquasi-neutrality ni ∼ ne ∼ n (Fig. 1-b). Figure 2 thePSD of the magnetic field fluctuations computed usinga windowed Fast Fourier Transform (FFT). The spec-trum shows two ranges of scales with distinct power-laws:∼ f−1.26 at low frequencies (f < 10−2Hz) and ∼ f−2.54

at higher frequencies. This observation shows a strik-

SW SWMSa

b

Figure 1. (a) The magnetic field modulus, (b) the electronplasma density measured by the Cassini spacecraft in the solar wind(SW) and in the magnetosheath (MS) of Saturn on 2005/03/17from 00:00-10:00.

Figure 2. The power spectral density of δB measured between02:00-08:30. The black lines are the power-law fits. The dottedcurve is a spectrum measured in the solar wind, considered hereto represent the upper bound of the sensitivity floor of FGM. Thearrow corresponds to the ion gyro-frequency, the gray and the redshaded bands indicate the Taylor shifted ion inertial length fdi andLarmor radius fρi , respectively (the width reflects the uncertaintydue to errors in estimating the ion moments).

ing result: turbulence transits directly from the “energycontaining scales” into the ion kinetic scales, withoutforming the so-called Kolmogorov inertial range with ascaling 5/3 (the terminology of energy containing scalesis borrowed from solar wind turbulence). The spectralbreak is closer to the local ion gyro-frequency than tothe Taylor-shifted ion inertial length fdi = Vf/2πdi andLarmor radius fρi = Vf/2πρi (Vf ∼ 300 km/s is the av-erage flow speed, Ti ∼ 258 eV, B0 ∼ 1.4 nT, ne ∼ 0.06cm−3 and βi ∼ 3.3). The reason might be that theselatter are subject to higher uncertainties due to errorsin estimating the plasma parameters using the ion mo-ments from the CAPS instrument. To confirm theabsence of the Kolmogorov f−5/3 spectrum in the mag-netosheath, we analyzed a list of 42 other time intervalsbetween 2004 and 2007, for a quasi-parallel and quasi-perpendicular shocks separately. For most of the time in-tervals we identified the structure of the shock by check-ing the angle θBn between the interplanetary magneticfield and the normal to the shock estimated using a semi-

3

Slope Values Slope Values

His

tog

ram

Den

sity

ba

Quasi-Perpendicular Shock Quasi-Parallel Shock

MHD scalesSub-ion scales

MHD scalesSub-ion scales

Figure 3. Histograms of the spectral slopes at MHD and sub-ion(kinetic) scales downstream of quasi-perpendicular (a) and quasi-parallel Shock (b).

empirical model of the global shock surface (Went et al.2011): θBn > 45◦ indicates a quasi-perpendicular shockwhereas θBn ≤ 45◦ indicates a quasi-parallel one (in fewcases, quasi-perpendicular shocks are simply identifiedby a sharp gradient in the magnetic field and the plasmameasurements). The results shown in Figure 3 confirmstatistically the absence of the Kolmogorov spectrum atMHD scales: the bulk of the spectra had slopes near −1in particular for quasi-perpendicular shocks (Czaykowskaet al. 2001). The histogram of the slopes at sub-ion scalespeaks between [−2.5,−3] in general agreement with pre-vious results reported in the solar wind and the mag-netosheath (Sahraoui et al. 2013; Huang et al. 2014). Aslight indication that steeper spectra are observed behindquasi-parallel shocks can also be seen.

2.2. Nature of the turbulent fluctuations at the MHDand kinetic scales

To identify the nature of plasma modes that carry theenergy cascade from the energy-containing to the sub-ionscales, we use the magnetic compressibility CB given bythe ratio between the PSDs of the parallel magnetic fieldcomponent and the magnetic field magnitude (parallel isw.r.t. the background magnetic field B0) (Gary & Smith2009; Salem et al. 2012):

CB(f) =|δB‖(f)|2

|δB‖(f)|2 + |δB⊥(f)|2(1)

Indeed, from linear theory, the Alfven and the magne-tosonic modes are known to have very different profilesof the magnetic compressibility (Sahraoui et al. 2012).This can allow us to verify easily the dominance (ornot) of the Alfvenic fluctuations in our data (Podesta& TenBarge 2012; Kiyani et al. 2013). We computedthe theoretical magnetic compressibilities from the lin-ear solutions of compressible Hall-MHD (Sahraoui et al.2003) and from the Maxwell-Vlasov equations using theWHAMP code (Ronnmark 1982). For the sake of sim-plicity, we keep using the terminology of the MHD slowand fast modes at kinetic scales even if it may be in-adequate (because of possible crossings between differ-ent dispersion branches). Since the slow mode is heavilydamped in kinetic theory at finite βi (Ito et al. 2004;Howes 2009), we used the limit βi = 0 and βe = 1

Fast modeSlow modeAlfven mode

Figure 4. Comparison between theoretical magnetic compress-ibilities, computed from the linear solutions of the compressibleHall-MHD (color dotted line) and of the Vlasov-Maxwell equations(colored solid line) for β = 1 and ΘkB = 87◦, with the observedone from the data of Fig. 1 (02:00-08:30) (solid black curve). TheTaylor hypothesis was used to convert the frequencies in the space-craft frame into wavenumber. The red, green and blue curves cor-respond respectively to the theoretical fast, slow and KAW modes.The horizontal dashed black line at CB = 1/3 indicates the powerisotropy level.

(therefore, β = βe + βi = 1) to suppress the ion Lan-dau damping and thus to capture the slow mode so-lution down to the scale kρi ≥ 1. In order to com-pare to spacecraft observations, the knowledge of thethree components of the k vector (or equivalently, thepropagation angle ΘkBo and the modulus k) from thedata is required. However, unambiguous determinationof those quantities requires having multi-spacecraft datathat is not available in planetary magnetospheres otherthan Earth (Sahraoui et al. 2006). Therefore, we usethe Taylor frozen-in-flow hypothesis, which assumes thatthe fluctuations have slow phase speeds w.r.t the flowspeed, to infer the component of the k along the flowdirection, i.e. ωsc ∼ k.Vf ∼ kVf . Under the assumptionthat turbulence is strongly anisotropic, i.e. k⊥ >> k‖,which is supported by previous observations in the mag-netosheath (Sahraoui et al. 2006; Mangeney et al. 2006)and in the solar wind (Sahraoui et al. 2010), the esti-mated wavenumber component along the flow is equiv-alent to k⊥ for data intervals when ΘVfB0

∼ 90◦. Inthe present data we estimated ΘVfB0

∼ 87◦, which weused to compute the theoretical solutions of Fig. 4 as-suming ΘVfB0

∼ ΘkB0. Nevertheless, we performed a

parametric study (not shown here) and verified that themagnetic compressibilities of the compressible Hall-MHDsolutions keep the same profile (but change its magni-tude) when varying β in the range [0.2, 100] for a fixedΘkBo = 87◦, and when varying ΘkBo from quasi-parallelto quasi-perpendicular angles for β = 1. Another conse-quence of using the Taylor hypothesis when ΘVfB0

∼ 90◦

is that the perpendicular component of the fluctuationδB⊥ in Eq. 2 is reduced to the component perpendicularto both Vf (or k, to fulfill k.δB = 0) and to B0 (Podesta& TenBarge 2012), namely

δB⊥(f) = δB(f).k×B0

|k×B0|∼ δB(f).

Vf ×B0

|Vf ×B0|(2)

Figure 4 shows a comparison between the observed mag-netic compressibility (from the data of Fig. 1) compared

4

δB||δne

Figure 5. (a) The plasma density and the magnetic field mag-nitude; (b) The local and averaged cross correlation of the densityand the parallel component of the magnetic fluctuations calculatedusing Pearson’s method.

to theoretical ones calculated using the observed plasmaparameters. First, one can see that the theoretical mag-netic compressibilities of the fast and slow modes in thefluid and kinetic models have the same profile being al-most constant at the MHD and sub-ions scales. TheKAW mode shows an increasing magnetic compressibil-ity as it approaches kinetic scales (Sahraoui et al. 2012).A similar profile has been reported in solar wind obser-vations (Podesta & TenBarge 2012; Kiyani et al. 2013).Second, the measured magnetic compressibility shows arelatively constant and high level (CB > 1/3) at the en-ergy containing scales and in the sub-ions range, whichindicates the dominance of the parallel component δB‖(most of the 42 studied intervals showed a similar pro-file). This clearly rules out the Alfvenic fluctuations as adominant component of the turbulence at least at MHDscales (f < 0.05Hz).

Figure 4 shows that the magnetic compressibility can-not be used to distinguish between fast and slow modes.To do so, we use instead the cross-correlation betweenthe magnetic field and the plasma density fluctuationsC(δB||, δn). Indeed, the fast (resp. slow) mode is knownto have a correlation (resp. anti-correlation) betweenits density and parallel magnetic component (Gary &Winske 1992). Figure 5 shows that locally and on averagethe density and the parallel component of the magneticfluctuations are anti-correlated, i.e. C(δB||, δne) < 0.This clearly rules out the fast mode fluctuation as thedominant component of the turbulence. This analy-sis establishes that the magnetosonic slow-like modedominates the turbulent fluctuations analyzed here inagreement with previous results on the Earth’s magne-tosheath (Kaufmann et al. 1970; Song et al. 1994; Cat-taneo et al. 2000), in outer planets (Violante et al. 1995;Erdos & Balogh 1996) and in the solar wind (Howes et al.2012; Klein et al. 2012). However, one cannot rule out thepossible presence of the ion mirror mode as previously re-ported in the terrestrial magnetosheath (Sahraoui et al.2006). The mirror mode, although is of purely kinetic na-ture (Southwood & Kivelson 1993), has indeed very sim-ilar properties than the slow mode which makes it chal-lenging to distinguish between the two modes in space-craft data. To check the possibility for the mirror modeto exist in our data requires measuring at least the ion

temperature anisotropy, which is not available onboardthe Cassini spacecraft.

2.3. Higher order statistics of the magnetic fieldfluctuations

To investigate the mono-fractal versus multi-fractal na-ture of the observed turbulence we analyzed the Proba-bility Density Function (PDF) of the magnetic field tem-poral increments, defined as δBτ (t, τ) = B(t+ τ)−B(t)where τ is the time lag. Intermittency is generally char-acterized by the presence of bursty increments whichyield heavy tails in the PDF of the field increments atsmall scales. In general, it is this deviation from Gaus-sianity that contains the pertinent information about theunderlying physics. Figure 6 provides three examples ofthe corresponding PDFs obtained from the list of the an-alyzed events downstream a quasi-parallel (Θ = 31◦) andquasi-perpendicular shocks (Θ = 86◦ and Θ = 60◦, re-spectively) for two values of τ , one from the MHD range(τ ∼ 975 s) and one from the sub-ion range (τ ∼ 25 s).Figures 6-(a-b) show that behind the quasi-perpendicularshock, the PDFs are found quasi-Gaussian for both val-ues of τ (i.e., at MHD and kinetic scales) indicating thequasi-randomness of the fluctuations in the “energy con-taining sacles” and in the sub-ion range. On the contrary,behind the quasi-parallel shock (Figure 6-c), the PDFsare non-Gaussian for τ ∼ 25 s (red PDF) showing theintermittent nature of turbulence at the kinetic scales asit was observed in the terrestrial magnetosheath (Sund-kvist et al. 2007; Yordanova et al. 2008). In the en-ergy containing scales the PDF is quasi-Gaussian as inthe case of quasi-perpendicular shock. These resultsagree with recent findings using global hybrid simula-tions (Karimabadi et al. 2014). We next calculate higherorder statistics given by the structure functions (SFs) ofthe magnetic field increments defined in Eq. 3.

Sm(τ) =

∞∫−∞

|δBτ (t)|mP (δB)dt =< |δBτ (t)|m > (3)

When increasing the order m, the SFs become progres-sively sensitive to rare and bursty events. Assuming apower-law dependence of the SFs as function of τ , i.e.,Sm(τ) ∼ τ−ζ(m), a linear (resp. nonlinear) dependenceof ζ(m) on the order m indicates a mono-fracatal (resp.a multi-fractal) behaviour of the turbulence. The scal-ing exponent in Figure 7-a shows a clear linear depen-dence of ζ on m at the sub-ion scales (≡ small values ofτ) behind a quasi-perpendicular shock supporting thusthe monofractal character of the turbulent fluctuations.However, downstream a quasi-parallel shock (Figure 7-c),the scaling exponent is a convex function of m, confirm-ing the multifractal nature of turbulence at the kineticscales. We recall that in the solar wind it has been shownthat the sub-ion scales were monofractal while the MHDscales were multi-fractal (Kiyani et al. 2009a).

3. DISCUSSION AND CONCLUSIONS

From the benefit of analyzing sufficiently long and rel-atively stationary time series measured by the Cassinispacecraft in the magnetosheath of Saturn, we were ableto probe into more than four decades of scales spanningfrom the MHD down to the sub-ion scales. We present

5

a b c

MHD Scales(τ=470 s)

Sub-ion scales(τ=12 s)

MHD Scales(τ=470 s)

Sub-ion scales(τ=12 s)

MHD Scales(τ=470 s)

Sub-ion scales(τ=12 s)

Quasi-Perpendicular Shock2007/04/28 13:30-17:00

Θ ~ 86°

Quasi-Perpendicular Shock2005/03/17 06:00-08:20

Θ ~ 60°

Quasi-Parallel Shock2005/03/17 02:05-04:15

Θ ~31°a b c

MHD Scales(τ=470 s)

Sub-ion scales(τ=12 s)

MHD(τ=470 s)

Sub-ion(τ=12 s)

MHD Scales(τ=470 s)

Sub-ion scales(τ=12 s)

Figure 6. PDFs of the magnetic field increments in the energycontaining and sub-ion scales (blue and red respectively) down-stream of quasi-perpendicular (a-b) and quasi-parallel (c) shocks(with Poisonnian error bars). Normalized histograms with 300 binseach were used to compute the PDFs. The same values of τ wereused in both cases, τ ∼ 12 s for the kinetic scales and τ ∼ 470 sfor the MHD scales. All the PDFs have been rescaled to have unitstandard deviation. A Gaussian distribution (black dashed curve)is shown for comparison.

log 10 (=[secs])1 2 3

log

10(S

m)

-2

0

2

4

6

Moment m0 2 4 6

1(m

)

0

0.5

1

1.5

2

log 10 (=[secs])1 2 3

log

10

(Sm

)

-2

0

2

4

6

8

Moment m0 2 4 6

1(m

)

0

1

2

3

Slope= 0.43033Slope=0.45

d e

log 10 (=[secs])1 2 3

log

10

(Sm

)

-2

0

2

4

6

8

Moment m0 2 4

1(m

)

0

0.5

1

1.5Scaling exponents Scaling exponents Scaling exponents

Quasi-Parallel Shock2005/03/17 02:05-04:15

Θ ~31°

Structure Functions

Quasi-Perpendicular Shock2007/04/28 13:30-17:00

Θ ~ 86°

Structure Functions

Quasi-Perpendicular Shock2005/03/17 06:00-08:20

Θ ~ 60°

Structure Functions

a b

c d

Quasi-Perpendicular Shock2070000-13:40

Structure Functions

Scaling exponents

Quasi-Parallel Shock2005/03/17 02:00-08:00

Structure Functions

a b

c d

Quasi-Perpendicular Shock2004/06/28 08:00-13:40

Structure Functions

Scaling exponents

Quasi-Parallel Shoc2005/03/17 02:00-08:

Structure Functionsa

c

Quasi-Perpendicular Sho2004/06/28 08:00-13:40

Structure Functions

Scaling exponents

f

a b c

Figure 7. Different orders of the structure functions of the mag-netic field increments δBτ (t) as function of the time lag τ down-stream of quasi-perpendicular (a-b) and quasi-parallel (c) shocks.(d), (e) and (f) represent the corresponding scaling exponent ζ(m).

the following plausible, albeit speculative, scenario to ex-plain the different observations: the interaction of thesolar wind with the bow shock may lead to the destruc-tion of all the pre-existing correlations between the tur-bulent fluctuations in the solar wind. This results insuppressing the Kolmogorov inertial range and generat-ing locally random-like fluctuations that have a scaling∼ f−1 over a broad range of scales. Those scales wouldplay the same role as the energy containing scales in so-lar wind turbulence. The absence of the inertial rangescale in our observations can be explained by the factthat the newly generated fluctuations behind the shockdo not have “enough time” to interact sufficiently witheach other to reach a fully developed turbulence state,hence the direct transition from the “energy containing”range that has ∼ f−1 scaling into the sub-ion range witha scaling ∼ f−2.6. In this scenario, turbulence may reacha fully developed state and the Kolmogorov 5/3 spectrummay be observed but only far away from the shock (e.g.,toward the flanks). However, a fundamental questionremains open: How the power-law spectra observed inthe sub-ion range are created in the absence of an iner-tial range? We note that existing theoretical models ofkinetic turbulence in the solar wind suggest that turbu-lence at sub-ion scales may result as a consequence of a

decoupling between the dynamics of Alfvenic fluctuationsand the rest of the MHD fluctuations that would carrythe cascade into the kinetic scales (Schekochihin et al.2009). Hence, if the MHD scales are not dominated byAlfvenic fluctuations as observed here, it is not clear as tohow turbulence is generated at kinetic scales. Is it gener-ated by local plasma instabilities occurring near the ionscale as proposed in Sahraoui et al. (2006)? In this case,would the f−1 spectrum observed at larger scales resultby an inverse cascade as observed in hydrodynamic tur-bulence (Paret & Tabeling 1997; Chertkov et al. 2007)?Another observation reported here that requires furtherinvestigations is the nature of the turbulence observedbehind the quasi-perpendicular shock. At sub-ion scales,turbulence was found to have a mono-fractal behaviouras in the solar wind (Kiyani et al. 2009b). However, thetails of the PDFs are clearly less pronounced in our ob-servations and the PDF of Figure 6-a looked close to aGaussian rather than a heavy tailed PDF. This resultrecalls recent numerical observations of weak turbulencein Electron-MHD (Meyrand et al. 2015). This similarityand the questions raised above require further investiga-tion in the future.

Authors thank the NASA’s Planetary Data System(PDS) for the CASSINI-CAPS data set availability,(Waite, J.H., Furman, J.D., CASSINI ORBITERSAT/SW CAPS DERIVED ION MOMENTS V1.0,CO-S/SW-CAPS-5-DDR-ION-MOMENTS-V1.0, NASAPlanetary Data System, 2013). Authors are alsoindebted to the program “Soleil Heliosphere etMagnetospheres” of CNES, the french space ad-ministration for the financial support on CASSINI. Dataanalysis were done with the AMDA science analysis sys-tem provided by the Centre de Donnees de la Physiquedes Plasmas (IRAP, Universite Paul Sabatier, Toulouse)supported by CNRS and CNES. LH thanks G. Belmontfor many fruitful discussions.

REFERENCES

Alexandrova, O., Lacombe, C., & Mangeney, A. 2008, Ann.Geophys., 26, 3585

Bagenal, F. 2007, Journal of Atmospheric and Solar-TerrestrialPhysics, 69, 387 , global Aspects of Magnetosphere-IonosphereCoupling Global Aspects of Magnetosphere-IonosphereCoupling

Bavassano, B., Dobrowolny, M., Mariani, F., & Ness, N. F. 1982,Journal of Geophysical Research: Space Physics, 87, 3617

Cattaneo, M. B. B., Moreno, G., Russo, G., & Richardson, J. D.2000, Journal of Geophysical Research: Space Physics, 105,23141

Chertkov, M., Connaughton, C., Kolokolov, I., & Lebedev, V.2007, Phys. Rev. Lett., 99, 084501

Chust, T., Roux, A., Kurth, W., Gurnett, D., Kivelson, M., &Khurana, K. 2005, Planetary and Space Science, 53, 395

Czaykowska, A., Bauer, T. M., Treumann, R. A., & Baumjohann,W. 2001, Annales Geophysicae

Dougherty, M. K., et al. 2004, Space Sci. Rev., 114, 331Erds, G., & Balogh, A. 1996, Journal of Geophysical Research:

Space Physics, 101, 1Gary, S. P., & Smith, C. W. 2009, J. Geophys. Res., 114, A12105Gary, S. P., & Winske, D. 1992, Journal of Geophysical Research:

Space Physics, 97, 3103Goldstein, M. L., Roberts, D. A., & Fitch, C. A. 1994, Journal of

Geophysical Research: Space Physics, 99, 11519Hamilton, K., Smith, C. W., Vasquez, B. J., & Leamon, R. J.

2008, J. Geophys. Res., 113, A01106Hollweg, J. V. 1999, J. Geophys.Res., 104, 14811Howes, G. G., Cowley, S. C., Dorland, W., Hammett, G. W.,

Quataert, E., & Schekochihin, A. A., 2006, The AstrophysicalJournal, 651, 590–614

6

Howes, G. G. 2009, Nonlin. Proc. Geophys., 16, 219Howes, G. G., Bale, S. D., Klein, K. G., Chen, C. H. K., Salem,

C. S., & TenBarge, J. M. 2012, The Astrophysical JournalLetters, 753, L19

Huang, S. Y., Sahraoui, F., Deng, X. H., He, J. S., Yuan, Z. G.,Zhou, M., Pang, Y., & Fu, H. S. 2014, The AstrophysicalJournal Letters, 789, L28

Iroshnikov, P. S. 1964, Soviet Astronomy, 7, 566Ito, A., Hirose, A., Mahajan, S. M., & Ohsaki, S. 2004, Phys.

Plasmas, 11, 5643Karimabadi, H., et al. 2014, Physics of Plasmas, 21Kaufmann, R. L., Horng, J.-T., & Wolfe, A. 1970, Journal of

Geophysical Research, 75, 4666Kivelson, M. G., Bagenal, F., Kurth, W. S., Neubauer, F. M.,

Paranicas, C., & Saur, J. 2004, Magnetospheric interactionswith satellites, ed. F. Bagenal, T. E. Dowling, & W. B.McKinnon, 513–536

Kiyani, K. H., Chapman, S. C., Sahraoui, F., Hnat, B.,Fauvarque, O., & Khotyaintsev, Y. V. 2013, ApJ, 763, 10

Kiyani, K. H., Chapman, S. C., & Watkins, N. W. 2009a, Phys.Rev. E, 79, 036109

Kiyani, K. H., Chapman, S. C., Yu. V. Khotyaintsev, Dunlop,M. W., & Sahraoui, F. 2009b, Phys. Rev. Lett., 103, 075006

Klein, K. G., Howes, G. G., TenBarge, J. M., Bale, S. D., Chen,C. H. K., & Salem, C. S. 2012, The Astrophysical Journal, 755,159

Kraichnan, R. 1965, Physics of Fluids, 8, 1385Leamon, R. J., Matthaeus, W. H., Smith, C. W., & Wong, H. K.

1998, ApJ Lett., 507, L181Leamon, R. J., Smith, C. W., Ness, N. F.,Matthaeus, W.H., &

Wong, H. K. 1998, J. Geophys. Res., 103, 4775Leamon, R. J., Smith, C. W., Ness, N. F.,Matthaeus, & Wong,

H. K. 1999, J. Geophys. Res., 104, 22331Lewis, G.R., Andr, N., Arridge, C.S., Coates, A.J., Gilbert, L.K.,

Linder, D.R., & Rymer, A.M. 2008, Planetary and SpaceScience, 56, 901-912

Mangeney, A., Lacombe, C.and Maksimovic, M. S.-A. A.,Cornilleau-Wehrlin, N., Harvey, C. C.and Bosqued, J.-M., &Trvnvcek, P. 2006, Annales Geophysicae, 24, 3507

Masters, A., et al. 2011, Journal of Geophysical Research: SpacePhysics, 116

—. 2013, Nature Physics, 9, 164Meyrand, R., Kiyani, K. H., & Galtier, S. 2015, Journal of Fluid

Mechanics, 770Paret, J., & Tabeling, P. 1997, Phys. Rev. Lett., 79, 4162Podesta, J. J., & TenBarge, J. M. 2012, Journal of Geophysical

Research: Space Physics, 117, n/aRomanelli, N., et al. 2014, Journal of Geophysical Research:

Space Physics, 119, 9992, 2014JA020391

Ronnmark, K. 1982, WHAMP - Waves in Homogeneous,Anisotropic, Multicomponent Plasmas, Tech. Rep. 179, KirunaGeophysical Institute

Russell, C. T., Lepping, R. P., & Smith, C. W. 1990, Journal ofGeophysical Research

Sahraoui, F., Belmont, G., & Goldstein, M. L. 2012, Astrophys.J., 748, 100

Sahraoui, F., Belmont, G., Rezeau, L., & Cornilleau-Wehrlin, N.2006, Phys. Rev. Lett., 96, 075002

Sahraoui, F., Goldstein, M. L., Belmont, G., Canu, P., & Rezeau,L. 2010, Phys. Rev. Lett., 105, 131101

Sahraoui, F., Goldstein, M. L., & Khotyaintse, Y. V. 2009, Phys.Rev. Lett., 102, 231102

Sahraoui, F., Huang, S. Y., Belmont, G., Goldstein, M. L.,Retino, A., Robert, P., & Patoul, J. D. 2013, The AstrophysicalJournal, 777, 15

Sahraoui, F., et al. 2003, J. Geophys. Res., 108, 1335Salem, C. S., Howes, G. G., Sundkvist, D., Bale, S. D., Chaston,

C. C., Chen, C. H. K., & Mozer, F. S. 2012, ApJ Lett., 745, L9Saur, J., Neubauer, F. M., Connerney, J. E. P., Zarka, P., &

Kivelson, M. G. 2004, Plasma interaction of Io with its plasmatorus, ed. F. Bagenal, T. E. Dowling, & W. B. McKinnon,537–560

Schekochihin, A. A., Cowley, S. C., Dorland, W., Hammett,G. W., Howes, G. G., Quataert, E., & Tatsuno, T. 2009, TheAstrophysical Journal Supplement Series, 182, 310

Song, P., Russell, C. T., & Gary, S. P. 1994, Journal ofGeophysical Research: Space Physics, 99, 6011

Southwood, D. J., & Kivelson, M. G. 1993, Journal ofGeophysical Research: Space Physics, 98, 9181

Sundkvist, D., Retino, A., Vaivads, A., & Bale, S. D. 2007, Phys.Rev. Lett., 99, 025004

Tao, C., Sahraoui, F., Fontaine, D., de Patoul, J., Chust, T.,Kasahara, S., & Retino, A. 2015, Journal of GeophysicalResearch: Space Physics, 120, 2477, 2014JA020749

Thomsen, M. F., Reisenfeld, D. B., Delapp, D. M., Tokar, R. L.,Young, D. T., Crary, F. J., Sittler, E. C., McGraw, M. A., &Williams, J. D. 2010, Journal of Geophysical Research: SpacePhysics, 115,2156-2202, 2010JA015267

Velli, M., Grappin, R., & Mangeney, A. 1989, Phys. Rev. Lett.,63, 1807

Violante, L., Cattaneo, M. B. B., Moreno, G., & Richardson,J. D. 1995, Journal of Geophysical Research: Space Physics,100, 12047

vonPapen, M., Saur, J., & Alexandrova, O. 2014, Journal ofGeophysical Research: Space Physics, 119, 2797

Went, D. R., Hospodarsky, G. B., Masters, A., Hansen, K. C., &Dougherty, M. K. 2011, Journal of Geophysical Research: SpacePhysics, 116, n/a, a07202

Yordanova, E., Vaivads, A., Andre, M., Buchert, S. C., & Voros,Z. 2008, Physical Review Letters, 100, 205003

Young, D. T., et al. 2004, Space Sci. Rev., 114, 1