Magnetic field structure in single late-type giants: The effectively single giant V390 Aur

0.41 MB·English

by  R. Konstantinova-Antova

Astronomy & Astrophysics manuscript no. v390aur˙zdi˙revision˙3R (cid:13)c ESO 2012
January 25, 2012
Magnetic field structure in single late-type giants: The effectively
⋆
single giant V390 Aur
R. Konstantinova-Antova1,2, M. Auri`ere2, P. Petit2, C. Charbonnel3,2, S. Tsvetkova1, A. L`ebre4, and R.
Bogdanovski1
1 Instituteof Astronomy and NAO,Bulgarian Academy of Sciences, 72 Tsarigradsko shose, 1784 Sofia, Bulgaria
e-mail: [email protected]
2 IRAP;UMR 5277; CNRSand Universit´ede Toulouse, 14, AvenueEdouard Belin, F-31400 Toulouse, France
3 GenevaObservatory,University of Geneva, 51, Chemin des Maillettes, 1290 Versoix, Switzerland
2 4 LUPM - UMR5299; CNRS and Universit´e de Montpellier 2, Place E. Bataillon, F-34095 Montpellier cedex 05,
1
France
0
2

n Received ;
a
J ABSTRACT

4
Aims.WehavestudiedtheactivegiantV390Aurusingspectropolarimetry toobtaindirectandsimultaneousmeasure-
2
mentsof themagnetic field and the activity indicators in order to get a precise insight of its activity.

Methods. We used the spectropolarimeter NARVAL at the Bernard Lyot Telescope (Observatoire du Pic du Midi,
]
R France) to obtain a series of Stokes I and Stokes V profiles. The Least Square deconvolution (LSD) technique was
applied to detect the Zeeman signature of the magnetic field in each of our 13 observations and to measure its longi-
S
. tudinal component. We could also monitor the CaIIK&H and IR triplet, as well as the Hα lines which are activity
h indicators. In order to reconstruct the magnetic field geometry of V390 Aur, we applied the Zeeman Doppler Imaging
p (ZDI) inversion method and present a map for the magnetic field. Based on the obtained spectra, we also refined the
- fundamental parameters of thestar and theLi abundance.
o Results.TheZDIrevealedastructureintheradialmagneticfieldconsistingofapolarmagneticspotofpositivepolarity
r
and several negative spots at lower latitude. A high latitude belt is present on the azimuthal field map, indicative of
t
s a toroidal field close to the surface. It was found that the photometric period cannot fit the behaviour of the activity
a
indicatorsformedinthechromosphere.Theirbehavioursuggestsslowerrotationcomparedtothephotosphere,butour
[
dataset is too short tobe able to estimate theexact periods for them.
1 All these results can be explained in terms of an α−ω dynamo operation, taking into account the stellar structure
v and rotation properties of V390 Aur that we study using up to-date stellar models computed at solar metallicity. The
2 calculated Rossby numberalso points to a very efficient dynamo.
4
Key words.star: individual:V390 Aur– magnetic field – dynamo
0
5
.
1 1. Introduction According to the literature V390 Aur has a mass of
0
about 1.8–2 M⊙ and is evolving along the base of the red
2
TheG8IIIstarV390Aur=HD33798isalreadyknownfor giantbranch(RGB)(Gondoin2003,Konstantinova-Antova
1
itsmagneticactivity.Amodulationofthephotometriclight et al. 2009). It is the primary of a wide multiple physical
:
v curveduetospotsisobservedbySpurr&Hoff(1987)anda system, ADS 3812, and it was shown that the giant could
Xi periodof9.825daysthatisconsideredtherotationalperiod be considered effectively single with respect to its activity,
of the star is determined by Hooten & Hall (1990). It has i.e. synchronization plays no role in its fast rotation and
r
a beenclassifiedasa chromosphericallyactive giantby Fekel activity (Konstantinova-Antova et al. 2008).
& Marshall(1991) on the basis of its CaIIK emission and
fast rotation (vsini = 29km.s−1). Later, enhanced X–ray
emission(Hu¨nschetal.1998,Gondoin1999,Gondoin2003)
Here we present the first map of the surface magnetic
andopticalflares(Konstantinova-Antovaetal.2000,2005)
field structure for this giant star applying the Zeeman
weredetectedforthisstar.Recently,Konstantinova-Antova
Doppler Imaging (ZDI) technique (Donati& Brown 1997,
et al. (2008)reporteddirect detection of magnetic field us-
Donati et al. 2006a). Section 2 describes our observations
ing the spectropolarimeter NARVAL at the 2m Telescope
andmethodsfordataprocessing.Section3givesourresults
Bernard Lyot at Pic du Midi Observatory,France.
from the ZDI and the behaviour of the activity indicators.
Based on up-to-date stellar evolution models we estimate
Send offprint requests to: R. Konstantinova-Antova in Section 4 the mass and evolutionarystatus of V390 Aur
⋆ Based on data obtained using the T´elescope Bernard Lyot anddetermineitsRossbynumberasanindicatorofthedy-
at Observatoire du Pic du Midi, CNRS and Universit´e Paul namo efficiency. Section 5 and 6 are for the discussion and
Sabatier, France. conclusions, respectively.
1
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
2. Observations and data processing Table 1. Data for activity indicators and B .
l
The observations of V390 Aur were performed at the Date Phase CaII K Hα CaII IR Bl σ
2-m Bernard Lyot Telescope (TBL) of Pic du Midi Gauss Gauss
Observatory with the new generation spectropolarimeter 14 Sep 08 0.00 0.56 0.339 0.441 −10 3
NARVAL (Auri`ere 2003) which is a copy of the instru- 15 Sep 08 0.10 0.55 0.341 0.435 −13 2
mentESPaDOnSinstalledattheCFHT attheendof2004 16 Sep 08 0.20 0.57 0.339 0.442 −9 3
19 Sep 08 0.51 0.57 0.383 0.463 1 3
(Donati et al. 2006b). NARVAL is a fiber–fed echelle spec-
20 Sep 08 0.60 0.61 0.391 0.477 −7 4
trometerallowingthewholespectrumfrom370nmto1000
21 Sep 08 0.705 0.58 0.387 0.476 −8 3
nm to be recorded in each exposure, in 40 orders aligned
24 Sep 08 0.01 0.53 0.387 0.458 −12 3
in the CCD frame by 2 cross-disperser prisms. We used it 25 Sep 08 0.12 0.52 0.377 0.448 −8 3
in polarimetric mode with a spectral resolution of about 26 Sep 08 0.22 0.55 0.351 0.444 −9 3
65000.StokesI (unpolarised)and StokesV(circular polar- 27 Sep 08 0.32 0.58 0.347 0.448 −6 3
ization) parameters were obtained by four sub-exposures, 28 Sep 08 0.42 0.59 0.341 0.436 −3 2
where in between each of them the retarders and Fresnel 29 Sep 08 0.53 0.61 0.350 0.451 −1 2
rhombswererotatedinordertoexchangethe beamsinthe 30 Sep 08 0.62 0.64 0.359 0.469 −13 4
instrument and to reduce the spurious polarization signa- Note: Rc values are given for Hα and CaII 854.2nm. For
tures (Semel et al. 1993). CaII K emission core I/I(395nm) is measured. Bl values for
themagnetic field and their accuracy are given in Gauss.
Thirteen observations were obtained in the period 14 – Phase is for 9.825d photometric period, considered as the
30September,2008(Table1).Theextractionofthespectra rotational period of V390 Aur.
was performed using Libre-ESpRIT (Donati et al. 1997), a
fullyautomaticreductionpackageinstalledatTBL.Forthe
Zeemananalysis,least-squaredeconvolutionanalysis(LSD,
Donatietal.1997)wasappliedtoallobservations.Weused -18
a mask calculated for an effective temperature of 5000K, -16 Bl
logg =3.0 and a microturbulence of 2.0km.s−1, consistent -14
-12
withphysicalparametersgivenbyGondoin(2003).Forthe
-10
case of V390Aur, the method enabledus to averageabout
12,300linesandtogetStokesIandStokesVprofileswitha Bl [G] --68
significantlyimprovedsignal–to–noiseratio(S/N).Asignif-
-4
icant Zeeman Stokes V signal was detected for each obser-
-2
vation.Thenullspectrumgivenbythestandardprocedure 0
(Donati et al. 1997) was also examined, which showed no 2
signal. We then computed the longitudinal magnetic field 4
B in G, using the first-order moment method (Donati et 24 26 28 30 32 34 36 38 40 42
l
al.1997,ReesandSemel1979).TodetermineB values,we HJD 2454700+
l
integrated the LSD profiles between -17 and +66km.s−1,
which reduced the significant blending by the profiles of
the faint stellar companion of V390 Aur (which is a SB2 0,68
0,66 CA II K
star, Konstantinova-Antovaet al. 2008).These profiles are 0,64 CA IR
0,62 H alpha
in some of our spectra. 0,60
0,58
The activity of the star for the same period has been nsity00,,5546
monitored with measurements of the relative intensity re- nte0,52
gCaarIdIinIgRth8e5c4o.n2tninmuuamnd(RHcα).foFrotrhCealiInIe–Kacttivhietyreinladtiicvaetoinrs- Relative I0000,,,,44454680
tensity of the emission core I/I(395nm) is measured. The 0,42
0,40
S/N is greater than 50 for the CaII K&H emission cores 0,38
0,36
and greaterthan 500 for the rest of the spectrallines men- 0,34
tioned above. The B and activity indicators behaviour is 0,32
l 24 26 28 30 32 34 36 38 40 42
presented in Table1 and in Fig.1. HJD 2454700+
In addition, we performed ZDI (Donati et al. 2006a)
Fig.1.VariabilityofB (top), CaII K,CaII IR854.2nm
l
for V390 Aur, using all these observations. In spite of a
and H (bottom) for V390 Aur.
α
large vsini value that makes V390 Aur a possible target
for classical Doppler Imaging, we do not present here a
photospheric map of brightness inhomogeneities, because
3. The magnetic field geometry of V390 Aur and
of the occasional Stokes I contamination by the double-
the behaviour of the activity indicators
linedstellarsystemassociatedwiththegiant.Withoutany
accurate orbital parameters available for this system, no
3.1. Zeeman Doppler Imaging
clean removal of its Stokes I signatures can be performed
following the approach of Donati (1999). The primary is InordertoreconstructthemagneticfieldgeometryofV390
the only detected contributor to the Stokes V line profiles, Aur based on modelling of the rotational modulation of
so that other components of the system do not affect the the StokesVprofiles,weappliedthe ZDIinversionmethod
magnetic map reconstruction. (Donati & Brown1997). The inversioncode we used is de-
2
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
scribedinDonatietal.(2006b)andenablesthesplitofthe
magnetic field in poloidal and toroidal components.
Tocalculatetherotationphasesofourobservations,we
tookasrotationalperiodthephotometriconeof9.825days,
vsini=29km.s−1, inclination angle of the axis of rotation
◦
56 (Konstantinova-Antova et al. 2008), and a limb dark-
eningcoefficientof0.75.We alsoassumedaconstantradial
velocity of 23.4km.s−1. Taking into account the complex
structure of the Stokes V profiles, we extend the order of
thesphericalharmonicsmodestoℓ≤25,sincenoimprove-
mentinthefittothe dataisachievedbyincreasingfurther
the maximum allowed value for ℓ.
Priortoreconstructingthefinalmagneticmap,wehave
estimatedthe surfacedifferentialrotationofV390 Aur,us-
ing the method proposed by Petit et al. (2002). We as-
sume a simple latitude dependence of the rotational shear
in the form Ω(l)=Ω −∆Ωsin2l, where Ω(l) is the rota-
eq
tion rate at latitude l, Ω the rotation rate of the equator
eq
and ∆Ω the difference in rotation rate between the po-
lar region and the equator. The best magnetic model is
achievedfor differenial rotation parametersequal to Ω =
eq
0.652±0.002rad.d−1and∆Ω=0.048±0.007rad.d−1,with
a reduced χ2 equal to 1. The value of the photometric
period lies between the equatorial and polar periods esti-
matedhereandcorrespondstotherotationperiodobtained
◦
arounda latitude of 30 , whichis the latitude seenorthog-
onally by the observer, assuming an inclination angle of
◦
56 .
The result of our modelling is presented in Fig.2. In
this figure, the upper pannel displays the Stokes V profiles
fitted with the model, and the three other panels show the
surface magnetic field distribution for V390 Aur as maps
of the radial, azimuthal and meridional components of the
magnetic field.
Ourmapoftheradialmagneticfieldrevealsastrongpo-
lar spot of positive polarity and several groups of spots of
negativepolaritysituatedatlowerlatitudes.Theazimuthal
component of the field presents a middle to high latitude
belt. We reconstruct about 80 percent of the surface mag-
neticenergyinthepoloidalmagneticcomponentwherethe
dipole structure dominates.
3.2. B and activity indicators
l
Figure 3, upper panel, shows the variations of B with re-
l
spect to the photometric phase already used in the ZDI
work.Thefoldingappearsgood,withtheunsigned|B |be-
l
ing maximum at phase 0 and minimum at phase 0.5. Fig.2. Upper pannel: LSD Stokes V profiles (solid lines)
The activity indicators vary smoothly (Fig.1). and their fit (dashed lines) after a correction of the mean
CaII 854.2 and Hα are synchronous, whereas CaII K radialvelocityofthe star.For displaypurposesthe profiles
deviates slightly. Though they appear to vary in op- are shifted vertically and the dotted horizontal lines indi-
position with |B |, folding of the data does not fit to cate the zero level. Lower pannels: Radial, azimuthal and
l
the photometric period, but suggests a longer period meridional maps for the reconstructed magnetic field for
(Fig.3). We performed a period search analysis for the V390Aur.Theverticalticksintheupperpartoftheradial
activity indicators CaII K, CaII IR andHα. We applied map show the phases when there are observations.
several period search techniques: Lomb-Scargle (Lomb,
1976; Scargle 1982) Lafler-Kinnman (Lafler & Kinman
1965), Bloomfield (Bloomfield 2000), CLEANest (Foster to CaII K, but as mentioned above, we are unable to
1995) and Deeming (Deeming 1975). Some of them gave determine precise values by a period search analysis due
indications ofperiodslongerthanthe photometric one and to the short dataset. Further study on this topic with a
mainly in the interval 10 - 15 days, but with an accuracy longer dataset is required.
of 2-3days due to the shortdatasetwe have.It might be a Taking into account possible longer rotational periods
tendencyforlongerperiodsofCaII IRandHα,compared of the activity indicators and comparing them with the
3
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
-16 1
Bl 1
-14 Bl 2
Bl [G]--11---02468 Normalized Intensity0,8
-2
0
2
4 0,6
6695 6700 6705 6710 6715 6720
0,0 0,2 0,4 0,6 0,8 1,0 Wavelength (A)
Phase ( P = 9,825 d )
Fig.4. Fit (solid line) with the model (see text) of the
spectrallinesintheLiregion(671nm).Thedashedvertical
line indicates the position of the LiI 670.8 nm.
0,66 Ca II K 1
Ca II K 2
0,64
4. Evolutionary considerations and dynamo
0,62
operation in V390 Aur
5)0,60
9
I / I (30,58 4.1. Determination of stellar parameters, lithium abundance,
0,56 and vsini
0,54 With the aim of improving the precision on lithium abun-
0,52 dance(NLi =log[n(Li)/n(H)]+12)andstellarparameters
(T , logg and [Fe/H]), we performed a spectral synthe-
0,50 eff
0,0 0,2 0,4 0,6 0,8 1,0 sis analysis on the Stokes I spectra of V390 Aur. We used
Phase
MARCSmodelsofatmosphere(Gustafssonetal.2008)and
theTurboSpectrumcode(Alvarez&Plez,1998)inorderto
producehighresolutionsyntheticspectraofthelithiumline
0,50
H alpha 1 region around 671 nm (see Canto-Martins et al. 2006 and
0,48 H alpha 2 2011 for the complete method and for atomic and molecu-
Ca IR 1
0,46 Ca IR 2 lar line lists references).Bestfit is displayedin Figure 4. It
0,44 is obtained for Teff =4970K±50K, logg =3.0±0.2dex,
and[Fe/H]=−0.05±0.05dex,withamicroturbulenceve-
Rc0,42 locityof1.5 km.s−1 andtakingintoaccount(witharadial
0,40 tangent profile) a macroturbulence velocity of 8 km.s−1 .
0,38
Line broadening is then correctly reproduced with a rota-
0,36 tion profile and vsini = 25±1km.s−1 . We checked that
0,34 this solutionalsoeasily andcorrectly fits other spectralre-
gions throughoutthe NARVAL wavelengthrange.Our T
0,0 0,2 0,4 0,6 0,8 1,0 eff
Phase valueisinperfectagreementwiththatderivedbyGondoin
(2003)and Bell & Gustafsson (1989) (4970±200K), and is
Fig.3. Bl, CaII K, Hα and CaII IR 854.2nm behaviour very similar to the value of 5000K obtained by Fekel &
for V390 Aur. Phase is for the photometric period of Balachandran(1993).We determined a lithium abundance
9.825d. The different colors in each of them correspond N =1.6±0.15dex. This value is in very good agreement
Li
to a different rotation (see the legend on each pannel). withthe previousdeterminationof1.5±0.2made by Fekel
& Balachandran (1993), and slightly lower than the value
of 1.8 obtained by Brown et al. (1989).
4.2. Evolutionary status
The position of V390 Aur in the HR diagram is shown in
B behaviour regarding the surface rotational period (the Fig. 5. We adopt the effective temperature of 4970±50K
l
photometric one), there might be a correlationin their be- that we derived from the Stokes I spectra. The luminosity
haviour, and also a good coincidence with the ZDI map is obtained using the parallax of the star given in the New
(Fig.2). The indicators have a stronger intensity at phases ReductionHipparcoscataloguebyvanLeeuwen(2007),the
around 0.4−0.5 when we observe areas with positive and V magnitude from the 1997 Hipparcos catalogue, and the
negativepolaritiesandtheresultantB iscloseto0.Atthe bolometric correction following Buzzoni et al. (2010) pre-
l
opposite, they have a weaker intensity near phase 0 when scription (i.e., BC = −0.253); the error bar reflects only
the negative polarities are dominant in the resultantlongi- the uncertainty of the parallax. The adopted parameters
tudinal magnetic field. lead to a stellar radius of 6.4R⊙.
4
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
the dashed lines indicate the beginning and the end of the
first dredge-up (warmer and cooler lines respectively).
Based on these up-to-date models and taking into ac-
counttheobservationaluncertainties,wedetermineamass
of2.25±0.25M⊙forV390Aur.Notethatatthelocationof
V390 Aur the effect of rotation on the mass determination
is negligible compared to the uncertainties on the paral-
lax and on the effective temperature. The star appears to
startthe ascentoftheRedGiantBranch(RGB)andisnot
bright enough to lie in the clump. This agrees with previ-
ous studies by Gondoin (2003) and Konstantinova-Antova
et al. (2009) which identified V390 Aur as a star of about
1.8−2M⊙ located at the base of the RGB2
4.3. Lithium abundance and rotation
At the base of the RGB, the star experiences the so-called
first dredge-up (see the dashed lines in Fig.5). Although
its convective envelope has not yet reached its maximum
extent in depth, it already encompasses ∼ 0.83M⊙ and
∼0.54R∗atthelocationintheHRDofV390Aur.Starting
Fig.5. Position of V390 Aur in the H-R diagram. The fromthe lithiuminterstellarmediumabundanceof3.3,the
adoptedeffectivetemperatureisdeterminedfromouranal- corresponding standard Li dilution is found to lead to a
ysis of the Stokes I spectra of V390 Aur. The luminosity theoretical surface abundance N(Li) equal to 1.68 (where
is estimated using the parallax from the New Reduction N(Li)=log[n(Li)/n(H)]+12) at this evolutionary point.
Hipparcos catalogue by van Leeuwen (2007; see text for This prediction is in rather good agreement with our ob-
more details). Standard evolutionary tracks (solid lines) servationalvalueof1.6±0.15.We notehoweverthatinthis
computed for [Fe/H] = 0 are shown up to the RGB tip domaininstellarmassandeffectivetemperaturemanysub-
for different labelled stellar masses. For the 2.25M⊙ case, giant and giant stars exhibit Li abundances significantly
the predictions for models with rotation-induced mixing lower than predicted by the standard dilution models (e.g.
treated as in Charbonnel & Lagarde(2010) are also shown do Nascimento et al. 2000; de Laverny et al. 2003 ; Canto-
fortwoinitialrotationvelocitiesof50and110km.s−1(dot- Martins et al. 2011). This is well accounted for by addi-
tedlines). The dashedlines delimit the firstdredge-up,the tional Li depletion due to rotational mixing on the main
cooler line marking the deepest penetration of the convec- sequence (see e.g. Charbonnel & Talon 1999, Palacios et
tive envelope. The dashed-dotted line indicates the begin- al.2003,Pasquinietal.2004,Charbonnel& Lagarde2010,
ning of central He-burning for the mass range covered by and references therein). For initial rotation velocities of 50
the tracks. and 110km.s−1 our 2.25M⊙ rotating models predict re-
spectively N(Li) of 1.46 and 0.87 at the effective tempera-
ture ofV390Aur.The LicontentofV390Aur canthus be
Figure 5 also presents evolutionary tracks correspond- explained by assuming that the star arrived on the main
ing to up-to-date stellar models of various initial masses sequence with a modest rotation velocity of the order of
computed with the code STAREVOL1 at solar metallicity, 50km.s−1.
whichisveryclosefromthe [Fe/H]=−0.05±0.05dex de- Another possibility is that the hydrodynamical mecha-
rivedfromtheStokesIspectraandonlyslightlyhigherthan nisms induced by stellar rotation (i.e., meridional circula-
that derived by Fekel & Balachandran (1993; [Fe/H] = tionandshearturbulence)werehamperedinmodifyingthe
−0.3). We plot standard (non-rotating) tracks, except in content of this fragile element during the evolution of the
the 2.25M⊙ case for which we also show predictions for star. Indeed large internal angular momentum gradients,
rotating models with initial rotation velocities on the zero whicharebuiltupinthecaseofpurehydrodynamicalmod-
agemainsequenceof50and110km.s−1 (thehigherthero- elssuchasthosepresentedinFig.5,favourthe transportof
tationvelocity,thebrighterthetrack).Rotatingmodelsare chemicals inside the star, and thus the depletion of fragile
computed using Zahn’s (1992) and Maeder & Zahn (1998) lithium as requested by Li data in most field and cluster
formalism where the transport of angular momentum and counterparts of V390 Aur (see e.g. Talon & Charbonnel
chemicals is dominated by the Eddington-Sweet merid- 2010 for a more detailed discussion and references). So if
ional circulation and shear instabilities (see Charbonnel & an additional chief mechanism for the internal redistribu-
Lagarde2010for details onthe model assumptions andin- tionofangularmomentumwasabletomaintainsolidbody
put microphysics). For clarity the tracks are drawn up to rotation, or at least to reduce strongly differential rotation
the RGB tip only, but the dashed-dotted line shows the in the radiative stellar interior, much less Li depletion is
location of the clump (i.e., the beginning of central He- expected than in the pure hydrodynamical rotating case,
burning) for the various masses considered. Additionally even if the star was rotating fast at its arrivalon the main
sequence.InthemassdomainofV390Aurwherethetrans-
1 Most of themodels shown in Fig. 5 arefrom Charbonnel &
Lagarde(2010)andLagardeetal.,inprep.The2.25M⊙models 2 Charbonnel & Balachandran (2000) determined a mass of
were computed especially for the present study with the same 1.1±0.5M⊙ forV390Aur,duetoanerrorintheassumedTeff
assumptions and inputphysics. for this star (see their Table1).
5
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
portofangularmomentumbyinternalgravitywavesshould On the other hand, recent discoveries of Jupiter-mass
playnoroleduringthemainsequence(Talon&Charbonnel planets orbiting their host star at inner-solar system dis-
2008),efficientredistributionofangularmomentum,which tances could also be considered as a possible explanation
determines the extent and magnitude of rotation-induced of the Li content and fast rotation in V390 Aur. Applying
mixing and the resulting Li depletion, could occur by hy- the formula by Massarotti (2008) with the parameters ob-
dromagnetic means related either to a fossil magnetic field tained by our model and assuming a circular orbit with a
or to a dynamo driven by shear in the radiative layers of semi-major axis equal to the radius of the star (hypothesis
thestar(seee.g.,Charbonneau&MacGregor1993,Barnes for the just engulfed planet, that could explain not only
et al. 1999, Spruit 2002, Eggenberger et al. 2005, Duez et the fast rotation, but also the enchanced Li), we estimate
al. 2010). Rotating models, including the combined effects thatthe star hasto engulfabrowndwarfofabout20M
Jup
of magnetic fields, meridional circulation and shear turbu- to explain its V of 34km.s−1. In the same time, however,
eq
lence,willbecomputedinthenearfuturetoquantitatively Alibert et al. (2011) point out the observed lack of short
test this hypothesis in the case of stars at the evolution period planets orbiting 2M⊙ stars.
stage of V390 Aur. As to the forcing of stellar rotation by a close compan-
Additionally, let us recall that V390 Aur has a rather ion, let us recall that V390 Aur was found to be a wide
high rotationvelocity for its evolutionary stage (35km.s−1 binarysystemandsynchronizationcannotplayaroleinits
for vsini = 29km.s−1 and i = 56◦), as already noted by fast rotation (Konstantinova-Antova et al. 2008).
Fekel & Balachandran(1993).This is rather peculiar since
most(i.e.,morethan95%)ofthestarsofthesamespectral
4.4. Dynamo
class have already been spun down to a rotational veloc-
ity of ∼ 5km.s−1 or less (Gray 1982; De Medeiros et al.
Onthebasisofthetheoreticalconvectiveturnovertimeand
1996;De Medeiros & Mayor 1999).This general behaviour
assumingthatthephotometricperiod(9.825d)istherota-
can be explained simply by changes in the moment of in-
tionalone,wecanestimatethevalueoftheRossbynumber
ertia and in the stellar radius when the stars evolve to the
forV390Aur.Ifthevalueof126days(foralayerthatcorre-
RGB,andthe transitionto a slowrotationoccursatanef-
spondsto1/2H abovethebaseoftheconvectiveenvelope)
p
fective temperature at the order of 5500K.Taking into ac-
isconsidered,whichisobtainedfromourstandard2.25M⊙
count this secular effect and assuming rigid-body rotation
model (computed with a mixing length parameter of 1.6)
and conservation of angular momentum for our 2.25M⊙ at the location of V390 Aur in the HR diagram, then the
model,onecanobtainamaximumrotationvelocityofonly
Rossby number is 0.08. In the case of convective turnover
∼ 15km.s−1 (for a rotation velocity of 180km.s−1 on the
time of 52 days, i.e. the value determined for the middle
zeroagemainsequence)atthe locationofV390Aurinthe
(in radius) of the convective envelope, the Rossby number
HRD.
is 0.19. In both cases, these values are indicative of a very
Although fast rotation is observed only in few percents efficient α–ω dynamo.
of red giant stars, it appears to be a common feature of
a group of chromospherically active single giants (Fekel
& Balachandran,1993;Konstantinova-Antovaet al. 2009). 5. Discussion
Forthemoment,possibleexplanationsfortheirfasterrota-
tioncouldbeangularmomentumdredge-upfromthefaster Thepresentstudyrevealedacomplexstructureofthemag-
rotatinginteriorduringthefirstdredge-upphase(Simon& netic field at the surface of V390 Aur. In addition to the
Drake1989),orspin-upfromplanet engulfmentthat could magneticpolarspotofpositivepolarity,groupsofclosesit-
alsoincreasethelithiumabundance(e.g.Siess&Livio1999; uated magnetic spots of negative polarity are detected at
Livio & Soker 2002; Massarotti et al. 2008; Carlberg et al. mid-latitudes. Such polar spots are predicted for fast rota-
2009).Thefirstoptionrequiresrotationgradientinthestel- torsbySchu¨ssler(1996),andobservedbyPetitetal.(2004).
lar interior with the near core region spinning much faster TemperaturepolarspotsarereportedforgiantsinRSCVn
than the more external layers. According to the theoreti- systemsandforFKCom–typestarsbyStrassmeier(1997),
cal works by Palacios et al. (2006) and Brun & Palacios Korhonen et al. (2002) and Strassmeier (2002).
(2009) rotational gradients, both radial and longitudinal This surface structure with close magnetic spots could
ones, should appear in the convective envelope and the ra- explain the flare activity properties of the star reported
diativezoneforstarswithmassupto2−2.5M⊙duringthe in optical and X–ray (Konstantinova-Antova et al. 2000;
firstdredge-upphaseandthebeginningascendoftheRGB. Gondoin 2003; Konstantinova-Antova et al. 2005), like the
Our observations support their predictions for gradients in groups of optical flares and the ”continuous flare activity”
the convective envelope. In addition, a gradient between in the corona.
the envelope and (presumable) fast rotating radiative inte- Theobservedazimuthalmapbeltathigh-latitudecould
riorcoulddredge–upangularmomentumandspeed-upthe be interpreted as a toroidal component near the surface
upper convective layers and the surface (as we observe in (like in the FK Com–type star HD 199178 and in the RS
V390 Aur and other fast rotating single giants). CVn star HR 1099,Petit et al. 2004aand b) and is indica-
However, one should be cautious with all these inter- tive of the site of the dynamo operation. In this star, it is
pretations, since the knowledge on the various transport possible for the dynamo to operate not close to the base of
mechanisms during the RGB phase is not complete yet. theconvectiveenvelope,butinawholeregionwithinit.The
It is rather possible, not only meridional circulation and otherpossibility isthe toroidalmagneticfieldto be formed
shear-induced turbulence to play role for transport of an- near the base of the convection zone and to rise up as one
gularmomentumandchemicals,butalsootherfactors,like whole. This possibility is less likely, because of the vigor-
magnetic fields etc. Further work in this direction is re- ous convection during the first dredge-up phase. The first
quired. assumption, however, presumes the existence of a gradient
6
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
of rotationin the convectiveenvelope closerto the surface, Palmerini for a useful discussion. The observations were funded un-
not only at its base. The fast surface rotationof V390 Aur der an OPTICON grant in 2008. R.K.-A. is thankful for the pos-
andthesignificantdifferentialrotation∆Ω=0.048rad.d−1 sibility to work for six months in 2010 as a visiting researcher in
LATT, Tarbes under Bulgarian NSF grant DSAB 02/3/2010. R.K.-
are in support of such a rotational gradient.
A.,S.Ts.andR.B.acknowledge partialfinancial supportunder NSF
Thecomparisontosolar-typedwarfsrevealedthatSun- contract DO02-85.R.K.-A.andR.B.alsoacknowledges supportun-
like stars seem to display a higher fraction of toroidal field der the RILA/EGIDE exchange program (contract RILA 05/10).
for a rotation period similar to that of V390 Aur (Petit C.C. aknowledges financial support from the Swiss National Science
Foundation (FNS)andtheFrenchProgrammeNationaldePhysique
et al. 2008). The mostly poloidal field geometry is more
Stellaire(PNPS)ofCNRS/INSU.
reminiscent of M dwarfs with deep convective envelopes
(Morinetal.2008,2010).But,onemustbe cautiousabout
thisquestion,becausethefractionoftoroidalfieldcanvary References
significantly during stellar cycles (Petit et al. 2009).
AlibertY.,MordasiniC.,BenzW.,2011,A&A,526,63
Alvarez,R.,Plez,B.1998,A&A,330,1109
Auri`ereM.2003,inConf.Proc.”MagnetismandActivityoftheSun
6. Conclusions
and Stars”, EAS Publ. Series 9, eds. J. Arnaud & N. Meunier,
(Paris,EDPSciences), 105
1. The complex structure of the surface magnetic field
Auri`ereM.,Konstantinova-AntovaR.,PetitP.etal.,2011,A&A,534,
in the ZDI map and the analysis of the poloidal and 139
toroidal field components suggest a dynamo operation Barnes,G.,Charbonneau, P.,MacGregor,K.B.,1999,ApJ,511,466
in V390 Aur. It is rather different than the ZDI map BellR.A.,GustafssonB.,1989,MNRAS,236,653
Bloomfield,P.,2000, Fourieranalysisoftimeseries:anintroduction,
of another late G giant, EK Eri, a descendant of an
secondedition,(NewYork,JohnWiley&Sons,Inc.),261
Ap star, where a strong dominating dipole structure is Brown,J.A.,Sneden, C.,Lambert,D.L.,Dutchover, E.,1989,ApJS,
observed (Auriere et al. 2011). In addition, it is rather 71,293
likely for the dynamo to be of α–ω type, taking into BrunA.S.,PalaciosA.,2009,ApJ,702,1078
Buzzoni A., Patelli L., Bellazzini M., Pecci F. Fusi, Oliva E. 2010,
account the fast rotation and the evolutionary stage of
MNRAS,403,1592
V390 Aur with a well-developed convective envelope.
CantoMartinsB.-L.,L`ebreA.,DeLavernyP.etal.,2006A&A451,
The calculated Rossby number obtained from tailor- 993
made stellar evolution models is also in support of an Canto Martins B.-L., L`ebre A., Palacios A, et al., 2011, A&A 527,
efficient α–ω dynamo.A significanttoroidalcomponent A94
Carlberg,J.K.,Majewski,S.R.,Arras,P.,2009,ApJ,700,832
is developed which results in low-latitude spots and an
Duez,V.,BraithwaiteJ.,Mathis,S.,2010,ApJLetters,724,L34
azimuthal field belt at the surface. Such a belt is al- Charbonneau, P.,MacGregor,K.B.,1993, ApJ,417,762
readyobservedinFKComandRSCVn- type stars.A CharbonnelC.,Talon,S.,1999,A&A,351,635
possibleexplanationisthatthereisagradientofthero- CharbonnelC.,BalachandranS.,2000,A&A,359,563
CharbonnelC.,LagardeN.2010, A&A,522,10
tation in the extended convectiveenvelope which exists
FosterG.1995,ApJ,109,1889
during the first dredge-up phase and it is possible for
Deeming,T.J.,1975,Ap&SS,36,137
the dynamo to operate within it, not only at the base de Laverny P., do Nascimento J.D.Jr., L`ebre A., De Medeiros J.R.,
ofthisenvelope,asitisthecaseofaclassicalsolar-type 2003,A&A410,937
dynamo, but also closer to the surface. DeMedeiros,J.R.,DaRocha,C.,Mayor,M.,1996,A&A,314,499
DeMedeiros,J.R.,Mayor,M.,1999,A&ASuppl.,139,433
2. The structure of the surface magnetic field with close
DonatiJ.-F.,BrownS.F.1997,A&A,326,1135
situated active areas could also explain the properties DonatiJ.-F.,SemelM.,CarterB.D.etal.1997,MNRAS,291,658
of the flare activity in V390 Aur, observed in the past: DonatiJ.-F.,1999, MNRAS,302,457
groups of optical flares and continuous flaring in the Donati, J.-F., Howarth, I. D., Jardine, M. M. et al., 2006, MNRAS,
370,629
corona.
Eggenberger,P.,Maeder,A.,Meynet, G.,2005,A&A,440,L9
3. Theactivityindicatorsbehaviorpointstoapossiblede-
FekelF.C.,MarschallL.A.1991,AJ,102,1439
crease of the rotation rate in the chromosphere com- FekelF.C.,BalachandranS.1993, ApJ,403,708
pared to the rate in the photosphere. Further study is GondoinP.1999,A&A,352,217
necessary to determine the exact rotation periods for GondoinP.2003,A&A,404,355
GrayD.F.,1982,ApJ,262,682
the activity indicators in the chromosphere.
Gustafsson, B., Edvardsson, B., Eriksson, K., Jorgensen U.G.,
Inthe future,itwouldbe interestingto studymoresin- NordlundA.,PlezB.,2008,A&A486,951
gle giants - their surface magnetic field configurations HootenJ.T.,HallD.S.1990,ApJS,74,225
and the dynamo, and to getknowledge of the structure Hu¨nschM.,SchmittJ.H.M.M.,VogesW.1998, A&A,127,251
Konstantinova-Antova R.K.,AntovA.P.,BachevR.S.2000,IBVS,
of their atmospheres, using ZDI. Recently, we started
4867
a ZDI programme of eight giants with rotational peri-
Konstantinova-Antova R.K.,AntovA.P.,ZhilyaevB.E.etal.2005,
ods from 5 days to 600 days, including stars from the AN,326,38
Hertzsprunggaptotheclump.Inthefuture,itwouldbe Konstantinova-Antova R.,Auri`ereM.,IlievI.Kh. etal.,2008, A&A,
also possible for ZDI to be applied to AGB stars. Even 480,475
Konstantinova-AntovaR.,Auri`ereM.,Schro¨derK.-P.,PetitP.,2009,
ifforsomeofthesegiantssmallvsinidonotallowgood
in Proc. IAUS 259 ”Cosmic Magnetic Fields: From Planets, to
resolution mapping, such mapping makes it possible to Stars and Galaxies”, eds. K.G. Strassmeier, A.G. Kosovichev &
determine the surface magnetic field strength, to disen- J.E.Beckman, (Cambridge,CambridgeUniversityPress),433
tangle poloidal and toroidal components and finally to KorhonenH.,BerdyuginaS.V.,TuominenI.2002,A&A,390,179
Lafler,J.&Kinman,T.D.,1965,ApJS,11,216
makediagramssimilartothoseobtainedforMdwarves
Lomb,N.R.,1976,Ap&SS,39,447
and solar twins (Morin et al. 2010).
Livio,M.,SokerN.,2002,ApJ,571,L161
Maeder,A.,Meynet,G.,2003, A&A,411,543
Acknowledgements. We thank the TBL team for providing service Massarotti, A., Latham, D.W., Stefanik, R.P., Fogel, J., 2008, AJ,
mode observing with NARVAL. R.K.- A. is thankful to Dr. Sara 135,209
7
Konstantinova-Antovaet al.: Magnetic field structure in V390 Aur
Massarotti,A.,2008,AJ,135,2287
Mestel,L.,Weiss,N.O.,1987,MNRAS,226,123
MorinJ.,DonatiJ.-F.,PetitP.etal.,2008,MNRAS,390,567
MorinJ.,DonatiJ.-F.,PetitP.etal.,2010,MNRAS,407,2269
Palacios,A.,Talon,S.,CharbonnelC.,Forestini,M.,2003,A&A,399,
603
PalaciosA.,Charbonnel C.,TalonS.,SiessL.,2006, A&A,453,261
Pasquini, L., Randich, S., Zoccali, M., Hill, V., Charbonnel, C.,
Nordstrom,B.,2004,A&A,424,951
PetitP.,DonatiJ.-F.CollierCameronA.,2002,MNRAS,334,374
PetitP.,DonatiJ.-F.,WadeG.A.etal.2004, MNRAS,348,1175
PetitP.,DonatiJ.-F.,OliveiraJ.M.etal.2004,MNRAS,351,826
PetitP.,DintransB.,SolankiS.K.etal.2008,MNRAS,388,80
PetitP.,DintransB.,Morgenthaler A.etal.,2009,A&A508L,9
ReesD.E.,SemelM.,1979, A&A,74,1
Scargle,J.D.,1982,ApJ,263,835
Schu¨ssler M. 1996, in Proc. IAUS 176 ”Stellar Surface Structure”,
eds.K.G.Strassmeier&J.Linsky,(Dordrecht,KluwerAcademic
Publishers),269
SemelM.,DonatiJ.-F.,Rees,D.E.,1993, A&A,278,231
SiessL.,LivioM.,1999,MNRAS,308,1133
SimonT.,DrakeS.A.,1989,ApJ,346,303
Spruit,H.C.,2002, A&A,381,923
SpurrA.J.,HoffD.1987,IBVS,3028
StrassmeierK.G.1997,A&A,319,535
StrassmeierK.G.2002,AN,323,155
Talon,S.,Charbonnel,C.,2008,A&A,482,597
Talon,S.,Charbonnel, C.,2010, inProc.IAUS268”Light Elements
in the Universe”, eds. C. Charbonnel, M. Tosi, F. Primas & C.
Chiappini,(Cambridge,CambridgeUniversityPress),365
vanLeeuwenF.2007,A&A,474,653
8