The evolution of surface magnetic fields in young solar-type stars
aa r X i v : . [ a s t r o - ph . S R ] O c t Young Stars & Planets Near the SunProceedings IAU Symposium No. 314, 2015J. H. Kastner, B. Stelzer, & S. A. Metchev, eds. c (cid:13) The evolution of surface magnetic fields inyoung solar-type stars
Colin P. Folsom , , Pascal Petit , J´erˆome Bouvier , , Julien Morin ,Agn`es L`ebre , and Jean-Fran¸cois Donati Universit´e Grenoble Alpes, IPAG, F-38000 Grenoble, France CNRS, IPAG, F-38000 Grenoble, Franceemail: [email protected] IRAP, CNRS and Universit´e de Toulouse, 14 avenue ´Edouard Belin, 31400, Toulouse, France LUPM, UMR 5299, CNRS and Universit´e Montpellier II - Place E. Bataillon, 34090Montpellier, France
Abstract.
Surface rotation rates of young solar-type stars display drastic changes at the end ofthe pre-main sequence through the early main sequence. This may trigger corresponding changesin the magnetic dynamos operating in these stars, which ought to be observable in their surfacemagnetic fields. We present here the first results of an observational effort aimed at characterizingthe evolution of stellar magnetic fields through this critical phase. We observed stars from openclusters and associations, which range from 20 to 600 Myr, and used Zeeman Doppler Imagingto characterize their complex magnetic fields. We find a clear trend towards weaker magneticfields for older ages, as well as a tight correlation between magnetic field strength and Rossbynumber over this age range. Comparing to results for younger T Tauri stars, we observe a verysignificant change in magnetic strength and geometry, as the radiative core develops during thelate pre-main sequence.
Keywords. stars: magnetic field, stars: evolution, stars: rotation, stars: pre-main sequence,stars: solar-type
1. Introduction
Solar type stars display a major change in their rotation rates as they cross the pre-main sequence (PMS) and settle on to the main sequence (e.g., Gallet & Bouvier 2013,2015). Early on the PMS, stars strongly interact with their disks, and this regulates theirrotation rates. After a few Myr, a star decouples from its disk and is still contracting,which results in a rapid spin up. On a longer timescale, the star loses angular momentumthrough its magnetized wind, and starts to spin down as it settles onto the main sequence.Magnetic fields in solar-type stars are thought to be generated by a dynamo process,possibly an α -Ω dynamo. Thus, the large change in rotation rates around the zero-agemain sequence (ZAMS) should have a large impact on magnetic properties. Conversely,the rotational spin down is controlled by magnetic fields (e.g. Matt et al. 2012). It is thuscrucial to understand the connection between magnetism, rotation, and age in order to tofully decipher early stellar evolution and its potential impact on, e.g., the early evolutionof planetary systems and their habitability.We present here the first results from an investigation of the magnetic properties ofyoung solar-type stars as a function of rotation rate and age.1 C.P. Folsom, P. Petit, J. Bouvier, J. Morin & A. Lebre
2. Observations
Targets for our study are members of known moving groups and clusters, in the agerange of ∼
20 to ∼
600 Myr, when most of the rotational evolution occurs (cf. Bouvieret al. 2014). We exclude T Tauri stars, to avoid stars whose rotation rate is still regulatedby disk interactions, and that have been studied in previous projects (e.g. Donati et al.2011). We also exclude older stars from our selection, which have converged onto thegyrochronologic sequence and whose magnetic properties have been already characterized(e.g. Petit et al. 2008). Indeed, our sample was devised to fill the evolutionary gap betweenearly PMS stars and mature main sequence stars. We focus on 0.7-1.0 M ⊙ stars withknown rotational periods, usually derived from photometric monitoring.In order to characterize their magnetic fields, the targets were observed with high-resolution spectropolarimetry. We used ESPaDOnS at the Canada France Hawaii Tele-scope and Narval at the T´elescope Bernard Lyot in France. These are essentially identicalspectropolarimeters, with a resolution of 65000, and a wavelength coverage of 3700 to10500 ˚A. Reduced observations contain a circularly polarized Stokes V spectrum as wellas a total intensity Stokes I spectrum (see Donati et al. 1997 for data reduction). Obser-vations of a target consist of a series of typically 15 observations, spread over the rotationcycle of the star. This phase-resolved series of observations allows us to both diagnosethe presence of a magnetic field, and characterize its geometry, as discussed below.
3. Analysis
The high resolution, high S/N spectra allow us to derive precise fundamental parame-ters for the stars. Using spectrum synthesis, and by fitting model spectra directly to theobservations, we derive T eff , log g , v sin i , microturbulence, and [Fe/H] for all the stars inour sample. The spectra can also be used to determine detailed chemical abundances (e.g.Lithium). Most of the stars have reliable distance estimates, and thus we can determineintrinsic luminosities, and from that radii and masses.Accurate rotation periods are essential for this analysis. Thus we re-derived the rotationperiods for the stars in this study. For most stars, our results support the literaturephotometric periods. However, for a few stars, the literature periods are inconsistent withour spectropolarimetric time series, and for those stars we use the periods we derive.In order to derive the magnetic field strengths and geometries of the stars in this study,we use the Zeeman Doppler Imaging (ZDI; Donati et al. 2006). ZDI is a tomographictechnique. The rotationally modulated variability of the line profile is inverted to recon-struct the stellar magnetic field. Stokes V line profiles generated by Zeeman splitting areused. Our version of the method uses a spherical harmonic decomposition to describe themagnetic field, and the inversion process is regularized using a maximum entropy regu-larization in order to provide a stable unique solution. The v sin i and rotation periodsderived above were used as input, as well as an inclination of the rotation axis based on v sin i , radius, and period. An example magnetic map from ZDI is presented in Fig. 1.
4. Discussion
Even though our survey of magnetic field properties in young stars is not fully com-pleted yet, some early trends are apparent. We observe a significant decrease of the globalaverage magnetic field strength with age, as shown in Fig. 2. The trend is already seen atthe ZAMS, albeit with a large dispersion at a given age, and considerable overlap betweenthe mean magnetic field distributions of stars over the age range from 20 to 250 Myr. agnetic fields in young solar-type stars Figure 1.
Example of a magnetic map for TYC 6349-0200-1 produced using ZDI. Arrowsindicate magnetic field orientation, red arrows have a positive radial component, and blue arrowshave a negative radial component.
The same trend extends all the way to the MS, as previously reported by Vidotto et al.(2014) using the Bcool sample (Petit et al., in prep.). We find an even tighter correlationbetween average magnetic strength and Rossby number ( R o ), as shown in Fig. 2. Indeed,our ZAMS sample complements and extrapolates to younger ages the h B i - R o correlationseen on the MS for the Bcool sample. Only the fastest rotator of our sample, LO Peg,may show signs of magnetic saturation at very low Rossby number.An interesting comparison sample are the classical T Tauri stars (cTTS) from theMaPP project (e.g. Donati et al. 2011). These are younger pre-main sequence stars thatare still actively accreting and are predominantly convective. We see a clear distinctionbetween the magnetic properties of the MaPP sample and our older stars. cTTS havemuch stronger magnetic fields, and their fields are much more poloidal and aligned withthe stellar rotation axis (Fig. 3). This dramatic difference likely reflects a change in theinternal structure of the stars, going from fully convective at the T Tauri stage to beinglargely radiative on the ZAMS. This structural change appears to strongly affect thetype of dynamo operating in PMS stars, as suggested by Gregory et al. (2012). A similardifference is seen between partially and full convective M dwarfs (Morin et al. 2010).The three samples are shown together in Fig. 3, where Rossby number is plotted as afunction of age. It illustrates two main trends: i) the evolution of magnetic properties fromthe early PMS to the end of the PMS appears to be mainly driven by structural changes,with significant differences seen in magnetic strengths and topologies of PMS and ZAMSstars at a given Rossby number, while ii) the evolution of magnetic properties from thelate-PMS to the ZAMS and MS strongly correlates with Rossby number, presumablyreflecting the decreasing efficiency of the stellar dynamo as the stars are spun down.Beyond these clear trends, it is worth noticing the large scatter of magnetic propertiesseen on the ZAMS. While intrinsic variability contributes to this scatter, it could berelated to a third parameter, possibly the large amount of internal differential rotationpredicted at this phase of evolution (Gallet & Bouvier 2013, 2015).
5. Conclusion
By studying the magnetic properties of a sample of young solar-type stars, we havefilled the evolutionary gap between young PMS T Tauri stars and mature main sequencestars. Comparing samples, the main finding is that magnetic properties scale primarilywith internal structure during PMS evolution, and with rotation during ZAMS/early-MSevolution. The transition from a fully convective to a partly radiative interior during thePMS yields complex non-axisymmetric fields on the ZAMS, while the spin down of starson the MS drives the decline of magnetic field strength on a longer timescale.While we have drawn some preliminary conclusions here, the study is ongoing with a C.P. Folsom, P. Petit, J. Bouvier, J. Morin & A. Lebre
10 100 1000 10000
Age (Myr) < B > ( G ) ToupiesBcool
LO Peg
Rossby number < B > ( G ) ToupiesBcool
Figure 2.
Trends in magnetic field strength with age (left) and Rossby number (right). Blackpoints are from our sample, and red points are from the Bcool sample of Petit et al. (in prep.).A power law fit of the Rossby number trend produces h B i ∝ R − . ± . o . T eff (K) L ( L ⊙ ) TW HyaAA Tau BP TauGQ Lup DN TauV4046 Sgr AV4046 Sgr B
HIP 12545TYC 6349-0200-1TYC 6878-0195-1BD -16351 HIP 76768TYC 5164-567-1 LO PegPW AndHII 296 TYC 0486-4943-1PELS 031DX LeoV447 LacV439 And AV 1826AV 2177 A x i s y mm e t r y ( s h a p e ) / P o l o i d a l ( c o l o u r ) (cid:0) B (cid:1) Age(Myr) R o ss b y nu m b e r TW HyaAA TauBP Tau GQ LupDN Tau V4046 Sgr AV4046 Sgr B
HIP 12545TYC 6349-0200-1TYC 6878-0195-1BD -16351 HIP 76768TYC 5164-567-1LO PegPW AndHII 296TYC 0486-4943-1HII 739PELS 031DX LeoV447 LacV439 AndAV 1826AV 2177
HD 201091HD 22049HD 101501 HD 10476HD 3651HD 39587HD 72905HD 131156AHD 131156B A x i s y mm e t r y ( s h a p e ) / P o l o i d a l ( c o l o u r ) (cid:0) B (cid:1) Figure 3.
Magnetic properties of stars in our sample (black labels), stars in the Bcool sample(red labels), and stars in the MaPP sample (blue labels). Left frame: our stars and the MaPPsample in an HR diagram, with pre-main sequence evolutionary tracks. Right frame: our stars,the Bcool stars, and the MaPP stars in the age-Rossby number plane. Symbol size correspondsto magnetic field strength, symbol color corresponds to the ratio of poloidal to toroidal field,and shape corresponds to the degree of axisymmetry of the field.
Large Program being performed at the CFHT (‘The History of the Magnetic Sun’, PIP. Petit). This should help us further investigate differences in magnetic properties withrotation at a specific age, and differences with age at a specific Rossby number.
References
Bouvier, J. 2014
Protostars and Planets VI , 433Donati, J.-F. et al. 2006,
MNRAS , 370, 629Donati, J.-F. et al. 2011,
MNRAS , 417, 1747Donati, J.-F. et al. 1997,
MNRAS , 291, 658Gallet, F., Bouvier, J. 2013
A&A , 556, A36Gallet, F., Bouvier, J. 2015
A&A , 577, A98Gregory, S. G. et al. 2012
ApJ , 755, 97Matt, S. P., MacGregor, K. B., Pinsonneault, M. H., Greene, T. P. 2012,
ApJ , 754, L26Morin, J. et al. 2010,
MNRAS , 407, 2269Petit, P. et al. 2008
MNRAS , 388, 80Vidotto, A. A. et al. 2014