arXiv:1001.0032v1 [astro-ph.SR] 30 Dec 2009Draft version November 15, 2018 Preprint typeset using L ATEX style emulateapj v. 08/22/09 ASTEROSEISMIC INVESTIGATION OF KNOWN PLANET HOSTS IN THE KEPLER FIELD J. Christensen-Dalsgaard1,2, H. Kjeldsen1,2, T. M. Brown3, R. L. Gilliland4, T. Arentoft1,2, S. Frandsen1,2, P.-O. Quirion1,2,5, W. J. Borucki6, D. Koch6, and J. M. Jenkins7 Draft version November 15, 2018 ABSTRACT In addition to its great potential for characterizing extra-solar p lanetary systems the Kepler mis- sionis providing unique data on stellar oscillations. A key aspect of Keplerasteroseismology is the application to solar-like oscillations of main-sequence stars. As an ex ample we here consider an ini- tial analysis of data for three stars in the Keplerfield for which planetary transits were known from ground-based observations. For one of these, HAT-P-7, we obt ain a detailed frequency spectrum and hence strong constraints on the stellar properties. The remaining two stars show definite evidence for solar-like oscillations, yielding a preliminary estimate of their mean dens ities. Subject headings: stars: fundamental parameters — stars: oscillations — planetary systems 1.INTRODUCTION The main goal of the Kepler mission is to character- ize extra-solar planetary systems, particularly Earth-like planets in the habitable zone (e.g., Borucki et al. 2009). The mission detects the presence of planets through the minute reduction of the light from a star as a planet crosses the line of sight. Several observations of such reductions at fixed time intervals for a given star, and extensive follow-up observations, are used to verify that the effect results from planet transits and to characterize the planet. To ensure a reasonable chance of detection Keplerobserves more than 100,000 stars simultaneously, in a fixed field in the Cygnus-Lyra region. Most stars are observed at a cadence of 29.4 min, but a subset of up to 512 stars can be observed at a short cadence (SC) of 58.85s. Keplerwas launched on 6 March 2009 and data from the commissioning period and the first month of regular observations are now available. The very high photometric accuracy required to detect planet transits (Borucki et al. 2010; Koch et al. 2010) also makes the Keplerobservations of great interest for asteroseismic studies of stellar interiors. In particular, the SC data allow investigations of solar-like oscillations in main-sequence stars. Apart from the great astrophys- ical interest of such investigations they also provide pow- erful tools to characterize stars that host planetary sys- tems (Kjeldsen et al. 2009). In stars with effective temperature Teff<∼7000K we expect to see oscillations similar to those observed in the Sun (e.g., Christensen-Dalsgaard 2002), excited stochas- ticallybythe near-surfaceconvection. Theseareacoustic modes of high radial order; in main-sequence stars such 1Department of Physics and Astronomy, Aarhus University, DK-8000 Aarhus C, Denmark: e-mail jcd@phys.au.dk 2Danish AsteroSeismology Centre 3Las Cumbres Observatory Global Telescope, Goleta, CA 93117 4Space Telescope Science Institute, 3700 San Martin Drive, B al- timore, MD 21218 5Canadian Space Agency, 6767 Route de l’A´ eroport, Saint- Hubert, QC, J3Y 8Y9 Canada (present address) 6NASA Ames Research Center, MS 244-30, Moffett Field, CA 94035, USA 7SETI Institute/NASA Ames Research Center, MS244-30, Mof- fett Field, CA 94035, USAmodes approximately satisfy the asymptotic relation νnl≃∆ν0(n+l/2+ǫ)−l(l+1)D0 (1) (Vandakurov 1967; Tassoul 1980). Here νnlis the cyclic frequency, nis the radial order of the mode and lis the degree, l= 0 corresponding to radial (i.e., spher- ically symmetric) oscillations. Also, ∆ ν0is essentially the inverse sound travel time across the stellar diameter; this is closely related to the mean stellar density ∝angbracketleftρ∗∝angbracketright: ∆ν0∝ ∝angbracketleftρ∗∝angbracketright1/2.D0depends sensitively on conditions near the center of the star; for stars during the central hydrogenburningphasethisprovidesameasureofstellar age. Finally, ǫis determined by conditions near the stel- lar surface. This regular form of the frequency spectrum simplifies the analysis of the observations, and the close relation between the stellar properties and the param- eters characterizing the frequencies make them efficient diagnostics of the properties of the star. This has been demonstrated in the last few years through observations of solar-like oscillations from the ground and from space (for reviews, see Bedding & Kjeldsen 2008; Aerts et al. 2009; Gilliland et al. 2010a). Even observations allowing a determination of ∆ ν0 provide useful constraints on ∝angbracketleftρ∗∝angbracketright. With a reliable de- termination of individual frequencies ∝angbracketleftρ∗∝angbracketrightis tightly con- strained and an estimate of the stellar age can be ob- tained. This can greatly aid the interpretation of obser- vations of planetary transits (e.g., Gilliland et al. 2010b; Nutzman et al. 2010). We note that photometric obser- vations such as those carried out by Keplerare predom- inantly sensitive to modes of degree l= 0−2. As indi- catedbyEq.(1)thesearesufficienttoobtaininformation about the core properties of the star. Ground-based transit observations have identified three planetary systems in the Keplerfield: TrES-2 (O’Donovan et al. 2006; Sozzetti et al. 2007), HAT-P-7 (P´ al et al. 2008), and HAT-P-11 (Dittmann et al. 2009; Bakos et al. 2010). These systems have been observed byKeplerin SC mode. Their properties (cf. Table 1) indicate that they should display solar-like oscillations at observable amplitudes, and hence they are obvious targets for Keplerasteroseismology. Here we report the results of a preliminary asteroseismic characterization of2 Christensen-Dalsgaard et al. TABLE 1 Properties of transiting systems. Name KIC No Teff(K) [Fe/H] L/L⊙log(g) (cgs) vsiniSource (kms−1) HAT-P-7 10666592 6350 ±80 0.26±0.08 4 .9±1.1 4.07±0.06 3.8±0.5 (a) 6525±61 0.31±0.07 4 .09±0.08 (b) HAT-P-11 10748390 4780 ±50 0.31±0.05 0.26±0.02 4.59±0.03 1.5±1.5 (c) TrES-2 11446443 5850 ±50−0.15±0.10 1.17±0.10 4.4±0.1 2 ±1 (d) 5795±73 0.06±0.08 4 .30±0.13 (b) Note. — Sources: (a): P´ al et al. (2008); (b): Ammler-von Eif et al . (2009); (c): Bakos et al. (2010); (d): Sozzetti et al. (2007 ). In some cases asymmetric error bars have been symmetrized. the central stars in the systems, based on the early Ke- plerdata. 2.OBSERVATIONS AND DATA ANALYSIS We have analyzed data from Kepler for three planet-hosting stars using a pipeline developed for fast and robust analysis of all Keplerp-mode data (Christensen-Dalsgaard et al. 2008; Huber et al. 2009). Each time series contains 63324 data points. SC data characteristics and minor post-pipeline processing are discussed in Gilliland et al. (2010c). In addition a limb- darkened transit light curve model fit has been removed and 5-σclipping applied to remove outlying data points from each of the time series. The frequency analysis con- tains four main steps: 1. We calculate an oversampled (factor of four) ver- sion ofthe power spectrum by using a least-squares fitting. We smoothed the spectrum to 3 µHz reso- lution to remove the fine structure caused by the finite mode lifetime. 2. We correlated the smoothed power spectrum with an equally spaced comb of delta functions, sepa- ratedby∆ ν0/2,andconfinedtoaGaussian-shaped band with a full width at half maximum of 5∆ ν0. We adopted the maximum of this convolution over lags between 0 and 0.5 ∆ ν0as the filter output for each ∆ν0. 3. After identifying the peak correlation for the best matched model filter and extracting the large sep- aration corresponding to this peak we calculate the folded spectrum (see Fig. 1b), i.e., the sum of the power as a function of frequency modulo the opti- mumlargeseparation(theonecorrespondingtothe peak correlation). The summed power is used to locate the p-mode structure and identify the ridges corresponding to the different mode degrees (based on the asymptotic relation). 4. From the asymptoticrelationandthe identification of mode degrees we finally identify the position of the individual p-mode frequencies in the smoothed version of the power spectrum; when more than one mode is seen near the expected frequency we use the power-weighted average of the two peaks. Those extracted frequencies and the mode identifi- cations are used in the modeling. For observations with low signal-to-noise ratio it may not be possible to identify the individual frequencies. In 0 1 2 Fig. 1.— (a)PowerspectrumofHAT-P-7forfrequencies between 300 and 3000 µHz. The spectrum is smoothed with a gaussian filter with a FWHM of 3 µHz. The noise level at high frequencies corre- sponds to 1.1 ppm in amplitude. The white curve is a smoothed power spectrum with a gaussian filter (150 µHz FWHM). A fit to the background (dashed white curve) is also shown. The exces s power and the individual p-modes are evident. (b) Folded pow er spectrum, between 750 and 1500 µHz, for HAT-P-7 for a large sep- aration of 59 .22µHz. Indicated are the positions corresponding to radial modes ( l= 0) and non-radial modes with l= 1 and 2. The measured positions are used to identify the individual osci llation modes in panel (a). (c) ´Echelle diagram (see text) for frequencies of degree l= 0, 1, and 2 in HAT-P-7; a frequency separation of 59.36µHz and a starting frequency of 10 .8µHz were used. The filled symbols, coded for degree as indicated, show the obser ved frequencies, while the open symbols are for Model 3 in Table 2 , minimizing χ2 ν.3 such cases the analysis is carried through step 2, to de- termine the maximum response and hence an estimate of the large separation. Results on the three individual cases are presented in §4. 3.MODEL FITTING Stellar evolution models and adiabatic oscillation frequencies were computed using the Aarhus codes (Christensen-Dalsgaard 2008a,b), with the OPAL equation of state (Rogers et al. 1996) and opacity (Iglesias & Rogers 1996) and the NACRE nuclear reac- tion parameters (Angulo et al. 1999). In some cases (see below) diffusion and settling of helium were included, using the simplified formulation of Michaud & Proffitt (1993). Convection was treated with the B¨ ohm-Vitense (1958) mixing-length formulation, with a mixing length αML= 2.00 in units of the pressure scale height roughly corresponding to a solar calibration. In some models with convective cores, overshoot was included over a dis- tance of αovpressure scale heights. Evolution started from chemically homogeneous zero-age models. The ini- tial abundances by mass X0andZ0of hydrogen and heavy elements were characterized by the assumed value of [Fe/H], using as reference a present solar surface com- position with Zs/Xs= 0.0245 (Grevesse & Noels 1993) and assuming, from galactic chemical evolution, that X0= 0.7679−3Z0. From the observed ∆ ν0, effective temperature and composition an initial estimate of the stellar parame- ters was obtained using the grid-based SEEK pipeline (Quirion et al., in preparation). Smaller grids were then computed in the vicinity of these initial parameters, to obtaintighterconstraintsonstellarproperties. ForHAT- P-7 the analysis of the observations yielded frequencies of individually identified modes; here the analysis was based on χ2 ν=1 N−1/summationdisplay nl/parenleftBigg ν(obs) nl−ν(mod) nl σν/parenrightBigg2 ,(2) whereν(obs) nlandν(mod) nlare the observed and model fre- quencies, σνis the standard error in the observed fre- quencies (assumed to be constant) and Nis the num- ber of observed frequencies. In addition, we considered χ2=χ2 ν+χ2 T, whereχ2 Tis the corresponding normalized square difference between the observed and model effec- tive temperature. When χ2 νwas available we minimized it along each evolution track and considered the result- ing minimum values, and the corresponding value of χ2, as a function of the parameters characterizing the mod- els (see Gilliland et al. 2010b, for details). When only the large separation ∆ ν0could be determined from the observations, we identified the model along each track which matched ∆ ν0and considered the resulting χ2 Tas a function of the model parameters. 4.RESULTS 4.1.HAT-P-7 The observed power spectrum for HAT-P-7 is shown in Fig. 1a. The presence of solar-like p-mode peaks, with a maximum power around 1.1mHz, is evident. At high frequency the noise level in the amplitude spectrum is1.1 parts per million (ppm), with some increase at lower frequency, likely due to the effects of stellar granulation. Carrying out the correlation analysis described in §2 we determined the large separation as ∆ ν0= 59.22µHz. Figure 1b shows the resulting folded spectrum. This clearly shows two closely spaced peaks, identified as cor- responding to modes of degree l= 0 and 2, and single peak separated from these two by approximately ∆ ν0/2, corresponding to l= 1. On this basis we finally deter- mined the individual frequencies, identifying the modes from the asymptotic relation; the final set includes 33 p-mode frequencies, determined with a standard error σν= 1.4µHz. These frequencies, corresponding to ra- dial orders between 11 and 24, are illustrated in Fig. 1c in an ´ echelle diagram (see below). A grid of models was computed for masses between 1.41 and 1 .61M⊙, [Fe/H] between 0.17 and 0.38, and αov= 0,0.1 and 0.2, extending well beyond the end of central hydrogen burning. The modeling did not include diffusion and settling. At the mass of this star the outer convection zone is quite thin, and as a result the set- tling timescale is much shorter than the age of the star. Including settling, without compensating effects such as partial mixing in the radiative region or mass loss, leads to a rapid change in the surface composition which is inconsistent with the observed [Fe/H]; for simplicity we therefore neglected these effects for HAT-P-7.8 The computed frequencies were corrected according to the procedure of Kjeldsen et al. (2008) for errors in the modeling of the near-surface layers, by adding a(ν/ν0)b wherea= 0.1158µHz,ν0= 1000µHz andb= 4.9. As discussed in §3, for each evolution track, characterized by a set of model parameters, we minimized the depar- tureχ2 νof the model frequencies from the observations, defining the best model for this set. We first consider χ2 νas a function of the effective temperature of the models (Fig. 2a). It is evident that there is a clear minimum in χ2 ν; this is consistent with the determination of Teffby P´ al et al. (2008) but not with the somewhat higher temperature obtained by Ammler-von Eif et al. (2009) (see also Table 1). Thus in the following we use the observed quantities from P´ al et al. (2008). Since the frequencies to leading order are determined by the mean stellar density ∝angbracketleftρ∗∝angbracketright, Fig. 2b,c show χ2 νand χ2as functions of ∝angbracketleftρ∗∝angbracketright. It is evident that the best-fitting modelsoccupyanarrowrangeof ∝angbracketleftρ∗∝angbracketright, withawell-defined minimum. Fittingaparabolato χ2inpanel(c)weobtain the estimate ∝angbracketleftρ∗∝angbracketright= 0.2712±0.0032gcm−1. In Fig. 2d χ2is shown against model age. Here the variation with model parameters is substantially stronger, resulting in a greater spread in the inferred age; in particular, it is evident, not surprisingly, that the results depend on the extent of convective overshoot. From the figure we esti- matethattheageofHAT-P-7isbetween1.4and2.3Gyr. Examples of evolution tracks are shown in Fig. 3; pa- rameters for these models are provided in Table 2. They were chosen to give the smallest χ2 νfor each of the three values of αovconsidered. Also shown are the locations 8Artificially suppressing settling in the outer layers, whil e in- cluding diffusion and settling in the core, leads to results t hat are very similar to those presented here.4 Christensen-Dalsgaard et al. TABLE 2 Stellar evolution models fitting the observed frequencies for HAT-P-7. No M ∗/M⊙Age Z0X0αovR∗/R⊙/angbracketleftρ∗/angbracketrightTeffL∗/L⊙χ2 νχ2 (Gyr) (gcm−3) (K) 1 1.53 1.758 0.0270 0.6870 0.0 1.994 0.2718 6379 5.91 1.08 1.2 1 2 1.52 1.875 0.0290 0.6809 0.1 1.992 0.2708 6355 5.81 1.04 1.0 4 3 1.50 2.009 0.0270 0.6870 0.2 1.981 0.2718 6389 5.87 1.00 1.2 4 Note. — Models minimizing χ2 ν(cf. Eq. 2) along the evolution tracks, illustrated in Fig. 3 . The models have been selected as providing the smallest χ2 νfor each of the three values of the overshoot parameter αov. The smallest value of χ2 νis obtained for Model 3. Fig. 2.— Results of fitting the observed frequencies to a grid of stellar models (see text for details). Plusses, stars and di amonds correspond to modelswith αov= 0 (no overshoot), 0.1, and 0.2. (a) Minimum mean square deviation χ2 νof the frequencies (cf. Eq. 2) along each evolution track, against the effective temperatu reTeff of the corresponding models. The vertical dashed and dotted lines indicate the effective temperatures found by P´ al et al. (200 8) and Ammler-von Eif et al. (2009). (b) Minimum mean square devia- tionχ2 νagainst the mean density /angbracketleftρ∗/angbracketrightof the corresponding models. (c) As (b), but showing the combined χ2. (d)χ2against the age for the models that minimize χ2 ν; the different ridges correspond to the different masses in the grid, the more massive models resu lting in a lower estimate of the age.Fig. 3.— Theoretical HR diagram with selected evolutionary tracks, corresponding to the models defined in Table 2. The ’+ ’ in- dicate the models along the full set of evolutionary sequenc es mini- mizing the difference between the computed and observed freq uen- cies. The box is centered on the LandTeffas given by P´ al et al. (2008), with a size matching the errors on these quantities. of the models minimizing χ2 νalong each of the computed tracks; these evidently fall close to a line in the HR di- agram, corresponding to the small range in ∝angbracketleftρ∗∝angbracketright. The range of luminosities, from P´ al et al. (2008), is based on modeling and hence has not been used in our fit; even so, it is gratifying that the present models are essentially consistentwiththesevalues. Also,asindicatedbyFig.2a andTable 2, the best-fitting models areclose to the value ofTeffobtained by P´ al et al. (2008). The match of the best-fitting model (Model 3 of Ta- ble 2) to the observed frequencies is illustrated in a so- called´ echelle diagram (Grec et al. 1983) in Fig. 1c. In accordance with Eq. (1) the frequency spectrum is di- vided into slices of length ∆ ν, starting at a frequency of 10.8µHz; the figure shows the location of the observed (filled symbols) and computed (open symbols) frequen- cies within each slice, against the starting frequency of the slice; the model results extend to the acoustical cut- off frequency, 1930 µHz, of the model. There is clearly a very good overall agreement between model and obser- vations, including the detailed variation with frequency which reflects the frequency dependence of the large sep- aration, as a possible diagnostics of the outer layers of the star (e.g., Houdek & Gough 2007). We have finally made a fit of the inferred ∝angbracketleftρ∗∝angbracketright, as well asTeffand [Fe/H] from P´ al et al. (2008), to com- puted evolutionary tracks from the Yonsei-Yale compi- lation (Yi et al. 2001). This was based on a Markov Chain Monte Carlo analysis to obtain the statistical properties of the inferred quantities (see Brown 2010, for details). This resulted in M= 1.520±0.036M⊙, R= 1.991±0.018R⊙and an age of 2 .14±0.26Gyr. We note that the age estimate reflects the specific as-5 sumptions in the Yonsei-Yale evolution calculations; as indicated by Fig. 2d the true uncertainty in the age de- termination is likely somewhat larger. 4.2.HAT-P-11 For HAT-P-11 the oscillation amplitudes were much smaller than in HAT-P-7, as expected from the general scaling of amplitudes with stellar mass and luminosity (e.g., Kjeldsen & Bedding 1995). Thus with the present short run of data it has only been possible to determine thelargeseparation∆ ν0= 180.1µHzfromthemaximum in the correlation analysis. We have matched this to a grid of models, including diffusion and settling of helium, with masses between 0.7 and 0 .9M⊙and [Fe/H] between 0.21 and 0.41. These models provide a good fit to the observed TeffandL/L⊙; note that in the presentcase the luminosity is based on a reasonably well-determined par- allax. We havedetermined an estimateof ∝angbracketleftρ∗∝angbracketrightbyaverag- ing the results of those models which match the observed ∆ν0and lie within 2 standard deviations ( ±100K) from the value of Teffprovided by Bakos et al. (2010); the re- sult is∝angbracketleftρ∗∝angbracketright= 2.5127±0.0009gcm−3. Although the for- mal error is extremely small, owing to a tight relation between the large separation and the mean density for stars in this region in the HR diagram, the true error is undoubtedly substantially larger. In particular, we ne- glected the error in the determination of ∆ ν0and these data have not allowed a correction for the systematic errors in the modeling of the near-surface layers of the star. 4.3.TrES-2 Here also we were unable to determine individual fre- quencies from the present set of data. The expected am- plitudes are smaller than for HAT-P-7, and the noise level higher due to the fainter magnitude of TrES-2. The correlation analysis yielded two possible values of ∆ν0: 97.7µHz and 130 .7µHz. For this star ∝angbracketleftρ∗∝angbracketrighthas been determined from the analysis of the transit light curve. Sozzetti et al. (2007) obtained ∝angbracketleftρ∗∝angbracketright= 1.375± 0.065gcm−3, while Southworth (2009) found ∝angbracketleftρ∗∝angbracketright= 1.42±0.13gcm−3. From the scaling with ∝angbracketleftρ∗∝angbracketright1/2thesmaller of the two possible values of ∆ ν0is clearly incon- sistent with these values of ∝angbracketleftρ∗∝angbracketright, while ∆ ν0= 130.7µHz yields models that are consistent with the observed Teff and log(g) of Sozzetti et al. (2007) as well as with these values of the mean density. Here we considered a grid of models with helium diffusion and settling, masses be- tween 0.85 and 1 .1M⊙and [Fe/H] between −0.25 and −0.05. Determining again the mean value of ∝angbracketleftρ∗∝angbracketrightfor those models that matched ∆ ν0and had Teffwithin two standard deviations of the value of Sozzetti et al. (2007) we obtained ∝angbracketleftρ∗∝angbracketright= 1.3233±0.0027gcm−3. As in the case of HAT-P-11 the true error is likely substantially higher. 5.DISCUSSION AND CONCLUSION The present preliminary analysis provides a striking demonstration of the potential of Keplerasteroseismol- ogyanditssupportingroleintheanalysisofplanethosts. Thesestarswill undoubtedly be observedthroughout the mission and hence the quality of the data will increase substantially. For HAT-P-7 the detected frequencies are already close to what will be required for a detailed anal- ysis of the stellar interior, beyond the determination of the basic parameters of the star. Thus here we can look forward to a test of the assumptions of the stellar mod- eling; the resulting improvements will further constrain the overall properties of the star, in particular its age. Also, given the observed vsiniwe expect a rotational splitting comparable to that observed in the Sun, and hence likely detectable with a few months of observa- tions. For the other two stars there is strong evidence for the presence of solar-like oscillations; thus continued observations will very likely result in the determination of individual frequencies and hence further constraints on the properties of the stars. Funding for this Discovery mission is provided by NASA’s Science Mission Directorate. We are very grate- ful to the entire Keplerteam, whose efforts have led to this exceptional mission. The present work was sup- ported by the Danish Natural Science Research Council. Facilities: The Kepler Mission REFERENCES Aerts, C., Christensen-Dalsgaard, J., & Kurtz, D. W. 2009, Asteroseismology, Springer, Heidelberg, in the press Ammler-von Eif, M., Santos, N. C., Sousa, S. G., Fernandes, J ., Guillot, T., Israelian, G., Mayor, M., & Melo, C. 2009, A&A, 507, 523 Angulo, C., et al. 1999, Nucl. Phys. A, 656, 3 Bakos, G. ´A., et al. 2010, ApJ, in the press (arXiv:0901.0282v2) Bedding, T. R., & Kjeldsen, H. 2008, in Proc. 14thCambridge Workshop on Cool Stars, Stellar Systems, and the Sun, ed. G. T. van Belle, ASP Conf. Ser., San Francisco, 384, 21 B¨ ohm-Vitense, E. 1958, ZAp, 46, 108 Borucki, W., et al. 2009, in Proc. IAU Symp. 253, Transiting Planets, eds F. Pont, D. Sasselov & M. Holman, IAU and Cambridge University Press, 289 Borucki, W., et al. 2010, Science, submitted Brown, T. M. 2010, ApJ, in the press (arXiv:0912:1639) Christensen-Dalsgaard, J. 2002, Rev. Mod. Phys., 74, 1073 Christensen-Dalsgaard, J. 2008a, Ap&SS, 316, 13 Christensen-Dalsgaard, J. 2008b, Ap&SS, 316, 113Christensen-Dalsgaard, J., Arentoft, T., Brown, T. M., Gil liland, R. L., Kjeldsen, H., Borucki, W. J., & Koch, D. 2008, in Proc. HELAS II International Conference: Helioseismology, Asteroseismology and the MHD Connections, eds L. Gizon & M. Roth, J. Phys.: Conf. Ser., 118, 012039 Dittmann, J. A., Close, L. M., Green, E. M., Scuderi, L. J., & Males, J. R. 2009, ApJ, 699, L48 Gilliland, R. L., et al. 2010a, PASP, in the press Gilliland, R. L., McCullough, P. R., Nelan, E. P., Brown, T. M ., Charbonneau, D., Nutzman, P., Christensen-Dalsgaard, J., & Kjeldsen, H. 2010b, ApJ, submitted Gilliland, R. L., et al. 2010c, ApJL, in the press Grec, G., Fossat, E., & Pomerantz, M. 1983, Sol. Phys., 82, 55 Grevesse, N., & Noels, A. 1993, in Origin and evolution of the Elements, eds N. Prantzos, E. Vangioni-Flam & M. Cass´ e (Cambridge: Cambridge Univ. Press), 15 Houdek, G., & Gough, D. O. 2007, MNRAS, 375, 861 Huber, D., Stello, D., Bedding, T. R., Chaplin, W. J., Arento ft, T., Quirion, P.-O., & Kjeldsen, H. 2009, Comm. in Asteroseismology, 160, 74 Iglesias, C. A., & Rogers, F. J. 1996, ApJ, 464, 943 Kjeldsen, H., & Bedding, T. R. 1995, A&A, 293, 876 Christensen-Dalsgaard et al. Kjeldsen, H., Bedding, T. R., & Christensen-Dalsgaard, J. 2 008, ApJ, 683, L175 Kjeldsen, H., Bedding, T. R., & Christensen-Dalsgaard, J. 2 009, in Proc. IAU Symp. 253, Transiting Planets, eds F. Pont, D. Sasselov & M. Holman, IAU and Cambridge University Press, 309 Koch, D., et al. 2010, ApJL, submitted Michaud, G., & Proffitt, C. R. 1993, in Proc. IAU Colloq. 137: Inside the stars, eds A. Baglin, & W. W. Weiss, ASP Conf. Ser., San Francisco, 40, 246 Nutzman, P., Gilliland, R. L., McCullough, P. R., Charbonne au, D., Christensen-Dalsgaard, J.,Kjeldsen, H., Nelan, E. P., Brown, T. M., & Holman, M. J. 2010, ApJ, submitted O’Donovan, F. T., et al. 2006, ApJ, 651, L61P´ al, A., et al. 2008, ApJ, 680, 1450 Rogers, F. J., Swenson, F. J., & Iglesias, C. A. 1996, ApJ, 456 , 902 Southworth, J. 2009, MNRAS, 394, 272 Sozzetti, A., Torris, G., Charbonneau, D., Latham, D. W., Holman, M. J., Winn, J. N., Laird, J. B., & O’Donovan, F. T. 2007, ApJ, 664, 1190 Tassoul, M. 1980, ApJS, 43, 469 Yi, S., Demarque, P., Kim, Y.-C., Lee, Y.-W., Ree, C. H., Lejeune, T., & Barnes, S. 2001, ApJS, 136, 417 Vandakurov, Yu. V. 1967, AZh, 44, 786 (English translation: Soviet Ast., 11, 630)