KEPLER TELESCOPE AND SOLAR-LIKE PULSATORS

KEPLER TELESCOPE AND SOLAR-LIKE PULSATORS J. Molenda Żakowicz 1, T. Arentoft 2,3, H. Kjeldsen 2,3, and A. Bonanno 4 1 Instytut Astronomiczny Uniwersytetu Wrocławskiego, Kopernika 11, 51-622 Wrocław, Poland 2 Institute of Physisc and Astronomy, University of Aarhus, Bldg. 1520, DK-8000 Aarhus, Denmark 3 Danish AsteroSeismology Centre (DASC), University of Aarhus, Bldg. 1520, DK-8000 Aarhus, Denmark 4 Osservatorio Astrofisico di Catania, Via S.Sofia 78, 95123 Catania, Italy ABSTRACT 25 We present 104 stars selected for a long-time monitoring for Kepler, a new NASA space telescope. 30 of these stars are primary asteroseismological targets for Kepler, the remaining 74, secondary asteroseismological targets for the telescope. They are mostly F K stars that we expect to show solar like oscillations. We determine fundamental astrophysical parameters and estimate pulsational properties of the selected targets. Then, we use simulations to present Kepler s ability of detection oscillations with frequencies and amplitudes typical for solar-like pulsators. Key words: Stars: variables: solar like; Stars: variables: other; Space missions: Kepler. N 20 15 10 5 0 B A F G K M Figure 1. Histogram of spectral types for 91 of 104 Kepler asteroseismic targets that have spectral classification. 2. TARGET SELECTION 1. INTRODUCTION Kepler is a new NASA space telescope designed for detection of Earth size and larger planets with the method of photometric transits (Borucki et al. [6]). Kepler instrument is a 0.95 m Schmidt telescope with a 105 deg 2 field of view centered in the Cygnus Lyra region at α 2000 = 19 h 22 m 40 s, δ 2000 = +44 o 30. The telescope will be launched in October 2008. The mission is planned to last for four years and then continue for another two years. For the entire life time of the mission, Kepler will perform continuous and simultaneous observations of more than 200 000 stars (according to the estimations of the Kepler team) that fall into Kepler field of view and into its nominal magnitude range that is V = 9 14 mag. We present 104 most promising targets for an asteroseismic study that can be performed with the use of the Kepler telescope. Our list includes stars that may show variability of different type including SPB, δ Scuti, ro Ap, γ Doradus and solar like pulsators. We focus on the last type of pulsators and we discuss the Kepler mission in the context of discovery of new variables of this kind. In the process of selection the most promising asteroseismic targets for Kepler, we focused on stars for which fundamental astrophysical parameters, i.e., luminosity, effective temperature, gravity, metalicity, etc., can be determined and precise pulsational models computed. We searched the Hipparcos Catalogue (ESA [16]) and selected 104 stars from the magnitude range V = 9 11 mag. We selected the brightest stars from the Kepler nominal magnitude range because they will have the highest signal to noise, S/N, ratio in the Kepler photometry. We used the pixel calculator C code kindly provided by D. Koch to check the position of these stars on Kepler CCD panels. We found 87 stars to fall onto active CCD panels and 17, into the star tracker corner. In Fig. 2, we show the finding chart with Kepler asteroseismic targets (dots) and other stars brighter than V = 6 mag (circles). Due to the clarity of the plot, we do not show stars from the magnitude range V = 7 8 mag. The total selected sample contains five stars classified as B, 22, as A, 19, as F, 16, as G, 24, as K and five, as M. The remaining 13 stars do not have spectral classification. In Fig. 1, we show a histogram of spectral types of 91 of the selected targets that have assiged a spectral type. Proceedings SOHO 18 / GONG 2006 / HELAS I, Sheffield, UK 7-11 August 2006 (ESA SP-624, October 2006)

δ 2000 [deg] 55 50 45 40 35 96600 96061 96343 95506 96776 97168 96699 94565 93943 97219 95637 93577 96146 94743 93945 97513 95495 93011 97671 95092 92247 92053 97247 92637 97657 96277 93755 91841 94701 94438 92259 98486 92941 97486 96532 95631 94976 94335 93607 98655 96538 94967 94292 95843 93879 92961 97576 96735 93320 91128 94145 93522 92132 98797 98793 98160 96634 95098 97582 94918 91178 99267 96846 94670 98551 95638 97439 94898 94409 92775 97974 98829 98381 96642 95237 94809 93941 98814 96210 95032 92922 99336 96751 95174 98037 94952 97337 95859 94734 97254 96762 95913 94497 97724 95733 95069 93687 96074 93070 96299 93469 95654 95024 94022 93924 94704 94137 95184 300 295 290 α 2000 [deg] 285 280 Figure 2. The finding chart for 104 Kepler asteroseismic targets. Dots, V = 9 11 mag Kepler asteroseismic targets labeled with Hipparcos HIP numbers, circles, other stars brighter than V = 6 mag.

3. ASTROPHYSICAL PARAMETERS We derived effective temperatures of the 104 stars using either Strömgren uvbyβ, Johnson UBV or 2 MASS JHK photometry (Cutri et al. [13]). For HIP 92637, 94145, 96146, 92637 and 97439, we used Strömgren photometry (Hauck & Mermilliod [19], Arellano et al. [4]) and the calibration of Napiwotzki et al. [27]. For the other hot stars, namely, HIP 93924, 95069, 95174, 96210, 96762, 97254, 98037, 98160, 98551 and 98814, that do not have Strömgren photometry, we used (B V ) indices and the calibration of Johnson [20]. For HIP 91178, 92247, 92259, 93070, 93522, 93943, 94137, 94670, 95092, 95495, 96061, 96343, 96776, 97486, 97582, 97724 and 98486, we used JHK photometry and the calibration of Kinman & Castelli [21]. For the remaining stars we used JHK photometry and the calibration of Alonso et al. [1] or Alonso et al. [2]. Finally, for both components of HIP 94335, a eclipsing binary of Algol type, we adopted the log T eff values listed by Ribas et al. [30]. Computing T eff from the metalicity dependent calibrations of Kinman & Castelli [21], Alonso et al. [1] or Alonso et al. [2], we used the values of metalicity published by Pilachowski et al. [28], Reid et al. [29], Carney et al. [10], Feltzing & Gustafsson [17] and Zhang & Zhao [33] for HIP 92775, 94704, 95184, 97219 and 99267, respectively. For the remaining stars we used the solar value. Then, we used the photometric errors to compute the standard deviations of log T eff. The computed values increase with log T eff that is partly due the the internal precision of the applied calibration, partly to the increase of photometric errors. For FGK stars for which we used 2MASS photometry, as well as for early type stars for which we used Strömgren photometry, the average error in log T eff is equal to 0.002. For A type stars for which we used 2MASS photometry, the average error in log T eff is equal to 0.006. The highest errors occur for B type stars for which we used (B V ) indices from the Hipparcos Catalogue. The typical errors of this index is equal to 0.03 mag. This results in a high uncertainty in the effective temperature, that is on the average equal to 0.010. The average scatter on the effective temperature, however, is low since it does not include systematic errors of the calibration that are in general larger that the internal scatter. The main source of such systematic errors is E(B V ) that is not directly measured but either calculated from the formula of di Benedetto [15] or derived from the dust maps of Schlegel et al. [31]. Therefore, the possible systematic errors of this parameter can influence the effective temperature and luminosity. Another source of systematic errors is the metalicity that has been measured for most of the studied stars and therefore assumed to be equal to the solar value. 30 of the 104 targets, mainly F K stars, have the ratio of σ π /π lower than 0.175 (see the Hipparcos Catalogue, Table 1. Effective temperatures and luminosities of 30 Kepler primary asteroseismic targets. The standard deviation of T eff is expressed as σ 10 3. HIP SpT V 0 log T eff log L/L ±σ ±σ 91128 M0 9.83 3.579±2-1.057± 0.050 92922 G5 9.03 3.751±2 0.355± 0.034 93011 F2 9.47 3.840±2 0.863± 0.098 94145 F0III 8.96 3.897±4 1.020± 0.104 94335A F8V 9.56 3.786±2 0.373± 0.047 94335B G8V 11.02 3.720±2-0.139± 0.051 94497 K2 9.82 3.693±2-0.448± 0.035 94565 G0 9.20 3.797±2 0.790± 0.087 94704 K2 11.01 3.735±2-0.452± 0.085 94734 9.39 3.768±1 0.470± 0.072 94743 F5 9.03 3.850±2 0.645± 0.055 94898 G5 9.43 3.730±2-0.123± 0.031 95098 F8 9.33 3.797±2 0.730± 0.080 95631 G5 9.04 3.725±2 0.043± 0.031 95637 F0 8.87 3.866±2 1.114± 0.137 95638 G5 10.47 3.752±2 0.020± 0.097 95733 10.93 3.726±2-0.177± 0.136 95843 F2 9.13 3.820±2 0.577± 0.050 96146 F0 8.96 3.823±2 0.700± 0.048 96634 K0 9.11 3.712±1-0.343± 0.032 96735 K0 9.15 3.697±1-0.503± 0.034 97168 10.27 3.761±2-0.118± 0.051 97219 G5 8.96 3.718±1-0.263± 0.031 97337 M0 10.98 3.612±3-0.945± 0.047 97657 K2 9.43 3.675±1-0.570± 0.037 97974 G0 9.91 3.757±2 0.051± 0.044 98381 K2 9.88 3.654±2-0.585± 0.040 98655 K0 10.33 3.719±2 0.085± 0.097 98829 G5 9.62 3.742±2-0.195± 0.031 99267 F3 9.78 3.778±2-0.199± 0.034 ESA [16]). In Fig. 3, we show the histogram of spectral types of 27 of these stars. We computed log L/L for these stars using the formula of Smith [32] that includes the Lutz Kelker bias (Lutz & Kelker [24]) in computations of M V. Computing the bolometric magnitudes, M bol, we used the values of bolometric corrections, BC, listed by Flower [18] and the solar bolometric magnitude, M bol, equal to 4.74 mag. We calculated the standard deviation of log L/L using the photometric errors, the parallax errors and the uncertainty of BC. We list these 30 stars, hereafter Kepler primary targets, their spectral types, de reddened V magnitudes, and computed log T eff and log L/L in Table 1. We plot them on the log T eff log L/L diagram in Fig. 4. The results obtained for the remaining 74 stars, hereafter secondary targets, we list in Table 2. 4. SOLAR-LIKE VARIABILITY IN KEPLER PHOTOMETRY The sample presented in the previous sections is perfect for an extended study of solar like oscillations in stars of

Table 2. Effective temperatures of 74 Kepler secondary asteroseismic targets. The standard deviation of T eff is expressed as σ 10 3. HIP SpT V 0 log T eff HIP SpT V 0 log T eff HIP SpT V 0 log T eff ±σ ±σ ±σ 91178 A0 8.93 4.006±7 94292 G8V 9.82 3.708±2 96538 10.96 3.862±2 91841 K0 9.30 3.698±2 94409 M0 9.18 3.565±1 96600 11.15 3.826±2 92053 M0 9.46 3.559±2 94670 B9 9.07 4.013±8 96642 F5 9.35 3.852±2 92132 G5 9.31 3.699±2 94809 F 9.12 3.890±2 96699 K0 8.95 3.729±2 92247 A2 9.08 3.917±3 94918 G2V 9.39 3.772±4 96751 K2 8.53 3.694±1 92259 A5 9.30 3.906±5 94952 K5 8.82 3.614±1 96762 A2 8.55 4.112±14 92637 B4V: 9.66 4.309±1 94967 G8V: 9.59 3.725±3 96776 A3 9.09 3.940±4 92775 G2 9.72 3.755±1 94976 K0V 9.57 3.720±3 96846 K2 8.56 3.697±2 92941 K0 9.22 3.694±3 95024 C 10.90 3.832±2 97247 K5 9.45 3.607±1 92961 F8 9.17 3.796±2 95069 A2 8.56 4.034±8 97254 A2 8.35 4.066±9 93070 A2 9.08 3.956±5 95092 A0 9.63 3.980±9 97439 G2 8.41 3.752±2 93320 K3 11.13 3.693±2 95174 A 8.84 4.022±4 97486 A0 9.54 4.021±8 93469 G0 10.14 3.829±3 95184 11.05 3.762±1 97513 K0 8.61 3.637±1 93522 A3 9.99 3.994±10 95237 G5 8.71 3.700±3 97576 K2 9.28 3.699±2 93577 F0 8.86 3.871±2 95495 A2 9.04 3.917±3 97582 F0 8.51 3.960±5 93607 F5 8.95 3.838±2 95506 A0 8.74 4.055±6 97671 K0 8.99 3.648±1 93687 K0 8.67 3.718±2 95654 F5 8.85 3.849±2 97724 A0 8.54 3.985±6 93755 K0 9.73 3.660±1 95859 F8 9.00 3.708±4 98037 A2 8.53 4.137±19 93879 F8 9.79 3.802±2 95913 K0 9.03 3.719±2 98160 B9V 8.65 4.016±3 93924 A0 8.72 4.072±9 96061 9.34 4.010±7 98486 A2 9.08 4.021±8 93941 10.49 4.201±14 96210 B8 8.82 4.203±20 98551 A0 9.79 4.008±10 93943 A5 9.54 3.912±5 96277 F0 9.66 3.907±2 98793 11.07 3.719±2 93945 11.57 3.634±1 96299 F5 8.90 3.885±2 98797 10.78 3.942±9 94022 K0 10.39 3.578±1 96343 A0 9.32 3.997±7 98814 B8V 9.53 4.004±7 94137 A5 9.45 3.909±3 96532 11.07 3.893±2 +G2III 10 1.5 1 8 0.5 N 6 4 log L/L sun 0-0.5 2-1 0 F G K M Figure 3. Histogram of spectral types for 27 of 30 Kepler primary asteroseismic targets that have spectral classification. -1.5 3.9 3.87 3.84 3.81 3.78 3.75 3.72 log T eff Figure 4. log T eff log L/L diagram for 30 Kepler primary asteroseismic targets. 3.69 3.66 3.63 3.6 3.57

different evolutionary status and different astrophysical parameters. With the use of asteroseismology, the interior of these stars can be examined (Christensen Dalsgaard [12]) allowing detailed tests of pulsational and evolutionary models. The solar like oscillations are driven stochastically by convection in the outer layers of the stellar atmosphere. They are characterized by high regularity. At least for low-degree modes, the p-mode frequencies of radial order n and angular degree l, follow approximately the asymptotic relation, (see, e.g., Bedding & Kjeldsen [5]): ν n,l = ν(n + 1 2 l + ǫ) l(l + 1)D 0. (1) Modes of same l but adjacent radial order, n, are separated by the large separation, ν, while modes of the same n but adjacent l are separated by 1 2 ν. The large separation is sensitive to the mean density of the star, while ǫ is sensitive to conditions at the surface, and D 0 to conditions near the center. The advantage of this regularity is that information about the star (the mean density) can be derived even from low signal-to-noise observations, e.g., by autocorrelation analysis of the amplitude spectrum, using all modes simultaneously to derive the large separation. These stars, however, show very tiny amplitudes which can be detected either in precise Doppler measurements or in space photometry (see Kjeldsen & Bedding [23]). Kepler, although it is a planet finding mission, will provide photometric measurements that will be suitable for observing pulsating stars, including so low amplitude variables as solar-like pulsators. We used the scaling relations from Brown et al. [8] and Kjeldsen & Bedding [22] to compute the acoustic cut off frequencies and amplitudes of solar like oscillations that can be estimated as functions of mass, luminosity and effective temperature: ν ac = M/M (R/R ) 2 5.5 mhz (2) T eff /5777 dl L = L/L (M/M ) (T eff /5777) 2 4.7 ppm (3). and we compared those values to the time sampling and noise levels of Kepler. We compared the Nyquist frequencies of Kepler (8333 and 278µHz, for the 1 and 30-min cadence planned for the Kepler mission, respectively) with the upper frequency limit of solar-like oscillations, i.e., the acoustic cut off frequency, expected for stars across the HR-diagram. We based our computations on models of Christensen Dalsgaard [11]. We show our results in Fig. 5, where we use dashed and dotted lines to indicate stars have periods sufficiently long to be resolved with the 1 or 30 minute cadence, respectively. We show that with respect to time resolution, in most stars Figure 5. Comparison between the Nyquist frequency and the acoustic cut off frequencies of stars across the HRdiagram, for observations at cadences of one data point per minute and per 30 minutes, respectively. oscillations can be observed with the 1-min cadence. The 30-min cadence may be used only for evolved stars. In a similar way we compared expected amplitudes of solar like oscillation to the Kepler noise level for stars across the HR-diagram. Requiring a signal-to-noise of 3 after one month of observations we obtained the plot shown in Fig. 6. The different lines show the smallest L/M-values for which the 3σ criterion is fulfilled. This plot concerns stars of magnitudes V = 10 to V = 14 and was obtained with the use of the noise levels stated in the caption of the Figure. We found that for a V = 11, all stars more massive or more evolved than the Sun, will have amplitudes sufficiently high to be detected at the 3σ level after one month of observations at the 1-min cadence. After 3 months, solar amplitudes can be detected in stars slightly fainter than V = 12. We stress, however, that these estimations do not take into account the finite life time of the modes and also that results on Procyon and on stars in M67 indicate that the L/M scaling relation for the amplitudes overestimates the amplitudes for F type stars (e.g., Bedding & Kjeldsen [5]). In order to assess what can actually be obtained from Kepler observations of solar like stars, we preformed series of simulations of the stochastically excited modes carried out with the aim of estimating the precision to which stellar radii can be determined from seismic data. The simulations were done using algorithms described in De Ridder et al. [14]. The idea is that stellar radii can be obtained from the large frequency separation of solar like oscillations, as it depends on the mean sound speed of the star. This possibility is of potentially great importance for characterizing planet-hosting stars detected with Kepler.

Figure 6. Lines of S/N>3 for stars across the HRdiagram at distances corresponding to V = 10 to V = 14, using the scaling relation for estimating amplitudes and assuming noise-levels in the amplitude spectrum of 0.96, 1.5, 2.41, 3.83 and 6.06 ppm in one month for V = 10 14, respectively. Figure 7. Amplitude spectra for two of the simulated stars (simulations #8,10), after 30 and 90 days of observations at the one-minute cadence, respectively. Ordinate scales differ from panel to panel. We simulated time series of 10 stars of different brightness and pulsational periods, amplitudes and mode lifetimes. Noise levels were scaled according to Table 1 in Borucki et al. [7], assuming a white-noise level of 71 ppm/min in the time series of a V = 9 star (0.61 ppm per month in the amplitude spectrum). We show amplitude spectra for two of the simulated stars in Fig. 7. We found that for all 10 simulations, the large separation could readily be obtained, even in cases of low signal to noise in the amplitude spectra (cf. Fig. 7, upper panels). Autocorrelations for the two stars in Fig. 7 are shown in Fig. 8. As above, input was the part of the amplitude spectrum where the signal is (about 2 4 mhz depending on the specific simulation), including only peaks above 1.5σ (the noise at high frequency). The autocorrelations show clear, regular spaced peaks at integer and half-integer values of the large separation as expected from the asymptotic relation. For all 10 stars, subsequent comparison with the input frequencies showed that the large separations derived from the data were correct. We estimated also the precision of derivation of the large separations from our simulated data. We found that the separations extracted from the data of different chunks of months are consistent within a few tens of a µhz. We found that the large separations found in the individual bins typically fall within 0.2µHz with the scatter equal to 0.04µHz. Therefore, we conclude that on the average the large separation can be determined to an estimated precision of better than at least 0.5µHz. 5. SUMMARY We presented a list of 30 primary and 74 secondary asteroseismic targets for the Kepler space telescope. We computed effective temperatures for all the targets and also luminosities, for the primary ones. In most cases, log T eff computed by us is the first determination of this parameter. The obtained values slightly differ from the preliminary log T eff, log L/L listed in Molenda Żakowicz [26] because of more precise de-reddening procedure and slightly different approach to estimation E(B V ). For stars with effective temperatures derived also by other authors, i.e., HIP 96146, 97219 and 99267 Masana et al. [25], HIP 92775 Pilachowski et al. [28], HIP 95184 Carney et al. [10], HIP 98814 Budding [9] and 18 other stars from our sample listed by Ammons et al. [3], our values agree well with those from the literature. We showed that both the Kepler 1 min (high) cadence and the 30-min (low) cadence modes can be used for detection of solar like oscillations and for precise determination of seismic parameters in a wide range of HR diagram. Finally, we show that this detection, i.e., discovery of new solar like pulsators, can be made in a short time after the start of the mission.

Figure 8. Autocorrelations for the data shown in Fig. 7. The dotted lines show the large separations found for the two simulations. Ordinate scales differ from panel to panel. REFERENCES [1] Alonso, A., Arribas, S., Martínez-Roger, C. 1996, A&A 313, 873 [2] Alonso, A., Arribas, S., Martínez-Roger, C. 1999, A&AS 140, 261 [3] Ammons, S.M., Robinson, S.E., Strader, J., Laughlin, G., Fischer, D., Wolf, A. 2006, ApJ 638, 1004 [4] Arellano F.A., Rojo A.E., Gonzalez Bedolla, S., Rosenzweig, P. 1998, ApJS 117, 167 [5] Bedding T.R., Kjeldsen H. 2003, PASA 20, 203 [6] Borucki, W.J., Koch, D.G., Dunham, E.W., Jenkins, J.M. 1997, in Planets Beyond the Solar System and the Next Generation of Space Missions ASP Conf. Ser. vol. 119, p.153, ed. David Soderblom [7] Borucki et al. 2003, in Future EUV/UV and Visible Space Astrophysics Missions and Instrumentation. Edited by J. Chris Blades, Oswald H. W. Siegmund. Proceedings of the SPIE, Volume 4854, p. 129 [8] Brown, T.M., Gilliland, R.L., Noyes, R.W., Ramsey, L.W. 1991, ApJ 368, 599 [9] Budding, E. 1985, PASP 97, 584 [10] Carney, B.W., Latham, D.W., Laird, J.B., Aguilar, L.A. 1994, AJ 107, 2240 [11] Christensen-Dalsgaard, J. 1993, Asp Conf. Ser. Vol. 40, Proceedings of IAU Coll. 137: Inside the Stars, eds. W. W. Weiss and A. Baglin, 483 [12] Christensen Dalsgaard, J. 2004, Solar Physics, 220. 137 [13] Cutri, R.M. et al. 2003, 2 MASS All Sky Catalog of Point Sources [14] De Ridder, J., Arentoft, T., Kjeldsen, H. 2006, MN- RAS 365, 595 [15] di Benedetto, G.,P. 1998, A&A 339, 858 [16] ESA 1997, The Hipparcos Catalogue, ESA SP- 1200 [17] Feltzing, S., Gustafsson, B. 1998, A&AS 129, 237 [18] Flower, P.J. 1996, ApJ 469, 355 [19] Hauck, B., Mermilliod, M. 1990, A&AS 86, 107 [20] Johnson H.L. 1966, ARA&A 4, 193 [21] Kinman, T., Castelli, F. 2002, A&A 391, 1039 [22] Kjeldsen, H., Bedding, T.R. 1995, A&A 293, 87 [23] Kjeldsen, H., Bedding, T.R. 2004, Proc. of the SOHO 14 / GONG 2004 Workshop (ESA SP-559). Helio- and Asteroseismology: Towards a Golden Future p.101, Ed.: D. Danesy., [24] Lutz, T.E., Kelker, D.H. 1973, PASP 85, 573 [25] Masana, E., Jordi, C. Ribas, I. 2006, A&A 450, 735 [26] Molenda Żakowicz J. 2006, Mem. Soc. Astron. It. vol. 77, p. 545 [27] Napiwotzki, R., Schönberner, D., Wenske, V. 1993, A&A 268, 653 [28] Pilachowski, C.A., Sneden, C., Booth, J. 1993, ApJ 407, 699 [29] Reid, I.N., Wyk, F.V., Marang, F., Roberts, G., Kilkenny, D., Mahoney, S. 2001, MNRAS 325, 931 [30] Ribas, I., Gimenez, A., Torra, J., Jordi, C., Oblak, E. 1998, A&A 330, 600 [31] Schlegel, D., Finkbeiner, D., Davis, M. 1998, ApJ, 500, 525 [32] Smith, H., Jr. 1987, A&A 188, 233 [33] Zhang, H.W., Zhao, G. 2005, MNRAS 364, 712