The HARPS search for southern extrasolar planets XXI. Three new giant planets orbiting the metal-poor stars HD5388, HD181720, and HD190984
N.C. Santos, M. Mayor, W. Benz, F. Bouchy, P. Figueira, G. Lo Curto, C. Lovis, C. Melo, C. Moutou, D. Naef, F. Pepe, D. Queloz, S. G. Sousa, S. Udry
aa r X i v : . [ a s t r o - ph . E P ] J a n Astronomy & Astrophysics manuscript no. 13489 c (cid:13)
ESO 2018August 28, 2018
The HARPS search for southern extrasolar planets ⋆ XXI. Three new giant planets orbiting the metal-poor stars HD5388,HD181720, and HD190984
N.C. Santos , M. Mayor , W. Benz , F. Bouchy , P. Figueira , G. Lo Curto , C. Lovis , C. Melo , C.Moutou , D. Naef , F. Pepe , D. Queloz , S. G. Sousa , and S. Udry Centro de Astrof´ısica, Universidade do Porto, Rua das Estrelas, 4150-762 Porto, Portugal Observatoire de Gen`eve, Universit´e de Gen`eve, 51 ch. des Maillettes, 1290 Sauverny, Switzerland Physikalisches Institut Universit¨at Bern, Sidlerstrasse 5, 3012 Bern, Switzerland Institut d’Astrophysique de Paris, UMR7095 CNRS, Universit´e Pierre & Marie Curie, 98bis Bd Arago, 75014 Paris, France ESO - European Southern Observatory, Karl-Schwarzschild-Strasse 3, 85748 Garching bei M¨unchen, Germany Laboratoire d’Astrophysique de Marseille, Traverse du Siphon, 13376 Marseille 12, FranceReceived ; accepted
ABSTRACT
We present the discovery of three new giant planets around three metal-deficient stars: HD 5388 b (1.96 M
Jup ), HD 181720 b(0.37 M
Jup ), and HD 190984 b (3.1 M
Jup ). All the planets have moderately eccentric orbits (ranging from 0.26 to 0.57) and longorbital periods (from 777 to 4885 days). Two of the stars (HD 181720 and HD 190984) were part of a program searching for giantplanets around a sample of ∼
100 moderately metal-poor stars, while HD 5388 was part of the volume-limited sample of theHARPS GTO program. Our discoveries suggest that giant planets in long period orbits are not uncommon around moderatelymetal-poor stars.
Key words. planetary systems – planetary systems: formation – Stars: abundances – Stars: fundamental parameters – Techniques:spectroscopic – Techniques: radial velocities
1. Introduction
The discovery of about 400 exoplanets orbiting solar-typestars has opened a number of questions about the ori-gin of the newfound planets (for a review see Udry &Santos 2007). The theories of planet formation are thusconfronted with new and fascinating problems, whose so-lution may give us a new insight into the processes ofplanet formation and evolution.Two major theories have been proposed to explainthe formation of the discovered giant planets. On theone hand, the “traditional” core-accretion model (Pollacket al. 1996), more recently improved to include the effectsof planet migration and disk evolution (e.g. Ida & Lin2004; Mordasini et al. 2009), suggests that giant planets ⋆ Based on observations collected at the La Silla ParanaObservatory, ESO (Chile) with the HARPS spectrograph atthe 3.6-m telescope (ESO runs ID 72.C-0488 and 082.C-0212).Tables 3, 4, and 5 (with the radial-velocities) are only availablein electronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/ are formed by the accretion of gas around a pre-formedmetal-rich (icy) core. On the other hand, the so-calleddisk instability models advocate that giant planets canbe formed by the direct collapse of gas and dust in theproto-planetary disk (Boss 1997).Based on recent observational results (see discussion inUdry & Santos 2007) the core-accretion model is presentlyfavoured. However, the existence of giant planets at morethan 20 AU from their host stars (Kalas et al. 2008; Maroiset al. 2008) was recently suggested to be more easily ex-plained by the disk instability model (Dodson-Robinsonet al. 2009). To understand which of the two processes ofgiant planet formation best explains the detected popula-tion of exoplanets we need more observational data.The solution to this problem may include the under-standing of the well known stellar metallicity-giant planetcorrelation. It is known that giant planets are more easilyfound orbiting metal-rich dwarfs (Gonzalez 1997; Santoset al. 2004b; Fischer & Valenti 2005). This result stronglyfavors the core-accretion model to explain the formationof the majority of the giant planets discovered so far (e.g.Matsuo et al. 2007). Still, it is also known that planet
Santos et al.: The HARPS search for southern extrasolar planets
Table 1.
Stellar parameters for the stars analyzed in thepresent paper.Parameter HD 5388 HD 181720 HD 190984Spectral type F6V G1V F8V m v B − V ± ± ± † Distance [pc] 53 ± ± +37 − † M v ⊙ ] 4.60 1.94 2.69Radius [R ⊙ ] 1.91 1.39 1.53log R ′ HK − − − v sin i [km s − ] 4.2 1.5 3.4 T eff [K] 6297 ±
32 5781 ±
18 5988 ± g ± ± ± ξ t ± ± ± / H] − ± − ± − ± M ⊙ ] 1.21 0.92 0.91 † Derived in this paper (see text) formation is not completely inhibited around metal-poordwarfs (e.g. Cochran et al. 2007; Santos et al. 2007). Thesmall number of planets detected around metal-poor ob-jects still prevents a clear analysis of the low metallic-ity tail of the abundance distribution of planet-host starsthough.To bridge this gap, a number of projects have beenstarted to search for planets in metal-poor samples(Sozzetti et al. 2006; Cochran et al. 2007; Santos et al.2007). In addition to these, large volume-limited samplesof solar neighborhood stars are adding new planets to thelists, some of them orbiting metal-poor stars.In this paper we present the detection by two HARPSGTO programs of three new planets which orbit metal-poor stars. One of these was discovered as part of a largevolume-limited survey (Pepe et al. 2004) for giant plan-ets. The remaining two are the second and third plan-ets discovered in the HARPS metal-poor sample (Santoset al. 2007). In Sects. 2 and 3 we briefly present our sampleas well as the parameters of the host stars. In Sect. 4 wepresent the radial-velocity measurerements and the fittedorbital solutions. We conclude in Sect. 5, where we discussthe implications of the present findings.
2. Sample and observations
The planets presented in this paper were discovered or-biting stars which were followed as part of two differentHARPS GTO programs. The first consists of a sampleof ∼
850 solar-type stars in a volume-limited complement(up to 57.5 pc) of the “closer” CORALIE sample (Udryet al. 2000). More details about this sample can be foundin Naef et al. (2007). The second program consists of asample of ∼
100 metal-poor dwarfs that was searched forthe presence of giant planets. A description of this lattersample can be found in Santos et al. (2007). Observations of all the targets were obtained duringthe Guaranteed Time Observations GTO (Mayor et al.2003). For HD181720 and HD 190984, however, comple-mentary observations were obtained in a separate HARPSprogram whose goal was to search for neptune-mass plan-ets orbiting moderately metal-poor stars.Radial-velocities were derived using the latest versionof the HARPS pipeline. From each spectrum, other pa-rameters of the HARPS cross-correlation function (CCF)such as the bisector inverse slope (BIS – Queloz et al.2001) were also derived.Most of the radial-velocity measurements were notdone in the simultaneous ThAr calibration mode. A strat-egy to average out the noise due to stellar oscillations(Santos et al. 2004a) was also not used in most of themeasurements, since this is not required for the detec-tion of giant planets (the main goal of the two mentionedprograms). We can thus expect the residuals around theorbital solutions to be above 1 m s − . This is particularlytrue for early-G and late-F stars (like our targets), sincethe oscillation and granulation noise is stronger for theseobjects (Dumusque et al. 2009, in prep.).
3. Stellar parameters
The global stellar parameters for each of the planet-hoststars analyzed in the present paper are listed in Table 1.The spectral type, m v , B − V , and parallax (and de-rived distance), were taken from the Hipparcos catalog(ESA 1997), except for HD 190984 (see discussion below). M V , L, and the stellar radius were computed from theabove values, using a bolometric correction from Flower(1996) and the T eff obtained from the spectroscopy (seebelow).The values for the effective temperature, surface grav-ity, microturbulence, and iron abundance were obtainedusing the high-S/N combined HARPS spectra for eachtarget. The values were derived following the method andline-lists described in Santos et al. (2004b) and Sousa et al.(2008). The final values, together with their errors, arepresented in Table 1. We refer the reader to these authorsfor more details.Stellar masses were also derived by interpolating thetheoretical isochrones of Girardi et al. (2000) , using T eff and [Fe/H] obtained from the spectroscopy, as well as V and the parallax from the Hipparcos catalog. We estimatethat the uncertainties are on the order of 10% due to theerrors in the input parameters and also due to differentsystematic effects (Fernandes & Santos 2004).Values for the projected rotational velocity, v sin i , andthe stellar activity level (based on the Ca ii H and K lines)were derived from the HARPS spectra following the gen-eral recipes described in Santos et al. (2000) and Santoset al. (2002), respectively. The computed values show thatall targets have a low chromospheric activity level. See the web interface at http://stev.oapd.inaf.it/cgi-bin/paramantos et al.: The HARPS search for southern extrasolar planets 3
Fig. 1.
Top : Radial-velocity measurements of HD 5388 as afunction of time, and the best Keplerian fit to the data witha period of 777-days, eccentricity of 0.40, and semi-amplitudeof 42 m s − . The residuals of the fit are shown in the lowerbox. Bottom : phase-folded radial-velocity measurements ofHD 5388, and the best Keplerian fit.
Masana et al. (2006) found an effective temperature andstellar radius of 6201 K and 1.91 R ⊙ for this star in verygood agreement with the values derived in the present pa-per (6297 K and 1.91 R ⊙ ). This value for T eff is similar tothe one derived by Nordstr¨om et al. (2004, 6208 K) and tothe value obtained using the (B-V,[Fe/H]) calibration ofSousa et al. (2008, 6260 K) . The stellar metallicity derivedin the present paper ( − Fig. 2.
Top : Radial-velocity measurements of HD 181720 asa function of time, and the best Keplerian fit to the datawith a period of 956-days, eccentricity of 0.26, and semi-amplitude of 8.4 m s − . The residuals of the fit are shown in thelower box. Bottom : phase-folded radial-velocity measurementsof HD 181720, and the best Keplerian fit. the one derived by these latter authors ( − − R ′ HK = − Santos et al.: The HARPS search for southern extrasolar planets
Fig. 3.
Top : Radial-velocity measurements of HD 190984 asa function of time, and the best Keplerian fit to the datawith a period of 4885-days, eccentricity of 0.57, and semi-amplitude of 48 m s − . The residuals of the fit are shown inthe lower box. Bottom : phase-folded radial-velocity measure-ments of HD 190984, and the best Keplerian fit.
The spectroscopic effective temperature obtained for thisstar (5781 K) superbly agrees with the values of 5776,5745, and 5800 K, derived using the calibration presentedin Sousa et al. (2008), and obtained by Masana et al.(2006) and Nordstr¨om et al. (2004), respectively. Themetallicity of − − − R ′ HK = − R ′ HK index of − − The effective temperature derived for HD 190984 (5988 K)agrees well with the value found by Masana et al. (2006,– 5921 K) and Nordstr¨om et al. (2004, 5902 K), as wellas with the value obtained using the calibration of Sousaet al. (2008, 5868 K). The stellar metallicity derived byour spectroscopic analysis ( − − − π =5.28 ± π =5.45 ± v and luminosity of 2.37 and 9.3, respectively. Thesevalues seem to contradict the results from our detailedspectroscopic analysis.We have thus decided to rederive the parallax usingan iterative procedure that makes use of Eq. 1 in Santoset al. (2004b), the relation between luminosity, radius, andparallax, and the isochrones of Girardi et al. (2000). Wefirst fixed the bolometric correction and the visual magni-tude of the star to the values derived by the calibration ofFlower (1996) and the value listed in the Hipparcos cata-log, respectively. An initial value for the stellar mass wasalso obtained using the Hipparcos parallax and V mag-nitude and the metallicity and temperature derived fromspectroscopy. Then, using this value for the mass (withan associated error of 10%), the effective temperature andthe surface gravity (and their respective errors), we took1000 randomly selected Gaussian distributed values forthese three parameters. These were then used to create adistribution for the stellar luminosities and stellar radii,which where in turn used to calculate a distribution ofparallaxes. Once the new value for the parallax was de-rived, a new mass was obtained. Following this procedurewe found a convergence after only three iterations. Thefinal value for the parallax (see Table 1) is significantlydifferent from the value listed in the Hipparcos catalog.The reason for this discrepancy is not clear to us.
4. Radial-velocities
Sixty-eight radial-velocity measurements of HD 5388 wereobtained between September 2003 and September 2009 antos et al.: The HARPS search for southern extrasolar planets 5
Fig. 4.
Bisector inverse slope (BIS) as a function of radial-velocity for the three stars. The x and y scales were set the same forcomparison purposes Table 2.
Elements of the fitted orbits.HD 5388 HD 181720 HD190984 P ± ±
14 4885 ± T ± ±
30 2 449 572 ± a e ± ± ± V r ± − ± ± − ] ω ± ±
12 318 ± K ± ± ± − ] f ( m ) 4.47 10 − − ⊙ ] σ ( O − C ) 3.33 1.37 3.44 [m s − ] N
68 28 47 m sin i Jup ] (Table 3) . The data show a clear periodic signal that canbe well fitted with a Keplerian function with a period of777 days, an eccentricity of 0.40, and a semi-amplitude of42 m s − . This signal is compatible with the radial-velocityvariation induced by a 1.96 Jupiter mass companion or-biting the 1.21M ⊙ dwarf HD 5388 (see Table 2 and Fig. 1).The analysis of the residuals of the Keplerian fit(3.33 m s − ) suggest that some residual trends may bepresent in the data. However, we find no clear evidencefor any extra component. In any case, the average errorof the individual data points is 2.8 m s − , only 1.8 m s − below the residuals (after quadratic subtraction). Giventhe observational strategy used, the relatively high valuefor the stellar projected rotational velocity and the spec-tral type of the star, this value is likely not significant (seeSect. 2).To understand if the periodic radial-velocity signal ob-served could have a non-planetary origin (see e.g. Saar& Donahue 1997; Queloz et al. 2000; Santos et al. 2002,2009), we also analyzed the bisector inverse slope (BIS) Tables 3, 4, and 5, with the HARPS radial-velocities forHD 5388, HD 181720, and HD 190984, respectively, are onlyavailable online. of the cross-correlation Function (Queloz et al. 2001).No correlation between radial-velocity (RV) and BIS isseen (Fig. 4), which suggests that stellar activity of stellarblends cannot explain the RV variation observed. Togetherwith the low activity level of the star, we conclude that the777-day orbital period observed can be better explained bythe presence of a Jupiter-like planet which orbits HD 5388.
A total of 28 radial-velocity measurements of HD 181720were taken between September 2003 and September 2009(Table 4). The velocities show a clear low amplitude andlong period variation (Fig. 2). The data are well fitted witha single Keplerian function with a period of 956 days, aneccentricity of 0.26, and an amplitude 8.4 m s − . Given themass for HD 181720, this signal can be explained by thepresence of a 0.37 Jupiter-masses (minimum-mass) com-panion to HD 181720 (Table 2).The residuals of the Keplerian fit (1.4 m s − ) are slighlyhigher than the average error of the individual radial-velocity measurements (1.0 m s − ). A visual inspection ofthe residuals (Fig. 2) suggests that some structure may Santos et al.: The HARPS search for southern extrasolar planets exist in the data. However, the small number of pointsand the timeline of the measurements does not allow usto confirm this possibility.The analysis of the bisector inverse slope shows thatno clear correlation exists between the BIS and the radial-velocities (Fig. 4). Together with the low activity level ofthe star, we conclude that the 956-day orbital period ob-served can be better explained by the presence of a Saturn-like planet which orbits HD 181720.
From June 2004 to September 2009 we obtained a total of47 radial-velocity measurements HD 190984 (Table 5). Alook at the data (Fig. 3) shows a clear long period radial-velocity signal. Though a complete period is still not ob-servable, the data can be well fitted with a Keplerian func-tion with a period of 4885 days, an eccentricity of 0.57, anda semi-amplitude of 48 m/s. This signal is expected froma 3.1 Jupiter mass companion orbiting HD 190984.The average photon noise error of the radial-velocities(1.7 m s − ) is significantly below the observed residuals(3.44 m s − ). Given the stellar projected rotational veloc-ity, spectral type, and evolutionary stage together withthe observing strategy used (see Sect. 2), this result is notunexpected. No correlation was found between BIS andthe radial-velocities (Fig. 4). The observed radial-velocitysignal is better explained as due to the presence of a longperiod giant planet orbiting HD 190984.The fitted radial-velocity period is about twice as longas the baseline of our measurements. This increases theuncertainty of the orbital solution, as can be clearly seenin the error bars in Table 2. However, as preliminary asit can be, it seems very unlikely that the observed sig-nal is not due to the presence of a giant planet in orbitabout HD 190984. Since the HARPS program which in-cluded this star is now over, we cannot assure that a cor-rect follow-up will be done over the next 5 years to bettersettle this result. We thus prefer to publish the presentdata, with a word of caution to say that the listed orbitalparameters may be subject to some adjustments.
5. Concluding remarks
We present the detection of three new giant planets orbit-ing three moderately metal-poor stars from two separateHARPS GTO programs. This constitutes a strong addi-tion to the previously known number of planets orbitingstars with metallicities significantly below solar.Two of these stars (HD 181720 and HD 190984) werepart of a dedicated program to search for giant planetsorbiting a sample of ∼
100 metal-deficient stars. Togetherwith HD 171028 b (Santos et al. 2007), three planets havebeen discovered orbiting stars from this particular sam-ple. In a very simplistic analysis, this suggests that atleast 3% of stars with a metallicity below ∼− Fig. 5.
Metallicity distribution for the whole HARPS “metal-poor” sample. The three vertical lines denote the [Fe/H] of thethree stars in this sample for which a planet has been detected. of giant planets as a function of stellar metallicity israther flat for [Fe/H] values below solar. Interestingly,HD 171028, HD 181720, and HD 190984 all seem to be inthe metal-rich tail of the metallicity distribution of thesample (see Santos et al. 2007, the former has a metallic-ity of − ∼ ∼ − ) that induce RVsemi-amplitudes smaller than the typical precision of theirmeasurements ( ∼
10 m s − ).A debate exists of whether there is a dependence be-tween stellar metallicity and the orbital period of the dis-covered planets (see discussion in Santos et al. 2006). Ourresults support this tendency. We note however, that afew short period transiting hot Jupiters have been foundwhich orbit metal-poor stars (see table in Ammler-von Eiffet al. 2009).These results may have important implications for themodels of planet formation and evolution. A relativelyhigh frequency of planets orbiting metal-poor stars (in this antos et al.: The HARPS search for southern extrasolar planets 7 case in long period orbits) could imply that the disk insta-bility model is at play at these low [Fe/H] values. Planetformation through disk instability is relatively insensitiveto the metallicity of the disk (and of the host star). Anincrease in the number of known giant planets orbitinglow metallicity stars is crucial to settle these issues. Acknowledgements.
The HARPS spectrograph has been buildby the contributions of the Swiss FNRS, the Geneva University,the French Institut National des Sciences de l’Univers (INSU)and ESO. N.C.S. would like to thank the support bythe European Research Council/European Community un-der the FP7 through a Starting Grant, as well as the sup-port from Funda¸c˜ao para a Ciˆencia e a Tecnologia (FCT),Portugal, through program Ciˆencia 2007. We would also liketo acknowledge support from FCT in the form of grantsreference PTDC/CTE-AST/098528/2008 and PTDC/CTE-AST/098604/2008. S.G.S and P.F. would like to acknowledgethe support from the Funda¸c˜ao para a Ciˆencia e Tecnologia(Portugal) in the form of fellowships SFRH/BPD/47611/2008and SFRH/BD/21502/2005, respectively.
References
Ammler-von Eiff, M., Santos, N. C., Sousa, S. G., et al.2009, ArXiv e-printsBoss, A. P. 1997, Science, 276, 1836Cochran, D. C., Endl, M., Wittenmyer, R. A., & Bean,J. L. 2007, ApJ, in pressDodson-Robinson, S. E., Veras, D., Ford, E. B., &Beichman, C. A. 2009, ArXiv e-printsESA. 1997, The Hipparcos and Tycho CataloguesFernandes, J. & Santos, N. C. 2004, A&A, 427, 607Fischer, D. A. & Valenti, J. 2005, ApJ, 622, 1102Flower, P. J. 1996, ApJ, 469, 355Girardi, L., Bressan, A., Bertelli, G., & Chiosi, C. 2000,A&AS, 141, 371Gonzalez, G. 1997, MNRAS, 285, 403Henry, T. J., Soderblom, D. R., Donahue, R. A., &Baliunas, S. L. 1996, AJ, 111, 439Ida, S. & Lin, D. N. C. 2004, ApJ, 604, 388Kalas, P., Graham, J. R., Chiang, E., et al. 2008, Science,322, 1345Marois, C., Macintosh, B., Barman, T., et al. 2008,Science, 322, 1348Masana, E., Jordi, C., & Ribas, I. 2006, A&A, 450, 735Matsuo, T., Shibai, H., Ootsubo, T., & Tamura, M. 2007,ApJ, 662, 1282Mayor, M., Pepe, F., Queloz, D., et al. 2003, TheMessenger, 114, 20Mordasini, C., Alibert, Y., & Benz, W. 2009, A&A, 501,1139Naef, D., Mayor, M., Benz, W., et al. 2007, A&A, 470, 721Nordstr¨om, B., Mayor, M., Andersen, J., et al. 2004, A&A,418, 989Pepe, F., Mayor, M., Queloz, D., et al. 2004, A&A, 423,385Pollack, J., Hubickyj, O., Bodenheimer, P., et al. 1996,Icarus, 124, 62 Queloz, D., Henry, G. W., Sivan, J. P., et al. 2001, A&A,379, 279Queloz, D., Mayor, M., Weber, L., et al. 2000, A&A, 354,99Saar, S. H. & Donahue, R. A. 1997, ApJ, 485, 319Santos, N. C., Bouchy, F., Mayor, M., et al. 2004a, A&A,426, L19Santos, N. C., Israelian, G., & Mayor, M. 2004b, A&A,415, 1153Santos, N. C., Mayor, M., Bouchy, F., et al. 2007, A&A,474, 647Santos, N. C., Mayor, M., Naef, D., et al. 2000, A&A, 361,265Santos, N. C., Mayor, M., Naef, D., et al. 2002, A&A, 392,215Santos, N. C., Pont, F., Melo, C., et al. 2006, A&A, 450,825Santos, N. C., Silva, J., Lovis, C., & Melo, C. 2009, A&A,submittedSousa, S. G., Santos, N. C., Mayor, M., et al. 2008, A&A,487, 373Sozzetti, A., Torres, G., Latham, D. W., et al. 2006, ApJ,649, 428Sozzetti, A., Torres, G., Latham, D. W., et al. 2009, ApJ,697, 544Udry, S., Mayor, M., Naef, D., et al. 2000, A&A, 356, 590Udry, S. & Santos, N. 2007, ARAA, 45, 397van Leeuwen, F., ed. 2007, Astrophysics and Space ScienceLibrary, Vol. 350, Hipparcos, the New Reduction of theRaw DataWright, J. T. 2004, AJ, 128, 1273