Direct Imaging Search for Extrasolar Planets in the Pleiades
Kodai Yamamoto, Taro Matsuo, Hiroshi Shibai, Yoichi Itoh, Mihoko Konishi, Jun Sudo, Ryoko Tanii, Misato Fukagawa, Takahiro Sumi, Tomoyuki Kudo, Jun Hashimoto, Nobuhiko Kusakabe, Lyu Abe, Wolfgang Brandner, Timothy D. Brandt, Joseph Carson, Thayne Currie, Sebastian E. Egner, Markus Feldt, Miwa Goto, Carol Grady, Olivier Guyon, Yutaka Hayano, Masahiko Hayashi, Saeko Hayashi, Thomas Henning, Klaus Hodapp, Miki Ishii, Masanori Iye, Markus Janson, Ryo Kandori, Gillian R. Knapp, Masayuki Kuzuhara, Jungmi Kwon, Mike McElwain, Shoken Miyama, Jun-Ichi Morino, Amaya Moro-Martin, June Nishikawa, Tetsuo Nishimura, Tae-Soo Pyo, Eugene Serabyn, Hiroshi Suto, Ryuji Suzuki, Michihiro Takami, Naruhisa Takato, Hiroshi Terada, Christian Thalmann, Daigo Tomono, Edwin L. Turner, John Wisniewski, Makoto Watanabe, Toru Yamada, Hideki Takami, Tomonori Usuda, Motohide Tamura
aa r X i v : . [ a s t r o - ph . E P ] J un Direct Imaging Search for Extrasolar Planets in thePleiades
Kodai
Yamamoto , Taro Matsuo , Hiroshi Shibai , Yoichi Itoh , Mihoko Konishi ,Jun Sudo , Ryoko Tanii , Misato Fukagawa , Takahiro Sumi , Tomoyuki Kudo ,Jun Hashimoto , Nobuhiko Kusakabe , Lyu Abe , Wolfgang Brandner ,Timothy D. Brandt , Joseph Carson , Thayne Currie , Sebastian E. Egner ,Markus Feldt , Miwa Goto , Carol Grady , Olivier Guyon , Yutaka Hayano ,Masahiko Hayashi , Saeko Hayashi , Thomas Henning , Klaus Hodapp , Miki Ishii ,Masanori Iye , Markus Janson , Ryo Kandori , Gillian R. Knapp ,Masayuki Kuzuhara , Jungmi Kwon
15, 6
Mike
McElwain , Shoken Miyama ,Jun-Ichi Morino , Amaya Moro-Martin , June Nishikawa , Tetsuo Nishimura ,Tae-Soo Pyo , Eugene Serabyn , Hiroshi Suto , Ryuji Suzuki , Michihiro Takami ,Naruhisa Takato , Hiroshi Terada , Christian Thalmann , Daigo Tomono ,Edwin L. Turner , John Wisniewski , Makoto Watanabe , Toru Yamada ,Hideki Takami , Tomonori Usuda , and Motohide Tamura Department of Earth and Space Science, Graduate School of Science, Osaka University, 1-1Machikaneyama, Toyonaka, Osaka 560-0043, [email protected]@[email protected]@[email protected]@iral.ess.sci.osaka-u.ac.jp Department of Astronomy, Faculty of Science, Kyoto University, Kitashirakawa-Oiwake-cho,Sakyo-ku, Kyoto 606-8502, [email protected] Nishi-Harima Astronomical Observatory, 407-2 Nishigaichi, Sayo-cho, Sayo-gun, Hyogo 679-5313,[email protected] Graduate School of Science, Kobe University, 1-1 Rokkodai, Nada, Kobe, Hyogo 657-8501, Japan Subaru Telescope, 650 North Aohoku Place, Hilo, HI 96720, USA National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, [email protected] Laboratoire Lagrange, UMR7293, Universit ´ e de Nice-Sophia Antipolis, CNRS, Observatoire de laC ˆ o te d’Azur, 06300 Nice, France Max Planck Institute for Astronomy, Heidelberg, Germany Department of Astrophysical Sciences, Princeton University, NJ 08544, USA Department of Physics and Astronomy, College of Charleston, 58 Coming St., Charleston, SC29424, USA Department of Astronomy and Astrophysics, University of Toronto, 27 King’s College Circle,Toronto, Ontario, Canada M5S 1A1 Eureka Scientic, 2452 Delmer, Suite 100, Oakland CA 96002, USA Department of Astronomy, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033,Japan Institute for Astronomy, University of Hawaii, 640 North A’ohoku Place, Hilo, HI 96720, USA Department of Astronomical Science, The Graduate University for Advanced Studies(SOKENDAI), 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan ExoPlanets and Stellar Astrophysics Laboratory, Code 667, Goddard Space Flight Center,Greenbelt, MD 20771, USA Office of the President, Hiroshima University, 1-3-2 Kagamiyama, Higashi-Hiroshima, 739-8511,Japan Departamento de Astrofisica, CAB (INTA-CSIC), Instituto Nacional de T ´ e cnica Aeroespacial,Torrej ´ o n de Ardoz, 28850, Madrid, Spain Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA TMT Observatory Corporation, 1111 South Arroyo Parkway, Pasadena, CA 91105, USA Institute of Astronomy and Astrophysics, Academia Sinica, P.O. Box 23-141, Taipei 106, Taiwan Department of Astronomy, University of Washington, Box 351580 Seattle, WA 98195, USA Department of Cosmosciences, Hokkaido University, Sapporo 060-0810, Japan Astronomical Institute, Tohoku University, Aoba, Sendai 980-8578, Japan (Received ; accepted )
Abstract
We carried out an imaging survey for extrasolar planets around stars in thePleiades (125 Myr, 135 pc) in the H and K S bands using HiCIAO combined with theadaptive optics, AO188, on the Subaru telescope. We found 13 companion candidatesfainter than 14.5 mag in the H band around 9 stars. Five of these 13 were confirmedto be background stars by measurement of their proper motion. One was not foundin the second epoch observation, and thus was not a background or companion ob-ject. One had multi-epoch image, but the precision of its proper motion was notsufficient to conclude whether it was background object. Four other candidates arewaiting for second epoch observations to determine their proper motion. Finally, theremaining 2 were confirmed to be 60 M J brown dwarf companions orbiting aroundHD 23514 (G0) and HII 1348 (K5) respectively, as had been reported in previous2tudies. In our observations, the average detection limit for a point source was 20.3mag in the H band beyond 1 . ′′ σ ) aroundone star of the Pleiades cluster. Key words: infrared: stars — methods: statistical — stars: low-mass, browndwarfs — stars: planetary systems — techniques: high angular resolution
1. Introduction
Understanding planet-building and their evolutionary process is one of the most chal-lenging problems in astrophysics. Theoretically, there have been two main competing hypothe-ses regarding the formation of gas-giant planets: core accretion (e.g., Safronov 1969; Mizuno1980; Pollack et al. 1996) and disk instability (e.g., Kuiper 1951; Cameron 1978). Planet for-mation theories have been continuously updated or newly proposed (e.g., Inutsuka et al. 2010),but these two hypotheses have served as the basis for most studies. On the one hand, in thecore accretion model, relatively small giant planets such as Jupiter and Saturn are thought toform at about 10 AU or less from a solar-type host star in several Myr (Pollack et al. 1996; Ida& Lin 2004). On the other hand, in the disk instability model, planets of a few to 10 M J canbe created within a few 10 to 100 AU from the central star on a dynamical timescale of severalthousand years (Rafikov 2007; Rafikov 2011; Marois et al. 2008; Kratter et al. 2010; Janson etal. 2012). These formation models therefore predict two populations of giant planets segregatedby orbital distance, with the closer planets formed by core accretion and the outer ones by diskinstability.However, planets may experience subsequent orbital migration as a result of interactionwith the parent disk either inward or even outward in the case of type III migration (Masset &Papaloizou 2003). Furthermore, in a system with multiple planets, one can be ejected beyondthe outer radius of the disk through gravitational interaction between planets or their embryos(e.g. Ida & Lin 2004; Veras et al. 2009; Basu & Vorobyov 2012). In addition, free-floating planetsmight be captured at wide orbits, although such widely separated planets are likely rare (onthe order of a few percent, e.g., Kouwenhoven et al. 2010). Thus, a number of mechanisms toexplain the formation and evolution of planets have been theoretically explored, but it is mostimportant to observationally determine planet frequency over a wide range of orbital distances.Observationally, more than 830 extrasolar planets have been found to date, of whichabout 90% were detected by radial velocity (RV) and transit observations (e.g. Mayor et al.3011; Howard et al. 2010). This rapidly growing sample allows a statistical discussion of planetfrequency based on the properties of the planets and their host stars. However, these observingmethods have a limitation: it is difficult to detect planets that are far from host stars, i.e.,more than about 10 AU. Direct imaging, however, which is sensitive to such distant regions, canprovide critical and complementary information to that obtained by indirect detection methods(Marois et al. 2008; Marois et al. 2010; Lagrange et al. 2010; Currie et al. 2011; Carson et al.2012). Given its importance and with the development of instruments and observing techniques,direct imaging has been extensively performed in recent years with large-aperture telescopes.Lafreni`ere et al. (2007) calculated the planet frequency around a single star as less than 0.1 (forseparations in the range 50–250 AU and planet masses 0.5–13 M J ) on the basis of the Geminiobservations of 85 stars. In Nielsen & Close (2010), the frequency (8.9–911 AU, > M J )was estimated to be below 0.2, by compiling the data of 118 stars (Liu 2004; Masciadri et al.2005; Marois et al. 2006; Biller et al. 2007; Lafreni`ere et al. 2007). Moreover, Chauvin et al.(2010) reported VLT observations of 88 targets (10–500 AU, > M J ) that yielded a frequencyof below 0.1. Vigan et al. (2012) reported the frequency of a planet around early type stars(A–F) to be 8 . +10 . − . (1 σ ). The result of the previous direct imaging surveys for the frequencyof a planet summarized in Table 1. The problem with direct imaging is that the sample size issmall compared to that of indirect observations.In these imaging studies, the targets belong to the moving groups and local associationsincluding the β Pictoris moving group, TW Hya Association, Tucana-Horologium Association,and AB Doradus group (Lafreni`ere et al. 2007; Chauvin et al. 2010). Because these associationsare nearby ( ∼ >
50 AU around the member stars.The imaging is conducted with the near-infrared instrument HiCIAO with the AO188 adaptiveoptics on the Subaru telescope (Suzuki et al. 2010; Hodapp et al. 2008). Here we report theimaging results for the first 20 surveyed stars. 4 able 1.
Summary of the direct imaging observations.
Author Sp. Type Target Age Distance Number Investigated range Planet(median) cluster ∗ (Myr) (pc) Mass Separation † frequency(median) (median) ( M J ) (AU) (%)Lafreni`ere et al. (2007) F2–M4 1, 2, 3, 5, 10–300 3.2–34.9 85 0.5–13 50–250 ≤ < > < . +10 . − . (A3) (100) (50) (sma) ∗ Moving groups: (1) α Persei; (2) AB Doradus; (3) β Picoris; (4) Carina; (5) Carina-Near; (6) Columba; (7) η Cha; (8)Hercules-Lyra; (9) IC2391; (10) Local association; (11) Local association subgroup B4; (12) Tucana-Horologium; (13) TWHydrae association; (14) Ursa Major. † Separation: sma: semi-major axis; pro: projected
2. Target selection
Our purpose is to detect extrasolar planets of less than 10 Jovian masses as close aspossible to the central star. Therefore, we selected the Pleiades, a nearby young star clusterobservable from the northern hemisphere. The Pleiades cluster is significantly populous andthus it provides a better probe of the planet frequency at a given age and for a given commonstar-formation history. It is located at 133.5 ± . ± ± < M J . The luminosity of a planet depends on its age and mass. To beconsistent with previous studies, in our work, we have adopted the evolutionary model ofBaraffe et al. (2003) to predict the brightness of planets. The H -band magnitudes for a planetat 125 Myr are thus estimated to be 27.9, 22.5, and 20.4 magnitudes (mag) for 1, 5, and10 M J , respectively. The typical integration time in our observations is about 30 minutes withHiCIAO/AO188, as described later, which provides a detection limit (5 σ ) of 21.5 mag. Thismeans that it is possible to detect a planet less massive than 10 M J .We note that it has been predicted that the formation process itself is also related tothe luminosity evolution of a planet. There are two types of evolutionary models: hot start5nd cold start. Since the hot-start model assumes higher entropy for giant planets, it maycorrespond to planet formation by the collapse of a gaseous disk (Baraffe et al. 1998; Baraffeet al. 2002; Baraffe et al. 2003; Chabrier & Baraffe 2000), while the cold start condition mayrepresent core accretion process (Fortney et al. 2005; Fortney et al. 2008; Marley et al. 2007).It has been shown that higher initial entropy causes a planet to became brighter (Spiegel &Burrows 2012). Thus, the brightness of a planet at a certain age as derived by the hot startmodel serves as an upper limit, while the cold start model represents a lower limit. The modelby Baraffe et al. (2003) is a hot start model. Based on the cold start model (Spiegel & Burrows2012), the H magnitude is predicted to be 22.6 mag for a planet with 12 M J , indicating that wedo not have the sensitivity to detect such planets. Since planet mass estimates are dependenton the evolutionary model that is used, we should be aware of such uncertainties.Target stars in the Pleiades were selected on the basis of the following three criteria.1. The star is brighter than 12 mag in the R band.AO imaging requires a guide star to measure and correct the atmospheric distortion inoptical, so the star should be bright in R to obtain diffraction-limited performance. In thecase of Subaru/AO188, the guide star needs to be located within 30 ′′ of the target; thus,the target star itself is normally used as the AO guide star.2. The membership probability is high.Cluster membership for the target star is confirmed by using the following three criteria.First, the membership probability should be higher than 80% based on the proper motionmeasurements of Belikov et al. (1998) and the target star should not be classified as anon-member by the other proper motion tests of Lodieu et al. (2007). Second, if the starfails to fulfill the first sub-criterion, it needs to have a membership probability (Belikov etal. 1998) higher than 50% and be determined to be a member according to Lodieu et al.(2007). Third, if the star does not satisfy the above two sub-criteria, it should be classifiedas a Pleiades member on the basis of the proper motion and photometry of Stauffer et al.(2007).3. The star has no binary companion that might exert gravitational influence on planetformation.The target star should not be identified as a binary in literature (Bouvier et al. 1997;Raboud & Mermilliod 1998; Lodieu et al. 2007). In addition, there should be no otherbright ( <
15 mag in the H band) object in the field of view (FoV) of 20 ′′ × ′′ by 2MASSobservation.Finally, we selected 60 targets out of 455 stars in the Pleiades (Belikov et al. 1998; Micela etal. 1996; Pinfield et al. 2003; Raboud & Mermilliod 1998).6 able 2. Summary of the observations.
Name Sp. Type Date Obs. mode/ H / K S § R T exp N exp T total Ang.
FoV
Filter (mag) (mag) (sec) (min) (degree)BD +22 574 F8 ∗ H ∗ H H
10 175 29.2 72.82011-01-27 DI / H
10 30 5 -V1171 Tau G8 † H H
30 15 7.5 -HII 2462 G2 † H ∗ H † H H ∗ H ∗ H H
10 65 10.8 -V855 Tau F8 † H H
10 270 45 -HD 24132 F2V ∗ H ∗ H † K S † H † H † H † H ‡ H H H H DI; direct imaging. ADI; angular differential imaging. T exp ; integration time of each exposure. N exp ; total number ofexposures. T total ; total exposure time. Ang. FoV ; rotation angle of field of view during observation. ∗ Wright et al. 2003 † Skiff 2010 ‡ Belikov et al. 2002 § Hmag; Cutri et al. 2003, Rmag; Zacharias et al. 2005 . Observations Twenty of the 60 selected target stars were observed between October 2009 and January2012 (Table 2). The imaging observations were carried out as part of the Strategic Explorationsof Exoplanets and Disks with Subaru (SEEDS, Tamura 2009) by using HiCIAO, which is ahigh-contrast instrument installed on the Subaru telescope (Suzuki et al. 2010; Hodapp et al.2008). HiCIAO has a 2048 × ∼ ′′ × ′′ . The targets were observed either with the H or K S filter. The coronagraphic masks were not used.To obtain the high contrast needed to observe within the close vicinity of a host star,HiCIAO was used in combination with AO188 (Hayano et al. 2010). By using AO, a FWHMof 6–10 pixels (0 . ′′ . ′′
10) was achieved for a point source. In addition, angular differentialimaging (ADI; Marois et al. 2006) was implemented. ADI is an imaging method that allows therotation of the FoV with time but fixes the detector plane relative to the pupil plane by using animage-rotator. As a result, this method can effectively reduce quasi-static noise including thehalo of the star and speckles produced by the telescope, because the noise pattern is fixed onthe detector. The key to obtaining effective noise reduction is a large field rotation; therefore,the imaging was performed to cover the period of transit of the target stars over the meridian,giving a rotation angle of 25–150 degrees. Additionally, the target star was placed at the centerof the FoV to provide a wide area for the planet search.Our observational procedure consisted of three steps. First, 5–10 unsaturated frameswere taken as a reference for the point-spread-function (PSF) of the central star with 1.5 to2.5 s exposure time to avoid saturation. Second, the ADI observations were performed over anintegration of 5 or 10 s in the individual frames to obtain high sensitivity, but with no smearingcaused by the field rotation. The central star was saturated at the peak by this integrationtime, and the saturated area had a radius of 3–6 pixels. Third, several unsaturated frames wereretaken. Table 2 summarizes the information on the observed stars, observing mode, filters,and exposure times of saturated images.If sources were detected around a target star, they were considered to be candidatecompanions (CCs). For HD 23247, the bright (
H < . ′′ M J (brown dwarf mass), in the subsequent part of this paper sinceour focus is not on the stellar regime. The relative positions of CCs against the target starwere measured in the follow-up observations for HD 23912 and V855 Tau to determine whetherthey were co-moving. In the follow-up observations, the direct imaging (DI) mode withoutfield rotation was employed since the CCs have wide angular separation (more than about3 arcsec). V1171 Tau, BD+22 574, and HD 282954 have been observed with a different camera,Subaru/CIAO, in 2005, and the same CCs were detected (Itoh et al. 2011). Thus, our HiCIAO8bservations gave the proper motion measurements combined with the CIAO results. HD 23912was observed three times (in October 2009, January 2010, and January 2011). Since the fieldrotation by ADI was too small ( ∼
10 degree) for the first imaging in October 2009, it wasrevisited in January 2010.
4. Data reduction
The first step of the image processing was to remove the striped pattern caused by fluc-tuations in the bias levels in the individual raw images. The stripes consist of two components:32 horizontal stripes each with a height of 64 pixels, and thin vertical stripes, each 2048 pixelshigh, randomly distributed over the image. These patterns vary with time and are independentamong images. We created the striped pattern for the whole FoV by using the sky region ineach frame, and subtracting it from the raw frame, a process corresponding to sky subtraction.Next, the bad pixels and their clusters were corrected by subtracting the de-striped dark image.Then, we performed flat-fielding by using the dome-flats. Bad pixels randomly occurring inarbitrary pixel positions were interpolated from the surrounding pixels. These calibrations werecarried out by using our own reduction tool for HiCIAO data.The image processing that follows (described below) was performed with IRAF . Sub-pixel shifts cannot be avoided during the process of distortion correction and ADI reductions.They require the interpolation of adjacent pixels, which causes the smearing of pixel values. Asa result, the noise level is reduced. Moreover, the amount of sub-pixel shift was different for eachframe, and we confirmed that the degree of noise reduction could vary among multiple images.Such a non-uniform process, as well as artificial noise reduction, may affect our discussion ofdetection limits. Thus, before applying the distortion correction, all images were smoothed witha 2-D Gaussian filter with an FWHM of 3 pixels to obtain the same level of noise reduction forall pixels and images. The distortion was measured by comparison of images of the globularclusters (M5 and M15) with HiCIAO and HST/STIS (van der Marel et al. 2002). The distortionwas then corrected to obtain a pixel scale of 9.500 ± IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Associationof Universities for Research in Astronomy, Inc., under cooperative agreement with the National ScienceFoundation. ig. 1. The result of ADI reductions for HD 23912. The image was obtained in the H band. The whitecross indicates the position of the central star. A point source was detected at an angular separation of3 ′′
38. North is on the top and east is to the left.
Left panel : The field of view is 19 . ′′ × . ′′
5. The pixelvalue range is -0.4 to +0.6 ADU. The four corners cannot be discussed since these regions are outside theFoV in many frames.
Right panel : The zoom-in image of the companion candidate. The field of view is5 . ′′ × . ′′
8. The pixel value range is +0.0 to +5.0 ADU. was de-rotated to align the field so that north was on the top. Finally, the de-rotated imageswere median-combined with 5 σ clipping to obtain enough sensitivity to detect planetary-massobjects.An example of the final reduced image is presented in Figure 1. The image was obtainedusing data from HD 23912 in the H band taken in January 2010. The rotation angle of theFoV was 73 degrees and the total integration time was 29.2 minutes. At the center of the leftimage, the residual pattern of the subtraction of the stellar halo can be seen. A point source isdetected at 3 . ′′ ± . ′′
028 from the star with a position angle (P.A.) of 14 . ◦ ± . ◦
48. Duringthe ADI, more images are taken at similar P.A. of the field when the rotation is slow. Theemission from point sources in such images cannot be completely eliminated in the referenceimage, and consequently, self-subtraction occurs in the faint outskirts of the point source. Thisis consistent with the sculpting along the azimuthal direction. Images of all the target starsare shown in the Appendix.When a CC was detected, the relative position between the CC and the central star wasmeasured. The centroid position of the CC was determined using an aperture with radius of 1FWHM. A position measurement was performed in each frame or combined frame dependingon the brightness of the CC. In order to determine the position of the central star in saturated10mages, we first, in unsaturated images, determine the offset between the center derived byGaussian fitting and a centroiding algorithm with a mask equal in size to the saturated areain saturated images. Assuming the same offset holds true in the saturated case, we correct themeasurement derived by this masked centroiding algorithm accordingly. The uncertainty ofthe position measurement was checked by the deviation from the rotation center of the field byADI. The relative positions were measured in each (combined) image, and the rotation centerwas defined as the center of the fitted circular orbit for the CC in multiple images withoutde-rotation. Moreover, the deviation between the position of the CC and its fitted circularpath due to the ADI observation was below 0.7 pixels. This deviation encompasses possibledistortions left even after distortion correction (as the shape would not be perfectly circular),thus showing that as far as any measurable effect exists it is small. The results of the astrometrymeasurements are summarized in Table 3.The magnitude of the CCs was estimated by the target star as the flux calibrator. Themagnitudes of both for the central stars and the CCs were measured by aperture photome-try. For the central star, photometry was performed by using the unsaturated frames takenbefore and after the ADI observations as mentioned in section 3. The background level wasestimated as the centroid of the pixel-values histogram in an annulus with a radius of 50 pixelsand the width of 20 pixels. The aperture size varied from 2 to 40 pixels in radius, and theconverged magnitude, at a radius of about 20 pixels depending on the targets, was taken tobe its magnitudes. By comparing this instrumental magnitude with the 2MASS measurementunder the assumption that the star was not variable, the conversion from ADU to magnitudewas obtained. The photometry for the CCs was performed with the same aperture size as thatfor the central star. The flux loss by the image processing including the ADI reductions was ∼ ′′ ) in the raw image and applying the same reduction procedures. Theflux loss was independent of the separation beyond 1 ′′ . The photometry result obtained for theCCs was corrected for this flux loss. Finally, the magnitude of the CCs was calculated using theconversion from ADU to the magnitude derived from the photometry of the central star. Toimprove the signal-to-noise ratio (S/N), the photometry for a CC was performed with imagesin which 20–40 frames were combined and the results were averaged. The H magnitudes forthe CCs are shown in Table 3.
5. Results
The detection limit of our observations is defined by a signal-to-noise ratio (S/N) of 5.The noise was determined by the standard deviation of the background level in the azimuthaldirection measured at the same distance from the target star. The background level was11 able 3.
Astrometry and photometry of companion candidates.
Name Separation Angle P.A. H Mass ∗ UT Date Status( ′′ ) ( ◦ E of N) (mag) ( M J )V1171 Tau CC1 12.770 ± ± † - 2005-11-17 ‡ -12.629 ± ± ± ± ± ± ± ± † - 2005-11-17 ‡ -12.744 ± ± ± ± ± ± ± ± ± ± ± ± ± ± § - 2005-11-17 -3.288 ± ± ± ± ± k
14 2005-11-17 -8.501 ± ± ± ± ± k
33 2005-11-17 -9.031 ± ± ± ± ± ± ± ± ± ?HD23514 CC1 2.64 ± ± ∗∗ -2.64 ± ± ∗∗ -2.62 ± ± ∗∗ -2.642 ± ± ± ∗∗ -2.644 ± ± ± ∗∗ -2.646 ± ± ± ± ± †† -1.097 ± ± ± ‡‡ -1.12 ± ± §§ -1.12 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± Much brighter companion candidate was detected within 3 . ′′ ∼ M J ) are discussed in this paper.Status sign; U presents ”undefined” due to the uncertainty of the proper motion measurement. B presents the backgroundobject. C presents the co-moving object. N presents that the proper motion has not been measured yet. ∗ The masses are linearly interpolated by reference to Baraffe et al. (2003). † It was impossible to measure the individualbrightness of CC1 and CC2, because they were not spatially separated not well enough for aperture photometry. In addition,the error was difficult to determine due to the fluctuated PSF because of the poor seeing. ‡ Subaru/CIAO, Subaru/IRCS(Itoh et al. 2011). § It was impossible to measure the brightness of the CC due to the stellar halo. k K magnitude. Itwas impossible to estimate the error due to the variation of PSF because of the inclement weather. The companioncandidates were not found in the field of view. ∗∗ Keck/NIRC2 (Rodriguez et al. 2012). †† CFHT/PUEO (Bouvier et al.1997). ‡‡ Palomar Hale telescope/PHARO (Geißler et al. 2012). §§ Keck/OSIRIS (Geißler et al. 2012)
14 16 18 20 22 24 0 1 2 3 4 5 no i s e l e v e l ( H m agn i t ude ) angular separation (arcsec) Fig. 2.
Noise level (1 σ ) as a function of angular separation. Dotted lines indicate individual observationsobtained from October 2009 to January 2012. The total integration time in each observation is in therange of 5–45 minutes. obtained with an aperture size of approximately 2 × FWHM on the median-combined image byADI reductions. The relation between the standard deviation (S/N=1) and angular separationfrom the central star is plotted in Figure 2. The median of the detection limits for all ADIobservations becomes constant at 20.8 mag for S/N of 3 and 20.3 mag for S/N of 5 in theregion beyond 1 . ′′ ∼ . ′′
5, the detection limit is determined by thesubtraction residual of the stellar halo. It is 17.7 mag and 19.7 mag for separations of 0 . ′′ . ′′
0, respectively.We note that there are other ways to achieve better suppression of the stellar halothan the classical ADI reductions, such as Locally Optimized Combination of Images (LOCI:Lafreni`ere et al. 2007). The LOCI algorithm considers spatial correlations of the stellar haloand speckle noise with reference images. However, our primary focus in this paper is on therelatively distant region from the star (more than about 100 AU) where uncorrelated, randomnoise is dominant and classical ADI is more effective than LOCI. The results of standard ADIreductions are thus discussed in this work.
Among 13 companion candidates, a CC for HD 23912 is detected in our follow-up imagingwith HiCIAO while the CC of V855 Tau is not found in the follow-up. Another 7 CCs around5 stars (BD+22 574, V1171 Tau, HD 282954, HD 23514, and HII 1348) were observed withSubaru/CIAO, Subaru/IRCS, Keck/NIRC2, CFHT/PUEO, Palomar Hale telescope/PHARO,Keck/OSIRIS at the previous epochs (Itoh et al. 2011; Rodriguez et al. 2012; Bouvier et al.1997; Geißler et al. 2012). The relative distances to the central stars for these CCs are shownin Figure 3. The remaining 4 CCs, 2 for V1054 Tau and V1174 Tau respectively, are waitingfor the second epoch observations for proper motion measurements.13 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)V1171 Tau cc12005.Nov.17 (CIAO)2009.Nov.1 (Present work)2012.Dec.31 (Present work)2012.Dec.31 (Background star)Background star motion (a) V1171 Tau CC1 -9.4-9.3-9.2-9.1-9.0 8.68.78.88.99.0 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)V1171 Tau cc22005.Nov.17 (CIAO)2009.Nov.1 (Present work)2012.Dec.31 (Present work)2012.Dec.31 (Background star)Background star motion (b) V1171 Tau CC2 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)HD 23912 cc12010.Jan.232011.Jan.272011.Jan.27 (Background star)Background star motion (c) HD 23912 CC1 -0.4-0.3-0.2-0.1 3.23.33.43.5 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)BD+22 574 cc12005.Nov.17 (CIAO)2009.Oct.31 (Present work)2009.Oct.31 (Background star)Background star motion (d) BD+22 574 CC1
Fig. 3.
Top left panel:
V1171 Tau CC1.
Top right panel:
V1171 Tau CC2.
Lower left panel:
HD 23912CC1.
Lower right panel:
BD +22 574 CC1. .25.35.45.5 6.46.56.66.76.8 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)BD+22 574 cc22005 Nov 17 (CIAO)2009 Oct 31 (Present work)2009 Oct 31 (Background star)Background star motion (e) BD+22 574 CC2 -2.5-2.4-2.3-2.2-2.1-2.0 8.58.68.78.88.99.0 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)HD 282954 cc12005.Nov.17 (CIAO)2010.Jan.23 (Present work)2012.Sep.12 (Present work)2012.Sep.12 (Background star)Background star motion (f) HD 282954 CC1 -1.8-1.7-1.6-1.5 -2.1-2.0-1.9-1.8 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)HD 23514 cc12006.Dec.10 (NIRC2)2007.Oct.25 (NIRC2)2008.Nov.4 (NIRC2)2009.Nov.1 (NIRC2)2010.Oct.30 (NIRC2)2010.Nov.1 (Present work)2010.Nov.1 (Background star)Background star motion (g) HD 23514 CC1 D E C o ff s e t ( a r cs e c ) RA offset (arcsec)HII 1348 cc11996.Sep. (PUEO)2004.Oct.03 (PHARO)2005.Nov.21 (PHARO)2011.Dec.23 (Present work)2011.Dec.23 (Background star)Background star motion (h) HII 1348 CC1
Fig. 3.
Continued. Top left panel:
BD +22 574 CC2.
Top right panel:
HD 282954 CC1.
Lower leftpanel:
HD 23514 CC1.
Lower right panel:
HII 1348 CC1. .2.1. HD 23514, and HII 1348 HD 23514 and HII 1348 have a co-moving object respectively, which is most likely acompanion gravitationally bound to it (Figure 3(g), and Figure 3(h)). The companion objectswere first identified by the previous astrometry by Rodriguez et al. (2012) for HD 23514, andby Geißler et al. (2012) for HII 1348. The H magnitudes for the companion were measured tobe 15.39 ± ± M J , which is in the brown dwarf regime.Rodriguez et al. (2012) measured the separation and the P.A. of HD 23514 as 2 . ′′ ± . ′′
003 and 227 . ◦ ± . ◦
04 in November 2009, and in October 2010 they were 2 . ′′ ± . ′′
002 and227 . ◦ ± . ◦
05, respectively. In our observation in December 2010, the separation was 2 . ′′ ± . ′′
033 and the P.A. was 227 . ◦ ± . ◦
7. The H magnitude of the CC in December 2010 was15.37 ± . ′′ ± . ′′
005 and 346 . ◦ ± . ◦ . ′′ ± . ′′
02 and 346 . ◦ ± . ◦
6, respectively. In our observation in December 2011, the separationwas 1 . ′′ ± . ′′
03 and the P.A. was 346 . ◦ ± . ◦
9. The H magnitude of the CC in December2011 was 15.7 ± We observed HD 282954 and V1171 Tau two times for measurement of their propermotions with HiCIAO. We confirmed these 3 CCs were the background stars from comparison ofthe astrometry between the two epochs. Two CCs for BD+22 574 show changes in their relativedistances to the central star between the two epochs, and are likely to be background stars. Weconsider it likely that the distortion correction is not perfect for the CIAO data because thedistortion map for the CIAO data cannot be preperly generated due to a limited number of thefield stars in Trapezium, which was observed for the distortion correction. However, becausethe distortion is small at narrow separation, the CC1 of BD+22 574 (separation ∼ . ′′
3) isconfirmed as the background star. It is not clear whether BD+22 574 CC2 os the companionor the background star.
One CC was detected for V855 Tau in January 2011. Interestingly, however, it wasnot detected in January 2012. It is difficult to conclude that we had a false detection in 2011because it is not one of the known artifacts, it is seen in all of the several combined images,and its PSF has a reasonable FWHM without any peculiarity in its shape. It may therefore bea foreground object. HD 23912 also has one CC, but it turned out to be a background star onthe basis of ADI and DI observations with HiCIAO.16 .3. Statistical analysis for estimating the frequency of planets
The purpose of this subsection is to constrain the frequency of planets around a starbased on our observations. First, we define and calculate the detection efficiency ε n as theprobability of planet detection when host star n has one gas-giant planet.To begin with, we consider the separation range where we can detect a planet in ourobservations with HiCIAO/AO188. The detection limit of a point source far from the centralstar is determined solely by the total integration time without being affected by the stellarhalo. As already mentioned in section 5.1, the detection limit of our observations (5 σ ) was20.3 magnitudes with an integration time of 5–45 minutes beyond 1 . ′′
5. However, residualsof the stellar halo remain in the inner ( < . ′′
5) region as seen in Figure 2. In this area, onlybrighter planets, brown dwarfs, and stars can be detected, but we are interested in the regionwhere planets can be detected if they exist. The minimum separation for planet detection,which we define as the inner working angle (IWA), can depend not only on the sensitivity butalso on the field rotation of ADI. In this way, the IWA is determined only by the sensitivityof our observations, which is 0 . ′′ . ′′ > . ′′ F min is defined as the minimum angular separation that a planet with a given mass M P can bedetected.The H magnitude can be converted to planetary mass by using the evolutionary modelby Baraffe et al. (2003), assuming an age of 125 Myr and a distance of 135 pc for the Pleiades.Using this relation, the minimum detectable planet mass M min can be determined for eachseparation.Next, we calculate the detection efficiency, which is the probability that planets lie inthe detectable parameter space of the observation. The detection efficiency ε ( M P , a, e ) to finda planet with a certain orbit in the Pleiades is derived from the planet mass M P , semi-majoraxis a , and eccentricity e . Here, we assume that a host star always has one planet that has theorbital elements; a , e , inclination i (angle between line of sight and normal to the orbital plane),and azimuth φ (angle between line of sight and periapsis). As the planet moves along its orbit,the separation angle F from the central star to the planet varies with the true anomaly θ asdescribed as follows: F = a (1 − e ) D (1 + e cos θ ) q cos ( θ − φ ) cos i + sin ( θ − φ ) , (1)where D is the distance to the Pleiades cluster ( D =135 pc). We then introduce T d , which isthe time per orbital period T P for a planet of M P being in the range of F > = F min . Using T d ,the detection efficiency of a certain orbit is described as g ( M P , a, e, i, φ ) = T d /T P . Consideringthat the line of sight is randomly distributed and independent of a planet orbit the detectionefficiency for one orbit is 17 ( M P , a, e ) = R π/ i = − π/ sin i R πφ =0 g ( M P , a, e, i, φ ) dφdi R π/ i = − π/ sin i R πφ =0 dφdi . (2)Accordingly, the detection efficiency ε n for a host star n can be obtained from thedistribution of planet mass, semi-major axis, and eccentricity by ε n = R M max M min dNdM P R a max a min dNda R dNde ε ( M P , a, e ) dM P dade R M max M min dNdM P R a max a min dNda R dNde dM P dade . (3)Here, we need to consider the number distribution of planet mass, semi-major axis, and eccen-tricity, which are expressed as dN/dM P , dN/da and dN/de , respectively. The distribution ofplanet mass was derived as dN/dM P ∝ M − . ∼− . by the RV survey for planets with orbitalperiods longer than 100 days (Cumming et al. 2008). For the distribution of the semi-majoraxis, dN/da ∝ a − . was obtained from the RV survey for planets with long orbital periods(shorter than 2000 days: Cumming et al. 2008). Finally, the distribution of eccentricity wasderived as dN/de ∝ exp( − . e ) on the basis of data in The Extrasolar Planet Encyclopedia .We assume these distributions in our calculation.Adopting Baraffe et al. (2003), the minimum detectable mass in our observation was6–10 M J at separations larger than 1 . ′′
5. As shown in Figure 2, IWAs are 100 AU for a circularorbit and 50 AU for an eccentric orbit with an eccentricity of 0.9, respectively. Consideringthis result and using equation (3), the detection efficiency ε n ranges from 82–96% for a planetmass of 6–12 M J and semi-major axis of 50–1000 AU.In the above discussion, we calculated the detection efficiency for one planet orbitingone star, ε n . In the next step, we consider the probability of detecting at least one planet, p n ,around a star n ( n = 1 ...N ). p n is calculated from the detection efficiency ε n and the numberfrequency of planets around a host star η , since p n = η × ε n . (4)As noted above, ε n is uniquely determined by the orbital distribution of a planet and thedetection separation range in the observations. On the other hand, p n can be constrained byour imaging results for 20 stars. Therefore, it is possible to constrain the planet frequency η for a host star.In the following analytical approach, we employ Bayes’ theorem as described by Viganet al. (2012) and Lafreni`ere et al. (2007). The probability of detecting at least one planet is η × ε n while that of non-detection is (1 − η × ε n ). The likelihood of the data given ε n is describedas L ( { d n }| η ) = N Y n =1 (1 − ηε n ) − d n · ( ηε n ) d n , (5)where d n is the sign of detection, which equals 1 if at least one planet is detected around a http://exoplanet.eu/ { d n } shows the setof results from N observations. Using this likelihood, the conditional probability distributionthat the set of events { d n } occurs with frequency η is p ( η |{ d n } ) = L ( { d n }| η ) p ( η ) R L ( { d n }| η ) p ( η ) dη , (6)where p( η ) is the prior probability of η . Since η is unknown a priori, p ( η ) = 1.We can determine the range of η as a confidence interval (CI) on a given confidence level(CL) α , α = Z η max η min p ( η |{ d n } ) dη, (7)where η max and η min are the maximum and minimum values of η in the case of { d n } .In our observations, there are 8 companion candidates without proper motion measure-ments (V1171 Tau CC1, CC2, BD+22 574 CC2, HD 282954 CC1, V1054 Tau CC1, CC2,V1174 Tau CC1, and CC2). Even if they are companion objects, their masses are larger thanthat of planets ( > M J ). Thus, no planets is found around 20 stars in our observations, re-sulting in a value of η max of about 17.9% (CL = 95%) for planets in the mass range of 6–12 M J and the semi-major axis of 50–1000 AU. The minimum value η min is always 0 in this case.
6. Discussion
On the basis of our observations of 20 stars, the frequency of planets in the mass rangeof 6–12 M J orbiting at a distance of 50–1000 AU from a host star in the Pleiades (125 Myr,135 pc) is estimated to be 17.9% as an upper limit (2 σ ). This is the first time this constrainthas been obtained for a certain age ( ∼
125 Myr).In a previous direct imaging survey by Lafreni`ere et al. (2007), the frequency of planetsover the mass and separation ranges of 0.5–13 M J and 50–250 AU was below 10%, as derivedfrom observations of 85 stars with the Gemini North telescope. Similarly, the frequency ofplanets of > M J at 40–500 AU was not greater than 9.3% (2 σ ) by the VLT observations of88 stars within 100 pc (Chauvin et al. 2010). Therefore, our estimate is consistent with theseprevious results, indicating that the planet frequency in the Pleiades is not much higher thanin other moving groups and around field stars.According to these results, giant planets are very rare at larger separations (more thanabout 50 AU), although there are a few known candidate systems (e.g., Marois et al. 2008; Itohet al. 2005). Since current formation theory predicts that heavy giant planets can form onlyvia disk instability at distant regions, it is speculated that such instability is not a major in-situ formation process for giant planets. Furthermore, our observations cover a wide area evenbeyond a few 100 AU which is the typical size of protoplanetary disks (Andrews & Williams2007), thus it is difficult to expect that planets form in situ at such a distances from a hoststar. However, it has been suggested that giant planets or their natal fragments in multiple19 P l ane t f r equen cy η Semi-major axis (AU)Present studyDirect imaging (Lafreniere+2007)Direct imaging (Chauvin+2010)Direct imaging (NielsenClose2010)Direct imaging (Vigan+2012)Radial velocity (Cumming+2008)Microlensing (Cassan+2012) 0.01 0.1 1 0.1 1 10 P l ane t f r equen cy η Mass (M J )Present studyDirect imaging (Lafreniere+2007)Direct imaging (Chauvin+2010)Direct imaging (NielsenClose2010)Direct imaging (Vigan+2012)Radial velocity (Cumming+2008)Microlensing (Cassan+2012) Fig. 4.
Planet frequency η as a function of semi-major axis ( left panel ) and planet mass ( right panel ). Thecircle shows our work (50–1000 AU, 6–12 M J ), while the square indicates the direct imaging (50–250 AU,0.5–13 M J ; Lafreni`ere et al. 2007). The rice symbol , the triangle and the pentagon shows other directimaging (10–500 AU, 0.5–15 M J ; Chauvin et al. 2010), (8.9–911 AU, > M J ; Nielsen & Close 2010) and(5–320 AU, 3–14 M J ; Vigan et al. 2012), respectively. The triangle denotes the radial velocity (0.03–3 AU,0.3–10 M J ; Cumming et al. 2008), and the diamond shows microlensing (0.5–10AU, 0.3–10 M J ; Cassan etal. 2012). The dotted lines in the two panels indicate the distribution of the frequency of planets that isderived from our observations. The slopes of the lines are − .
31 and − .
61 in the lef t and right panels,respectively. planetary systems can be ejected into very wide orbits (10 –10 AU) through gravitationalinteraction (Basu & Vorobyov 2012; Veras et al. 2009). At present, the observed rareness is notinconsistent with theoretical predictions that invoke planet–planet scattering.In other planet surveys of the region near host stars using microlensing (OGLE: Beaulieuet al. 2006; Kubas et al. 2008, MOA: Sumi et al. 2010), the frequency of planets with 0.3–10 M J at 0.5–10 AU was 17 +6 − % (Cassan et al. 2012). In addition, the frequency of planets more massivethan 0.3–10 M J over 0.03–3 AU was 10.5 ± .
7% by RV survey (Cumming et al. 2008). Thoughthe detectable separation in these other surveys was different from that in direct imaging, thefrequency of planets according to our survey does not seem to be higher than those obtainedby microlensing and RV surveys (Figure 4, Table 4).In our observations, point sources fainter than 14.5 mag are detected around 9 of 20(40%) target stars whether or not they are real companion objects. The detection limit is 20.3mag in the H band at the separation of 1 . ′′ ′′ . This possibility of finding other point sourcesis consistent with previous direct imaging studies with similar survey depth and size of thefield of view. For instance, CCs were detected toward 32 stars (44%) in the galactic latitude20 able 4. Comparison of observations for planet frequency.
Observation Ref. Distribution index Planet frequencymethod Mass ( α ) Semi-major axis ( β ) ( η ) dN/dM P ∝ ( M P ) α dN/da ∝ a β Direct Imaging Present work -1.31 -0.61 ≤ . ≤ . . ± . +6 − % Our use of β is taken from Cumming et al. (2008). In direct imaging by Lafreni`ere et al. (2007), α and β were thevalues extrapolated from RV observations. Lafreni`ere et al. (2007) and Cassan et al. (2012) assumed a flat distribution inlogarithmic semi-major axis space. of > | | degrees in the imaging by Chauvin et al. (2010). Among them, 5 stars have alreadybeen confirmed as background objects while 78% remain to have their proper motion observed.It is highly likely that most of them are background stars, but we would like to point out thatas a by-product, deep direct imaging would also be useful to discuss galactic models. This,however, is beyond the scope of our paper.
7. Summary
We have carried out a SEEDS imaging survey for detection of extrasolar gas-giant planetsin the Pleiades with the near-infrared imaging instrument HiCIAO and the adaptive opticsinstrument AO188 on the Subaru telescope between October 2009 and January 2012. Thirteencompanion candidates were found around 9 host stars in H band by using ADI observations.The detection limit of our observations (5 σ ) was 20.3 magnitudes with an integration time of 5–45 minutes beyond 1 . ′′
5. For HD 23514 and HII 1348, we confirmed a brown dwarf respectively,which were detected by a previous study with proper motion measurement (Rodriguez et al.2012; Geißler et al. 2012). Five of the 13 candidates were confirmed to be background stars onthe basis of proper motion. One was not found in the second epoch observation; thus, this wasunlikely to be a background or companion object. Only one it was not confirmed whether ornot it is background star, as the precision of their proper motions was not sufficient. Four ofthe 13 remain to be observed to confirm whether they are co-moving.We determined the detection efficiency, which is the probability of finding a 6–12 Jovian-mass planet at 50–1000 AU from the host star in the Pleiades, to be about 90% on the basisof our detection limit. Because there was no detection of such a planet, we estimated that thefrequency of stars having gas-giant planets in the Pleiades is less than 17.9%. This result isconsistent with previous direct imaging studies, indicating that planet frequency in the Pleiadesis not considerably higher than those obtained in moving groups and field stars.21 ppendix. Observation images
Details regarding the images and reduction are described in Table 2 and section 4. Allimages are obtained through ADI reduction in the H and K S (only Figure 5(l); TYC 1800-2144-1). The field of view of all images is 19 . ′′ × . ′′
5. The circle in images represents theposition of the companion candidates (CCs). 22 a) BD+22 574 (b) V1171 Tau(c) HII 2462 (d) HD 23863
Fig. 5.
Top left panel:
BD+22 574.
Top right panel:
V1171 Tau. 2 CCs are in one circle.
Lower leftpanel:
HII 2462.
Lower right panel:
HD 23863. The unit of the color bar is ADU per each exposure time. e) HD 23912 (2010) (f) HD 282954(g) HD 23514 (h) HD 23247 (2011) Fig. 5.
Continued. Top left panel:
HD 23912 (2010).
Top right panel:
HD 282954.
Lower left panel:
HD 23514.
Lower right panel:
HD 23247 (2011). The unit of the color bar is ADU per each exposuretime. i) V855 Tau (2011) (j) HD 24132(k) HD 23061 (l) TYC 1800-2144-1 Fig. 5.
Continued. Top left panel:
V855 Tau (2011).
Top right panel:
HD 24132.
Lower left panel:
HD 23061.
Lower right panel:
TYC 1800-2144-1. The unit of the color bar is ADU per each exposure time. m) HII 1348 (n) Melotte 22 SSHJ G214(o) BD+23 514 (p) Melotte 22 SSHJ G213 Fig. 5.
Continued. Top left panel:
HII 1348.
Top right panel:
Melotte 22 SSHJ G214.
Lower left panel:
BD+23 514.
Lower right panel:
Melotte 22 SSHJ G213. The unit of the color bar is ADU per eachexposure time. q) Melotte 22 SSHJ G221 (r) V1054 Tau(s) V1174 Tau (t) Melotte 22 SSHJ K101 Fig. 5.
Continued. Top left panel:
Melotte 22 SSHJ G221.
Top right panel:
V1054 Tau.
Lower leftpanel:
V1174 Tau.
Lower right panel:
Melotte 22 SSHJ K101. The unit of the color bar is ADU per eachexposure time. eferences An, D., Terndrup, D. M., Pinsonneault, M. H., Paulson, D. B., Hanson, R. B., & Stauffer, J. R. 2007,ApJ, 655, 233Andrews, S. M., & Williams, J. P. 2007, ApJ, 659, 705Baraffe, I., Chabrier, G., Allard, F., & Hauschildt, P. H. 1998, A&A, 337, 403Baraffe, I., Chabrier, G., Allard, F., & Hauschildt, P. H. 2002, A&A, 382, 563Baraffe, I., Chabrier, G., Barman, T. S., Allard, F., & Hauschildt, P. H. 2003, A&A, 402, 701Basu, S., & Vorobyov, E. I. 2012, ApJ, 750, 30Beaulieu, J. P., et al. 2006, Nature, 439, 437Belikov, A. N., Hirte, S., Meusinger, H., Piskunov, A. E., & Schilbach, E. 1998, A&A, 332, 575Belikov, A. N., Kharchenko, N. V., Piskunov, A. E., Schilbach, E., Scholz, R. D., & Yatsenko, A. I.2002, A&A, 384, 145Biller, B. A., et al. 2007, ApJS, 173, 143Bouvier, J., Rigaut, F., & Nadeau, D. 1997, A&A, 323, 139Cameron, A. G. W. 1978, Moon Planets, 18, 5Carson, J., et al. 2012, eprint arXiv:1211.3744Cassan, A., et al. 2012, Nature, 481, 167Chabrier, G., & Baraffe, I. 2000, ARA&A, 38, 337Chauvin, G., et al. 2010, A&A, 509, 52Cumming, A., Butler, R. P., Marcy, G. W., Vogt, S. S., Wright, J. T., & Fischer, D. A. 2008, PASP,120, 531Currie, T., et al. 2011, ApJ, 729, 128Cutri, R. M., et al. 2003, 2MASS All Sky Catalog of Point Sources (The IRSA2MASS All-Sky Point Source Catalog, NASA/IPAC Infrared Science Archive.http://irsa.ipac.caltech.edu/applications/Gator/)Fortney, J. J., Marley, M. S., Hubickyj, O., Bodenheimer, P., & Lissauer, J. J. 2005, Astron. Nachr.,326, 925Fortney, J. J., Marley, M. S., Saumon, D., & Lodders, K. 2008, ApJ, 683, 1104Gratton, R. 2000, ASPC, 198, 225Geißler, K., Metchev, S. A., Pham, A., Larkin, J. E., McElwain, M., & Hillenbrand, L. A. 2012, ApJ,746, 44Hayano, Y., et al. 2010, Proc. SPIE, 7736, 21Hodapp, K. W., et al. 2008, Proc. SPIE, 7014, 42Howard, A. W., et al. 2010, Science, 330, 653Ida, S. & Lin, D. N. C. 2004, ApJ, 604, 388Inutsuka, S., Machida, M. N., & Matsumoto, T. 2010, ApJ, 718, 58Itoh, Y., et al. 2005, ApJ, 620, 984Itoh, Y., Oasa, Y., Funayama, H., Hayashi, M., Fukagawa, M., Hashiguchi, T., & Currie, T. 2011,Research in Astron. Astrophys., 11, 335Janson, M., Bonavita, M., Klahr, H., & Lafreni`ere, D. 2012, ApJ, 745, 4 ouwenhoven, M. B. N., Goodwin, S. P., Parker, R. J., Davices, M. B., Malmberg, D., & Kroupa, P.2010, MNRAS, 404, 1835Kratter, K. M., Murray-Clay, R. A., & Youdin, A. N. 2010, ApJ, 710, 1375Kubas, D., et al. 2008, A&A, 483, 317Kuiper, G. P. 1951, Proc. Natl. Acad. Sci., 37, 1Lafreni`ere, D., et al. 2007, ApJ, 670, 1367Lagrange, A.-M., et al. 2010, Science, 329, 57Liu, M. C. 2004, Science, 305, 1442Lodieu, N., Dobbie, P. D., Deacon, N. R., Hodgkin, S. T., Hambly, N. C., & Jameson, R. F. 2007,MNRAS, 380, 712Marley, M. S., Fortney, J. J., Hubickyj, O., Bodenheimer, P., & Lissauer, J. J. 2007, ApJ, 655, 541Marois, C., Lafreni`ere, D., Doyon, R., Macintosh, B., & Nadeau, D. 2006, ApJ, 641, 556Marois, C., Macintosh, B., Barman, T., Zuckerman, B., Song, I., Patience, J., Lafreni`ere, D., &Doyon, R. 2008, Science, 322, 1348Marois, C, Zuckerman, B., Konopacky, Q. M., Macintosh, B., & Barman, T. 2010, Nature, 468, 1080Masciadri, R., Mundt, R., Henning, Th., Alvarez, C., & Barrado y Navascues, D. 2005 ApJ, 625, 1004Masset, F. S., & Papaloizou, J. C. B. 2003 ApJ, 588, 494Mayor, M., et al. 2011, arXiv, 1109.2497Micela, G., Sciortino, S., Kashyap, V., Harnden, F. R., Jr., & Rosner, R. 1996, ApJS, 102, 75Mizuno, H. 1980, Prog. Theor. Phys., 64, 544Nielsen, E. L., & Close, L. M. 2010, ApJ, 717, 878Pinfield, D. J., Dobbie, P. D., Jameson, R. F., Steele, I. A., Jones, H. R. A., & Katsiyannis, A. C.2003, MNRAS, 342, 1241Pollack, J. B., Hubickyj, O., Bodenheimer, P., Lissauer, J. J., Podolak, M., & Greenzweig, Y. 1996,Icarus, 124, 62Raboud, D., & Mermilliod, J.-C. 1998, A&A, 329, 101Rafikov, R. R. 2007, ApJ, 662, 642Rafikov, R. R. 2011, ApJ, 727, 86Rodriguez, D. R., Marois, C., Zuckerman, B., Macintosh, B., & Melis, C. 2012, ApJ, 748, 30Soderblom, D. R., Nelan, E., Benedict, G. F., McArthur, B., Ramirez, I., & Spiesman, W. 2005, AJ,129, 1616Safronov, V. 1969, Evolution of the Protoplanetary Cloud and Formation of the Earth and PlanetsSkiff, B. A. 2010, VizieR Online Data Catalog, 1, 2023Spiegel, D. S., & Burrows, A. 2012, ApJ, 745, 174Stauffer, J. R., Schultz, G., & Kirkpatrick, J. D. 1998, ApJ, 499, 199Stauffer, J. R., et al. 2007, ApJS, 172, 663Sumi, T., et al. 2010, ApJ, 710, 1641Suzuki, R., et al. 2010, SPIE, 7735.101STamura, M. 2009, American Institute of Physics Conference Series, 1158, 11van der Marel, R. P., Gerssen, J., Guhathakurta, P., Peterson, R. C., & Gebhardt, K. 2002, AJ, 124,3255 an Leeuwen, F. 2009, A&A, 497, 209Veras, D., Crepp, J., & Ford, E. B. 2009, ApJ, 696, 1600Vigan, A., et al. 2012, A&A, 544, 9Wright, C. O., Egan, M. P., Kraemer, K. E., & Price, S. D. 2003, AJ, 125, 359Zacharias, N., Monet, D. G., Levine, S. E., Urban, S. E., Gaume, R., & Wycoff, G. L. 2004, A&AS,205,4815an Leeuwen, F. 2009, A&A, 497, 209Veras, D., Crepp, J., & Ford, E. B. 2009, ApJ, 696, 1600Vigan, A., et al. 2012, A&A, 544, 9Wright, C. O., Egan, M. P., Kraemer, K. E., & Price, S. D. 2003, AJ, 125, 359Zacharias, N., Monet, D. G., Levine, S. E., Urban, S. E., Gaume, R., & Wycoff, G. L. 2004, A&AS,205,4815