A Decade of Radial-Velocity Discoveries in the Exoplanet DomainSt´ephane UdryGeneva University, SwitzerlandDebra FischerSan Francisco State University, USADidier QuelozGeneva University, SwitzerlandSince the detection of planetary companion orbiting 51 Peg one decade ago, more than165 extra-solar planets have been unveiled by radial-velocity measurements. They exhibit awide variety of characteristics, including large masses with small orbital separations, higheccentricities, multi-planet architectures and orbital period resonances. Here, we discuss thestatistical distributions of orbital parameters and host star properties in the context of constraintsthey provide for planet-formation models. We expect that radial-velocity surveys will continueto provide important discoveries. Thanks to ongoing instrumental developments and improvedobserving strategies, Neptune-mass planets in short-period orbits have recently been detected.We foresee continued improvement in radial-velocity precision that will reveal Neptune-massplanets in longer-period orbits and planets down to a few Earth masses in short-period orbits.The next decade of Doppler observations should expand the mass distribution function ofexoplanets to lower masses. Finally, the role of radial-velocity follow-up measurements oftransit candidates is emphasized.1. INTRODUCTIONBefore 1995, the Solar System was the only known ex-ample of a planetary system in orbit around a sun-like star,and the question of its uniqueness was more a philosophi-cal than a scientific matter. The discovery of an exoplanetorbiting the sunlike star, 51 Peg (Mayor and Queloz, 1995),changed this fact and led to a steadily increasing number ofexoplanet detections. During the ensuing years, we learnedfirst that gas giant planets are common and that the plan-etary formation process may produce a surprising varietyof configurations: masses considerably larger than Jupiter,planets moving on highly eccentric orbits, planets orbitingcloser than 10 stellar radii, planets in resonant multi-planetsystems, and planets orbiting components of stellar bina-ries. Understanding the physical reasons for such widevariations in outcome remains a central issue in planet-formation theory. The role of observations is to provideconstraints that will help theoreticians to model the largevariety of properties observed for extra-solar planets.¿From the mere 7 or 8 exoplanets known at the time ofthe PPIV conference (and the 17 candidates published inthe proceedings; Marcy et al., 2000), the number of knownexoplanets has now surpassed 170. With this larger sample,statistically significant trends now appear in the distributionof orbital elements and host-star properties. The features ofthese distributions are fossil traces of the processes of for-mation or evolution of exoplanet systems and help to con-strain the planet-formation models.Here we present a census of the main statistical re-sults obtained from spectroscopic observations over thepast decade. In addition to the orbital properties describedin Sects. 2 and 4, and the primary-star characteristics dis-cussed in Section 5, we will discuss the evolution of radial-velocity measurements over the past 2 years, namely i) therole played by follow-up radial-velocity measurements inconfirming and characterizing planetary objects among themany candidates detected by photometric-transit programs(Section 6) and ii) the development of specially designedhigh-resolution spectrographs achieving precisions for ra-dial velocities below the 1 ms−1limit (Section 3). This ex-treme precision opens the possibility for detection of Earth-type planets with radial-velocity measurements (Section7).2. ORBITAL PROPERTIES OF EXOPLANETSAs a result of the increase in the temporal baseline ofthe large radial-velocity planet searches (Lick, Keck, AAT,ELODIE, CORALIE programs) and the initiation of newlarge surveys (e.g., HARPS planet search; Mayor et al.,2003) and metallicity-biased searches for Hot Jupiters (Fis-cher et al., 2005; Da Silva et al., 2006), there is a largesample of known extra-solar planets. This lends some con-fidence to observed trends in statistical distributions of theplanet properties. The most remarkable overarching featureof the sample is the variety of orbital characteristics. Thisvariety challenges the conventional views of planetary for-mation. A globalvisualillustration of these properties is1Fig. 1.— Separation-eccentricity diagram for the complete sam-ple of presently known extra-solar planets. The size of the dotsis proportional to the minimum mass of the planet candidates(m2sin i ≤ 18 MJup).given in Fig. 1 displaying orbital eccentricities as a func-tion of planet-star separations for the complete sample ofknown extra-solar planets. Several of the planet properties(close proximity to the star, large eccentricity, high mass)are clearly apparent in the figure. The goal now is to inter-pret the observed orbital distributions in terms of constraintsfor the planet-formation models.The determination of statistical properties of giant plan-ets should be derived from surveys that are themselves sta-tistically well defined (e.g., volume limited) and that havewell-understood detection thresholds in the various planet,primary-star and orbital parameters. There are several pro-grams that meet these requirements, including the volume-limited CORALIE planet-search program (Udry et al., 2000)and the magnitude-limited FGKM Keck survey (Marcy etal., 2005). In the diagrams, we present detected planet can-didates from all radial-velocity surveys and note that thediscussed properties agree with those presented from singlewell-defined programs as well.2.1 Giant Extra-solar Planets in NumbersThe most fundamental property that can be obtainedfrom a planet-search program is the fraction of surveyedstars that host detected planets. Given a typical Dopplerprecision of a few ms−1and duration of observations, thisplanet occurrence rate is only defined for a particular pa-rameter space: planets with masses larger than mlimandorbital periods shorter than Plim. The minimum rate is ob-tained just by counting the fraction of stars hosting plan-ets in this particular slice of parameter space. For plan-ets more massive than 0.5MJup, Marcy et al. (2005) findin the Lick+Keck+AAT sample that 16/1330=1.2% of thestars host Hot Jupiters (P ≤ 10d, i.e. a ≤ 0.1AU for a solar-mass star) and 6.6% of stars have planets within 5 AU. Inthe volume-limited
View Full Document
Unlocking...