(Institute for Advanced Study) Properties of Extrasolar Planets Roman Rafikov (Institute for Advanced Study) Methods of detection Characterization Theoretical ideas Future prospects

Methods of detection Methods of detection

Methods of detection Pulsar timing Planetary motion around pulsar causes periodic pulsar displacement around the center of mass by light-seconds for Earth-like planet ( g ) at AU from the neutron star ( ) – well within current detection capabilities. m sin i degeneracy (broken in PSR 1257+12 by planetary interaction) First discovered extrasolar system: PSR 1257+12 (Wolszczan and Frail 1992) A B C Mass, M_Earth 0.015 3.4 2.8 Eccentricity 0.0 0.018 0.026 Period, days 25.3 66.5 98.2 Sem. axis, AU 0.19 0.36 0.47 Second suggested system: PSR 1620-26 (Backer et al. 1993) Planet NS WD Similar to Solar System Recently confirmed (Sigurdsson et al 2003) to have a giant planet in a wide eccentric orbit. Similar to nearby EGP systems.

Radial Velocity Variations Methods of detection Radial Velocity Variations Planetary motion moves star around with velocity . Jupiter-mass planet at 1 AU causes stellar motion with velocity 30 m/s. Searching for periodic Doppler shifts can detect planet (like companion in spectroscopic binary). Method suffers from m sin i degeneracy Velocity noise can be pushed down to 1-2 m/s Up to now the most successful technique – more than 100 planets detected First detection - 51 Peg (Mayor & Queloz 1995) Butler et al 2004 GL 436 Keck

Methods of detection Planetary Transits Jupiter-size planet passing in front of the star causes stellar flux drop (eclipse) by for ~ hours. Transit probability for AU Combined with RV data gives mass (sin i =1) The only method which gives planetary size Expect signal also in reflected light Systematic effects (blends, etc.) First transit detection in previously known planetary system HD 209458 (Charbonneau et al 2000) More than 20 transit searches are underway 1 planet detected (4 more – OGLE transits ) Great ancillary science for time-variable phenomena searches (GRB monitoring campaigns, microlensing searches) TrES-1 (Alonso et al 2004)

Methods of detection Microlensing Bond et al 2004 Gravitational lensing by binary systems leads to easily recognizable pattern of magnification. Light curve fitting yields system parameters and may be used for planet searches. Cannot repeat observation Usually distance is unknown Stellar brightness unimportant Bond et al 2004 OGLE 2003-BLG-235/MOA 2003-BLG-53 (first planet detection by microlensing) Planetary mass Stellar mass Semimajor axis Distance to the system

Methods of detection Astrometry Jupiter-mass planet at 1 AU moves its star around by which for AU at 10 pc results in 100 µas displacement on the sky. Benedict et al (2002) may have detected astrometric signal from previously known planet Gliese 876b (~ 2 M_J, 0.3 AU) using fine guidance sensors on HST. Sensitive to massive and distant planets – requires long time baseline In combination with RV data breaks m sin i degeneracy and determines mass and orbit orientation (similar to observations of stars near the Sgr A*) Requires exquisite astrometric accuracy. Future missions should revolutionize this field (Keck Interferometer, SIM).

Methods of detection

Results: characteristics of extrasolar planets Characterization Results: characteristics of extrasolar planets

Dynamical Characteristics Characterization General Statistics 120 planetary systems, 138 planets, 15 multiple systems 5 by transits 4 by pulsar timing 1 by microlensing 110 by radial velocity searches Dynamical Characteristics Unusual distribution of semimajor axes – many planets very close to the star - Incomplete beyond several AU - Very significant within 1 AU (“Hot Jupiters”) Predominance of planets with large eccentricities for periods longer than several days - closest ones are circularized by stellar tides - e-distribution is close to that of binary stars Very different from Solar System planets!

Characterization Physical Properties Planetary sizes consistent with expectations except that Radius of HD 209458 is too large to be accomodated by simple models. High masses (only recently RV started probing Neptune mass range). Larger number of lower mass planets. “Brown dwarf desert” – relative lack of planets heavier than . Clear correlation between the metallicity of the parent main-sequence star and the probability of hosting planet. Pulsar planet in M4 may have formed in very low-Z environment; planets around PSR 1257+12 may have formed from very high-Z material.

Simulation from (Vidal-Madjar et al 2003) Characterization HD 209458 (a.k.a. “Osiris”) HST transit curve (Brown et al 2001) Mass , period 3.5 d, AU Parent star (G0 V) at a distance of 47 pc Radius of is a challenge for theory Direct probe of planetary atmosphere Na absorption detected in transit – atmospheric signature (Charbonneau et al 2002) Extended absorbing region in H-alpha – envelope of escaping hydrogen (Vidal-Madjar et al 2003) Evidence for oxygen and carbon in this escaping envelope Gas loss rate g/s Evaporation time is more than a Hubble time Simulation from (Vidal-Madjar et al 2003) (Vidal-Madjar et al 2004) Planet is safe from evaporation!

Theoretical ideas Theoretical Ideas

Dynamical peculiarities Theoretical ideas Dynamical peculiarities Tight planetary orbits Evidence for planet migration – planets did not form where we observe them They moved there as a result of the tidal interaction with the gaseous disk Why did they stop? What is the parking mechanism? Bryden Large orbital eccentricities Possible signature of planet-planet gravitational scattering Could have also been caused by the tidal planet-disk interaction Resonant interaction of migrating planets Rasio & Ford 1996 We don’t really know these things very well

1 M_J planet (Lubow et a al 2003) Theoretical ideas Peculiarities of physical structure Mass distribution Mass distribution arises from interplay of gas accretion and gravitational interaction of growing planet with surrounding gas disk Formation of “gap” can probably explain the “brown dwarf desert” Current observed overabundance of high mass planets can be purely a selection effect Metallicity dependence Can be accounted for assuming that giant planets form via the gas accretion onto the rocky cores Could be a signature of stellar pollution by accreting planetary material (e.g. migrating planets which failed to hit the brake) 1 M_J planet (Lubow et a al 2003)

Future prospects Future Prospects

Kepler Space-based Photometer ( 0.95-m aperture ) Future prospects Kepler Space-based Photometer ( 0.95-m aperture ) Photometric One-Sigma Noise <2x10-5 Monitor 100,000 main-sequence stars for planets. Mission lifetime of 4 years. Launch date – October 2007 Transits of terrestrial planets: - About 50 planets if most have R ~ R_Earth, Modulation of the reflected light from giant inner planets: - About 870 planets with periods less than one week. Transits of giant planets: - About 135 inner-orbit planet detections along with albedos for about 100 of them.

SIM (Space Interferometry Mission) Future prospects SIM (Space Interferometry Mission) Precision astrometry on stars to V=20 Optical interferometer on a 10-m structure Global astrometric accuracy: 4 microarcseconds (µas) Typical observations take ~1 minute; ~ 5 million observations in 5 years Launch date - February 2010 Search for terrestrial planets around the nearest 250 stars (1 µas) “Broad survey" for Neptune-size planets around 2000 stars (3 µas) Giant planets around young stars

Future prospects TPF (Terrestrial Planet Finder) Two complementary space observatories: a visible-light coronagraph and a mid-infrared formation-flying interferometer. TPF coronagraph launch is anticipated around 2014. TPF interferometer launch is anticipated before 2020. - Survey nearby stars looking for terrestrial-size planets in the "habitable zone“. - Follow up brightest candidates with spectroscopy, looking for atmospheric signatures, habitability or life itself. - Will detect and characterize Earth-like planets around as many as 150 stars up to 15 pc away. Goals:

Conclusions More than 100 planets are now known around main sequence stars and neutron stars, among them 15 multiple systems. They are detected by 4 different methods: - Radial velocity searches - Pulsar timing - Transits - Microlensing Most extrasolar planets exhibit unusual dynamical characteristic: small orbits and high eccentricities. Their physical properties are similar to those of giant planets in the Solar System (Jupiter and Saturn). In the next 10 years at least 3 space borne missions will open up new horizons in this field. More than 30 are currently underway.