Extra-Solar Planet Studies: the Italian Contribution

Mem. S.A.It. Suppl. Vol. 3, 24 Memorie della °c SAIt 2003 Supplementi

Extra-solar planet studies: the Italian contribution

R.G. Gratton1, R. Claudi1, S. Desidera1, F. Marzari2, and M. Barbieri3

1 INAF - Osservatorio Astronomico di Padova, Via Bergamo 5, 35122 Padova e-mail: [email protected] 2 Dipartimento di Fisica, Universit`adi Padova, Via Marzolo 8, 35131 Padova e-mail: [email protected] 3 CISAS, Universit`a di Padova, Via Venezia, Padovae-mail: [email protected]

Abstract. The search and characterization of extrasolar planets is one of the hot topic of astrophysical research. The italian community is beginning to contribute to this ﬁeld. Main activities concern: the study of protoplanetary disks; the studies of planet formation, dynamical stability and evolution of their orbits; the search for planets in binary systems using high precision radial velocities acquired with SARG; the search for transiting planets in open clusters; and the participation to the VLT Cheops Planet Finder and the ESA satellites Eddington and GAIA.

Key words. stars: extrasolar planets – stars: circumstellar matter

1. Background ferent observational techniques. More than a hundred giant planets orbiting stars are The quest for extra-terrestrial life is one known at present, with minimum masses of the major issues for the XXI century derived from Doppler shifts ranging from science. Life similar to that on the Earth 0.2 to 10 Jupiter masses. The detailed can likely only originate in planets around shape of the transit curve of the planet or- stars. For astronomy, the first step is a biting the star HD209458 has yielded the full understanding of the mechanisms that first direct measurements of the physical lead to their formation, so that conditions properties of the planet (Charbonneau et for habitable planets are solidly defined. al. 1999), while its atmosphere has been The first detection of a Jupiter-like planet probed by measuring the sodium absorp- around a main-sequence solar-type was ob- tion as the stellar light passes through the tained using precise radial velocity mea- planetary limb (Charbonneau et al. 2002). surements of 51 Peg by Mayor & Queloz Based on our present knowledge, more than (1995). Since then, the number of new ex- 5% of solar-type stars may harbour massive trasolar systems have steadily grown and planets and maybe an even higher percent- their nature has been investigated with dif- age may have planets of lower masses and larger orbital radii. Comprehensive reviews Send offprint requests to: R.G. Gratton of the status of extra-solar planet investi- R.G. Gratton et al.: Extrasolar Planet Studies 25 gations are given by Udry et al. (2003) and namical properties of binary (or multiple) Marcy et al. (2003). star systems with planets. Also, the possi- Notwithstanding the large number of ble existence of planets in dense environ- known extrasolar planet systems, their di- ments of star clusters is still not a settled versity in terms of dynamical properties has question. raised many questions concerning their ori- The metal-rich nature of stars with gin and evolution. It is matter of debate planets represents the only link between the whether these planets formed according to presence of a planet and a stellar photo- the conventional core-accretion model for spheric feature. However, this relationship the origin of Jupiter and Saturn (Pollack has to be strengthened by additional ob- et al. 1996) or by gravitational instabili- servations and purged by a possible sys- ties of the protoplanetary disks (Boss 1997, tematic effect related to the dependence of 1998). Important clues can be derived from the metallicity of stars in the solar neigh- the detailed studies of the physical proper- borhood on the galactic orbit of the stars ties of the circumstellar disks around pre- (Barbieri & Gratton 2002). main sequence stars. An additional conun- drum posed by the new (giant) planets is their orbits, that are in many cases well 2. Italian activities inside 1 AU from the parent stars. These The Italian groups involved in the extraso- planets may have originated at larger dis- lar planet research field investigates various tances and subsequently migrated inward. aspects, working from both a theoretical Theoretical studies are needed to detail the and observational point of view on topics possible migration processes that account related to the origin and evolution of ex- also for the substantial orbital eccentrici- trasolar planets. ties of the objects. Note that systems with giant planets in highly eccentric orbit are quite unlikely to host life: hence a clear un- 2.1. Proto-planetary disks derstanding of the process generating these eccentric orbits is basic in order to under- The group of the Arcetri Observatory stud- stand if the life-friendly circular orbits of ies the evolution of dust and structures of the giant planets of the solar system are a protoplanetary disks around pre-main se- rule or rather an exceptional case. quence stars via near and mid IR, as well Planetary formation is not limited to as radio high spatial resolution imaging single stars but can also occur in the more and spectroscopy. The goals are to pro- frequent multiple-star systems. Binary sys- vide constraints on the first stage of plan- tems and clusters can be considered as lab- etary formation process, and in particular oratories, where we may study the impact to investigate the growth of grains in cir- of dynamical perturbations on planetary cumstellar disks around pre-main sequence orbits. It is under investigation whether the stars (Testi et al. 2001a; Natta et al. 2001, gravitational perturbations of the binary Testi et al. 2003); and to provide a near- companion can influence in some way the infrared low-resolution spectroscopic clas- planetary formation process and the sub- sification scheme suitable for application sequent evolution of the planetary system to young giant extrasolar planets; this sys- (Marzari & Scholl 2000; Quintana et al. tem is based on spectra provided by the 2002; Barbieri et al. 2002). Available data TNG/Amici prism (Testi et al. 2001b). indicate that the characteristics of planets in binaries may be different from those or- 2.2. Dynamics biting single stars (Zucker & Mazeh 2002). Additional observational data and theoret- A necessary step to interpret the observa- ical modelling may better outline the dy- tional data on exoplanets is to study the 26 R.G. Gratton et al.: Extrasolar Planet Studies origin and long term evolution of extrasolar planetary systems. Within the ’stan- dard model’ for planet formation based on planetesimal accretion, the group at the Padua Department of Physics concentrated on the origin of terrestrial and giant planets in close binary star systems. A cru- cial parameter for planetesimal accretion is the relative velocity among the planetesi- mals that determines whether accretion or rather fragmentation dominates their evolution. The gravitational perturbations by a closeby companion star can pump up the Fig. 1. Radial velocities for τ Cet orbital eccentricities of the bodies, signifi- cantly increase their relative velocity, and then inhibit the formation of a terrestrial planet or core of a gaseous planet. This group addressed the dynamical questions concerning the migration of giant planets in exoplanetary systems by gravitational scattering (Marzari & Weidenschilling 2002) and the possible metal enrichment of stars by infall of planets on the convective envelope of the star. The group also studied planetesimal accretion and orbital stability of terrestrial planets and giant planet cores in binary systems Fig. 2. Differences in the metallicty values for (Barbieri et al. 2002). In particular, the ori- components of binary systems as a function of gin of the planet detected in the γ Cep the difference in temperature. Open symbols system is investigated. This is by far the are from Gratton et al. (2001); filled ones are system with the closest companion star lo- from Desidera et al. (2003b) cated at a = 21.36 AU on a highly eccentric orbit (e = 0.44). Preliminary results show that the planetesimal dynamics is strongly planet is located at a = 2.15 AU. The esti- perturbed by the secondary star and that mate of the stability limits for planetary or- only large values of the protoplanetary disk bits in the γ Cep system was also refined by density, as in Bodenheimer et al. (2000), al- using the Frequency Map Analysis method lows an ’in situ’ formation of the detected (Laskar et al. 1992, Laskar 1993a,b) tuned planet. The gas drag coupled to the pertur- by long term numerical integrations with bations of the star causes an alignment of the Hierarchical Jacobi Symplectic integra- the orbital pericenters that lead to low col- tor (Beust, 2003) designed to integrate mul- lisional velocities. A critical aspect of the tiple stellar systems. The outer borders of planetary formation in γ Cep is the migra- the stable region ranges from 4 AU for low tion of the planet during the giant impact eccentricity orbits to 2 AU for highly ec- phase when large planetary embryos collide centric orbits. to form its core. During this stage the core tends to migrate inwards by scattering the 2.3. The SARG extrasolar planet search other protoplanets. According to our sim- ulations the core drift stops preferentially These theoretical studies are comple- at 1 AU from the star, while the observed mented by the TNG long-term project R.G. Gratton et al.: Extrasolar Planet Studies 27

nents, the primary being more metal-rich than the secondary (Gratton et al. 2001). A similar analysis on 21 other systems (Desidera et al. 2003b) shows that this is the most exceptional case of abundance differences between the two components (see Figure 2). Differences significative at about 2 σ have been found for two other systems. The abundance differences between the two components of HD219542 are weakly cor- related with the condensation temperature of the various chemical elements, suggest- Fig. 3. Radial velocity curve for HD219542B. ing the ingestion of about 4 earth masses Superimposed is the best keplerian solution of rocky material. However, this last con- clusion needs to be strengthened by more observations. On the other hand, the ra- searching for extrasolar planets using the dial velocities (Desidera et al. 2003a) do high resolution spectrograph SARG at the not show any evidence for a planet around TNG, by the Padova, Catania and Teramo the primary: the upper limits are already Observatory groups, in collaboration with significative enough to exclude a system M. Endl (Texas). The observational sam- like 51 Peg. A Saturn-like planet orbit- ple includes 50 visual binary systems (sep- ing at a distance from the star similar to aration of a few arcsec). The selected bi- that of Mercury from the Sun is compati- naries have main sequence components of ble with our observations for the secondary nearly equal mass. The program uses the (see Figure 3). The probability that such SARG spectrograph (constructed by the a planet indeed exists is rather high (sta- same group), equipped with an absorbing tistical test show a significativity of about cell. The high precision radial velocities are 97%); however it is not possible to exclude obtained using AUSTRAL developed by at present that the observed signal is due to Endl and coworkers (Endl et al. 2000). The activity. If confirmed, this would be the first program is completed by an extremely ac- extrasolar planet discovered by an Italian curate analysis of the differences in chem- group. ical composition between the two components, with errors as small as a few per 2.4. Planets in open clusters cent, that allows to show the existence of even very small differences in the chemical In parallel to the SARG survey, a composition between the two components photometric monitoring of old, super- of the observed systems. metal rich open clusters is performed Figure 1 shows the radial velocity in a collaboration between the Padova curve obtained for the bright reference star Astronomy Department and Astronomical τ Ceti, which is supposed to have a con- Observatory (Piotto et al. 2003). The aim stant radial velocity. The internal errors of is to estimate the frequency of stars with our measures (about 1.4 m/s) represent one planets by the transit method, providing of the best results obtained insofar in this costraints on the planetary formation in kind of investigations. metal-rich stars. The effectiveness of gravi- The most interesting results concern tational encounters between stars of a clus- the system HD219542. For this system, ter in disrupting planetary systems (and the differential chemical abundance anal- prevent its formation) may eventually jus- ysis showed a significative offset in chem- tify a poor statistics in extrasolar planet ical composition between the two compo- detection. 28 R.G. Gratton et al.: Extrasolar Planet Studies

2.5. Polarization tion by several orders of magnitude. This raises the possibility of detecting young Finally, a group at Padova Astronomy planets around the closest young stars, in Department is also studying in a system- spite of their relatively large distances. The atic way the polarization properties of stars phase A study of CHEOPS will end within with and without planets (Tamburini et al. autumn 2004; if the project will be ap- 2002): this study is intended to provide ba- proved, observations are planned to start sic data to the hypothesis (drawn from a in 2008, three years before the JWST will smaller sample) that the planet formation be launched. efficiently removes the original dust. The direct detection of extrasolar planets is extremely difficult even with 8 m 3. New instrumentation for telescopes, due to the large intensity ra- 8 extrasolar planet researches tio (a typical value being 10 ) between the light from the star and that from the Italian groups interested in extrosolar plan- planet. From ground, the main difficulty is ets are actively participating to the design related to the presence of speckles (and re- of various instruments devoted to search, lated noise) due to the atmospheric turbu- detection and characterization of extraso- lence. An adaptive optics system able to lar planets. provide a high Strehl ratio is mandatory. Furthermore, CHEOPS will exploit differential techniques, considering subtraction 3.1. CHEOPS: a Planet Finder for the of images characterized by an ideally iden- VLT tical speckle pattern, but where the planet Direct detection of habitable planets, and signature is present/absent. CHEOPS con- possibly of signs of life on them, is siders two different strategies: one based one of the main goal of astrophysics on simultaneous acquisition of narrow band (see e.g. the Origin program by NASA). images obtained by means of an Integral Achievement of this goal will require ob- Field Spectrograph (IFS); and the sec- servations from space. However, giant plan- ond based on quasi-simultaneous polarime- ets might perhaps be discovered from try, where images obtained with differ- ground. ESO has included a Planet Finder ent polarization angles are acquired at a among the instruments proposed for the rate faster than the rate of variation of Second Generation at VLT. One of the the atmospheric turbulence. The Padua two groups who answered this call is the Astronomical Observatory has responsibil- CHEOPS project, leadered by M. Feldt at ity for the design and construction of the MPIA of Heidelberg, which includes vari- IFS (co-PI: M. Turatto). ous Italian institutions, under the coordina- tion of Padua Observatory. The main goal 3.2. Eddington of CHEOPS is the detection and characterization of planetary systems similar to Eddington is an ESA mission which has two our own, with giant planets in the outer primary science aims: to produce data on regions. In fact, stars with giant plan- stellar oscillations necessary to understand ets in such external orbits may be good stellar structure and evolution and charac- candidates for harbouring terrestrial plan- terize habitable - as well as other types of ets in the habitable zone. CHEOPS may - planets around other stars (Favata 2002). probably discover giant planets in differ- The planet finding will be conducted with ent phases of their evolution. In fact, dur- the method of transit detection. Eleven ing the on-going contraction and accre- italian institutes will be involved in the tion phases, the internal luminosity of giant Eddington Project with different aims. In planets exceeds the reflected light contribu- particular the scientific interest of Italian R.G. Gratton et al.: Extrasolar Planet Studies 29 community includes (i) partecipation in the Boss, A.P. 1998, ApJ, 503, 923 search for habitable planets; (ii) parteci- Charbonneau, D., et al. 1999, ApJ, 529, 45 pation to the asteroseismological observa- Charbonneau, D., et al. 2002, ApJ, 568, 377 tions and data analysis; (iii) interest in the Desidera, S., Gratton, R.G., Endl, M., et al. ”parallel science” results; (iv) partecipa- 2003a, A&A, 405, 207 tion in the development of the software for Desidera, S., Gratton, R.G., Claudi, R., et al. 2003b, A&A, in preparation on board and ground – based data analy- Endl, M., Kurster, M., & Els, S. 2000, A&A, sis; and (v) ground based observations in 362, 585 support of the mission. Eddington will be Favata,F., 2002, in Eddington Workshop on launched in 2007 just some months before Stellar Structure and Habitable Planet its major competitor: Kepler (recently ap- Finding, ESA SP-485, p.3 proved by NASA). Europe then, and Italy Gratton, R.G., et al. 2001, A&A, 377, 123 in particular, with the Eddington mission Laskar J. 1993a, Physica D 67, 257 aims to be the first in the rush for the am- Laskar J. 1993b, Cel. Mech. and Dyn. Astr. 56, bitious goal to find Earth twins. 191 Laskar J., Froeschlè C., Celletti A. 1992, Physica D 56, 253. 3.3. Gaia Lattanzi, M.G., et al. 2002, in GAIA: A European Space Project, O. Bienaymé& The astrometric group at Osservatorio C. Turon eds., EDP Sciences, p. 207 Astronomico di Torino is involved in the Marcy G.W., et al. 2003, in Scientific Frontiers design and characterization of NASA and in Research on Exstrasolar Planets, ASP ESA astrometric satellites SIM and GAIA, Conf. Series, in press and in particular in the definition of the re- Marzari, F., & Scholl, H. 2000, ApJ, 543, 328 quirements for detection of extrasolar plan- Marzari, M., & Weidenschilling, S.J. 2002, ets by means of astrometric perturbations Icarus, 156, 570 (Sozzetti et al. 2002; Lattanzi et al. 2002). Mayor, M., & Queloz, D. 1995, Nature, 378, 355 SIM and GAIA will be launched at the be- Natta, A., et al. 2001, A&A, 371, 186 ginning of the next decade. Thousands of Piotto, G., Montalto, M., Desidera, S. et al. extrasolar planets with periods in the range 2003, in Extrasolar Planets: today and to- from a few days to a few years could be dis- morrow, XIX IAP Coll. Paris, in press covered with these satellites, and full orbit Pollack, J.B., et al. 1996, Icarus, 124, 62 could be determined for a substantial frac- Quintana, E.V., et al. 2002, ApJ, 576, 982 tion of them. Sozzetti, A., et al. 2002, PASP, 114, 1173 Tamburini, F., Ortolani, S., & Bianchini, A. 2002, A&A, 394, 675 References Testi, L., et al. 2001a, ApJ, 554, 1087 Testi, L., et al. 2001b, ApJ, 552, L147 Barbieri, M. & Gratton, R.G. 2002, A&A, 384, Testi, L., et al. 2003, A&A, 403, 323 879 Udry S., Mayor M., Queloz D., et al. 2003, Barbieri, M., Marzari, F., & Scholl, H. 2002, in Scientific Frontiers in Research on A&A, 396, 219 Exstrasolar Planets, ASP Conf. Series, in Beust, H. 2003, A&A, 400, 1129 press Bodenheimer, P., Hubickyj, O., Lissauer, J.J. Zucker, S., & Mazeh, T. 2002, ApJ, 568, L113 2000 Icarus 143, 2 Boss, A.P. 1997, Science, 276, 1836