Exoplanets versus brown dwarfs: the CoRoT view and the future
EExoplanets versus brown dwarfs :the CoRoT view and the future Jean SchneiderLUTh – Observatoire de Paris
CoRoT has detected by transit several tens of objects (Moutou & Deleuil 2015)whose radii run from 1.67 Earth radius (CoRoT 7b, Leger et al. 2009) to 1.5 Jupiterradius (CoRoT-1 b, Barge et al. 2007) . Their mass run from less than 5.7 Earth mass(CoRoT-24 b, Alonso et al. 2014) to 63 Jupiter mass (CoRoT-15 b, Bouchy et al.2011). Their mass-radius diagram is represented in Figure 1 below. One could betempted to think that more massive the object is, the larger it is in size and that thereis some limit in mass and/or radius beyond which objects are not planets but very lowmass stars below the 80 Jupiter mass limit to trigger nuclear fusion (namely « browndwarfs » ). CoRoT findings contribute to the planet versus brown dwarf debate sincethe Figure 1 shows that there is no clear mass-radius relation. igure 1. Mass radius relation for CoRoT objects (15 Oct 2015) from exoplanet.eu.
One is thus facing two problems : terminology (what is a planet ? what is a browndwarf?) and classification (how to decide if a given object is a planet or a browndwarf according to a given definition ?). Let us discuss these two aspects and theCoRoT contribution.
What is a planet ?
The debate, open by several authors (see for example Baraffe et al. 2010, Schneideret al. 2011, Hatzes & Rauer 2015), is still ongoing and will not be closed by thepresent contribution. Names are arbitrary conventions, but the naturel trend is tomake them designate sufficiently elaborated concepts. Derived from the solar systemanalogy, exoplanets (in short planets) designate small bodies orbiting around starsand formed by condensation in a circumstellar dust disk. A first question is of coursehow small has the body to be for being a planet. The problem here is that there existsmall bodies orbiting stars which are probably not formed like planets, namely browndwarfs forming, like stars, by collapse of a (possibly dusty) gas cloud.
From the heaven of concepts to the hell of observations o we have a clear conceptual discrimination between planets and brown dwarfs(keeping in mind that it is a convention). But it is based on a criterion involving aninobsrvable concept, namely its formation scenario, because we do not have theformation movie at hand. We can only rely on actual observables. Standard basic bulkobservables are the object mass, radius, temperature. An ideal situation would be thatat least for one of these observables there exist two domains D planet and D brown dwarf of values which do not intersect. It is unfortunately not the case since there areplanets smaller or larger, heavier or lighter, cooler or hotter than objects we believe tobe brown dwarfs, Even worse, there are a few pulsar companions with masses lowerthan 30 Jupiter mass. They are probably the relict of stellar companions eroded bythe pulsar strong wind (Ray & Loeb 2015). One can argue that as such they are notplanets nor brown dwarfs, their formation process being very different. But onecannot exclude that such erosion mechanism happend also for low mass compnionsof main sequence stars with strong winds (see e.g.
Sanz-Forceda et al. 2010). Thechoice made by the Extrasolar Encyclopaedia at exoplanet.eu, based on Hatzes &Rauer (2015), is to take all objects below 60 Jupiter mass.
Figure 2
Empirical mass-density relation (Hatzes & Rauer 2015) he Hatzes & Rauer argument is that the mass-radius and the mass-density relationpresents no particular feature in the giant planet régime (i.e. more massive thanSaturn) and that there is a change in the slope of distribution at 60 Jupiter mass(Figure 2). But unfortunately their statistics in the 30-60 Jupiter mass region is poor(the so called brown dwarf desert) since they rely only on transiting planets and theauthors do no consider the mass histogramme in this region. Earlier data suggested adip around 40 Jupiter mass (Sahlman et al. 2011, Udry 2010 – Figures 3 and 4) in themass histogramme. More statistics will come in the near future including radialvelocity data from rge ground and astrometric data from Gaia to see if a featurearound 40 Jupiter mass in the mass-radius diagramme exists or not. igure 3
Mass histogramme of low mass objects (Udry et al. 2010)
Figure 4
Low mass objects histogramme in the 20 – 75 Jupiter mass region(Sahlman et al. 2011)
A future improvement to separate the planet and brown dwarf populations will comefrom advanced observables, like the spectral type and species composition. They willhelp to constrain the formation mechanism of the object (accretion in a dust disk orcollapse of a gas cloud).At least one conclusion is clear, the former mass limit of exoplanets at 13 Jupitermass, correspinding to the triggering of nuclear burning of deuterium, is not relevantsince an object can be formed by dust accretion and acquire a final mass larger than13 Jupiter.There is a second, more factual, problem : the value of observables can be veryncertain. This is especially the case for objects detected by imaging where the masscannot be infered from radial velocity measurements but only from spectra andmodels. A typical example is the object 2M1207 (Chauvin et al. 2005) with a Jupitermass of 4 ± 1 Jupiter mass can be derived from its spectra. Indeed, in these cases thestar-planet separation is so wide that the semi-amplitude K = √ ( GM star / a planet ) ofthe stellar radial velocity variation induced by the planet motion is too low to bemeasurable. Even more : when the mass determination is as precise as a percent (incase of radial velocity or astrometric measurements), one faces the absurd situation ofa sharp mass limit. For example, what to do with objects like CoRoT-15 b with M =63.3 ± 4 M Jup ? A last problem, which we do not address here because the concerned population isgenerally supposed to be small, is the « intersteller wanderers », i.e. planets ejectedby dynamical interaction from a well formed planetary system.ConclusionAssuming that the definition of a planet and a brown dwarf is adopted according totheir formation mechanism, to separate the two populations is not an easy task. Anycatalogue contains necessarily a mixture of both populations. Since catalogues areuseful not only to list characteristics of objects but also to make statistics on thesecharacteristics, I recommend to take low constrains (for our case a mass limit as highas 60 Jupiter mass) on the properties used to define a sample, in order not to missinteresting objects. Modern softwares used to read electronic catalogues allow toeliminate easily objects from a catalogue which do not fullfil the criteria of definitionof each user, which is free to impose his own criteria.ReferencesAlonso R., Moutou C., Endl M. et al. 2014 Astron. & Astrophys. 567, 112Baraffe I., Chabrier G. & Barman T. 2010 Rep. Progr. Phys. 73, 016901Barge P., Baglin A., Auvergne M. et al. 2008 Astron. & Astrophys. 482, L17Bouchy F., Deleuil M., Guillot T. et al. 2011 Astron. & Astrophys. 525, A85Chauvin G., Lagrange A.-M., Dumas C. et al. 2004 Astron. & Astrophys. 425, L29Csizmadia Sz. 2016 Exploration of the brown dwarf regime around solar-like stars byCoRoT in « The CoroT legacy book », Editors: Annie Baglin et al., Publisher: EDPSciences, arviv:1603.07597 Hatzes A. & Rauer H. 2015 ApJ. Letters 810, L25Leger A., Rouan D., Schneider J. et al. 2009 Astron. & Astrophys. 506, 287Moutou C. & Deleuil M. 2015 C.R. Acad. Sci. Paris 347, 153Ray A. & Loeb A. 2015 ApJ. submitted, arXiv:1510.06418Sahlman J., Ségransan D., Queloz D. et al. 2011 Astron. & Astrophys. 525, A95Sanz-Forcada J., Ribas I., Micela G. et al. 2010 Astron. & Astrophys. 511, L8Schneider J., Dedieu C., Le Sidaner P. et al. 2011 Astron. & Astrophys. 532, A79Udry S. 2010