Panel Discussion on the Brown Dwarf { Exoplanet Connection

Mem. S.A.It. Vol. 84, 1154 c SAIt 2013 Memorie della

Panel discussion on the brown dwarf { exoplanet connection

D. J. Pinﬁeld1, J.-P. Beaulieu2, A. J. Burgasser3, P. Delorme4, J. Gizis5, and Q. Konopacky6

1 Centre for Astrophysics Research, University of Hertfordshire, Hatﬁeld, UK, AL10 9AB, UK, e-mail: [email protected] 2 Institut d’Astrophysique de Paris, UMR7095, CNRS, Universite´ Paris VI, 98bis Boulevard Arago, 75014 Paris, France 3 Center for Astrophysics and Space Science, University of California San Diego, La Jolla, CA 92093, USA 4 UJF-Grenoble 1/CNRS-INSU, Institut de Plantologie et d’Astrophysique de Grenoble (IPAG) UMR 5274, 38041, Grenoble, France 5 Department of Physics and Astronomy, University of Delaware, Newark, DE 19716, USA 6 Dunlap Institute for Astronomy and Astrophysics, University of Toronto, 50 St. George Street, Toronto, ON M5S 3H4

Abstract. We present a summary of the panel discussion session on the brown dwarf exoplanet connection, at the Brown Dwarfs Come of Age conference in Fuerteventura. The discussion included an audience vote on the status (planet or brown dwarf) of a selection of interesting objects, as well as a vote on the current components of the IAU deﬁnition separating planets and brown dwarfs, and we report the results. In between these two opinion tests we discussed a set of key questions that helped us explore a variety of important areas, and we summarise the resulting discussion both from the panel and the conference audience.

Key words. Stars: brown dwarfs, planetary systems

1. Introduction also ambiguity from the very start. The relatively unambiguous brown dwarf Teide 1 has a Eighteen years on from the discovery of the mass of 40-70 MJup (just below the Hydrogen first brown dwarfs and exoplanet, the Brown burning limit) and is a free-floating M8 dwarf Dwarfs Come of Age conference provides an in the Pleiades cluster (Rebolo et al. 1995). opportunity to re-examine and discuss the con- In contrast the first exoplanet, 51 Peg b, has a nection between these two exotic populations, likely sub-Jupiter mass (∼0.5 MJup) and orbits and consider how we might better understand at a tight (hot) 0.05 AU separation (Mayor & their nature now that the field has matured. The Queloz 1995). Early ambiguity appeared in the very first discovered brown dwarfs and exo- form of Gl229B (Nakajima et al. 1995), with planets showed some clear disconnection, but a mass (20-50 MJup) much closer to the giant Send offprint requests to: D. Pinfield planet regime, a 40 AU orbit, and a spectrum Pinfield et al.: The brown dwarf – exoplanet connection 1155 showing strong similarities with that of Jupiter. 1.2. Structure of the panel discussion So the distinction between brown dwarfs and exoplanets has been blurred since the earliest Our discussion session took the following discoveries, and we would like to explore de- structure, which we describe in the rest of this veloped/growing opportunities for clarity. write-up. – To canvass opinion from the community, 1.1. Where do things stand? - the IAU and more generally encourage clarity over ambiguity we began with a vote on a se- definition ries of interesting objects - brown dwarf or As part of the re-classification exercise that planet? made Pluto a “dwarf planet” following new – We then presented five key questions to Kuiper-Belt discoveries, the IAU Working foster discussion, with panel and audience Group on Extrasolar Planets (WGESP) agreed member participation. on a “Planet” definition in a position statement – And at the end we took a vote on the main first issued on 28 Feb 2001 (and last modified elements of the existing IAU definition, to 28 Feb 2003; see Boss et al. 2003). The state- elicit an overview of how current opinion ment consists of the following points: maps onto existing definition. – Objects with true masses below the limiting mass for thermonuclear fusion of 2. Community view - individual deuterium (currently calculated to be 13 objects Jupiter masses for objects of solar metal- To stimulate discussion on what properties licity) that orbit stars or stellar remnants might best distinguish brown dwarfs from ex- are “planets” (no matter how they formed). oplanets, we asked the audience to vote on the The minimum mass/size required for an ex- planethood of a sample of 11 systems which trasolar object to be considered a planet span the brown dwarf/exoplanet “regimes”, should be the same as that used in our Solar covering a variety of masses, primary types System. (including no primary), ages and separations. – Substellar objects with true masses above This was a binary selection - either an object is the limiting mass for thermonuclear fusion or is not a planet. There were 55–60 votes per of deuterium are “brown dwarfs”, no mat- system, implying a sampling error of 4–6%. ter how they formed nor where they are lo- Tables 1 and 2 summarize the results (sys- cated. tem parametres and polling respectively) in – Free-floating objects in young star clusters order of decreasing “planetness”. The bench- with masses below the limiting mass for mark companions to HR 8799 and that of thermonuclear fusion of deuterium are not Gliese 229 were extremities on the planet- “planets”, but are “sub-brown dwarfs” (or brown dwarf spectrum; these systems have whatever name is most appropriate). similar separations and possibly ages, but the The WGESP noted that these statements clear distinguishing factors were companion are a compromise between definitions based mass and mass ratio, the latter differing by an purely on the deuterium-burning mass or on order of magnitude. The presence of more than the formation mechanism, and as such did not one substellar object and a debris disk in HR fully satisfy anyone on the working group. 8799 but not Gliese 229 may also have a signif- Very young free-floating objects are not icant impact on the outcome of the planetness just found in young clusters of course, and have vote. The most split vote came for the com- been identified in the field e.g. as members of panion to WD 0806-661, which is even within kinematic moving groups. As such the final the margin of error; the companion to Corot-3 point is naturally extended to include all free- and the “free-floating” planetary mass objects floating objects, so we assume this modifica- CFBDS 2149−04 and those in σ Orionis were tion hereafter. all within 2.5 σ of the margin of error. The last 1156 Pinfield et al.: The brown dwarf – exoplanet connection of these reflects the even split on whether such desert separates the stellar and planetary pop- unbound objects should be considered planets ulations. Brown dwarf researchers, however, in a later poll (below). are now working in regimes where these def- We examined the correlation between frac- initions are not very helpful. Direct imaging tion of planet votes to the systems’ various searches are detecting objects above and be- physical parameters. The most significant fac- low the deuterium-burning limit at wide sep- tors were object mass and separation from arations from the primary. These may or may primary star, both of which had marginally not have anything to do with formation in a significant (1.6 σ and 1.2 σ) negative correla- disk. tions (larger values decreased “planethood”). The panel discussion emphasizes the prob- Primary mass, system mass ratio and presence lems of formation-based and mass-based defi- of a primary were insignificantly correlated. A nitions. The HR 8799 system had wide support linear regression of these data provides a sim- as a planetary system, as it consists of multiple but probably unreliable prediction of this ple low mass objects that seem to have formed audience’s planethood for a given object based in a disk, though not necessarily through core on these parameters: accretion. If disk formation is not important, then 2M1207b seems to meet most planet definitions. Current work suggests that forma- %Planethood = 54.8 + 9.78MA − 2.22MB tion scenarios are not a simple binary choice. Theorists at this conference predicted a new +57.5q − 11.9 log10 a population of objects that form through disk in- +36.8C (1) stabilities. These may lie in the range 3 to “42” jupiter masses. Many of these are predicted to where M is primary mass in solar masses, M A B be ejected and become isolated, “free-floating” is companion mass in Jupiter masses, q is the objects. Some observers of star forming re- mass ratio M /M , a is the separation in AU B A gions reported evidence that the mass function and C = 1 if a primary is present or 0 if not. increases below the deuterium-burning limit, This is accurate to within 9% for the most sig- which if confirmed, also suggests a new for- nificant factors mation mechanism. There was a discussion of the latest microlensing evidence of a new population of free-floating or wide compan- %Planethood = 105 − 2.30MB ion objects below the deuterium-burning limit. −12.1 log10 a (2) Ultimately, the panel wondered if the disk instability should be considered a planetary or a is accurate to within 16%. stellar formation mechanism. There was spirited challenge from some 3. Discussion questions audience members on whether there are truly fundamental differences between planets and 3.1. How might we define the brown dwarfs. Perhaps these limits are arbi- fundamental differences (e.g. trary and astronomers should be more open- properties, formation) between minded. For example, if we lived on an earth- brown dwarfs and giant planets? like planet orbiting Gliese 229A, would we consider Gliese 229B a planet? Perhaps the de- The voting brought out that even at this spe- sire to define planets to be like those in our cialized conference there is a wide range of own system is just an emotional attachment, views on definitions, or rather what differences a psychological bias that astronomers should are fundamental. The IAU definition, which strive to overcome. Panelists warned that this is largely influenced by the Solar System, psychological bias is also a sociological effect. worked well in the radial velocity and tran- An object called a “planet” might attract fund- siting planet regimes, where the brown dwarf ing that the same object called a “brown dwarf” Pinfield et al.: The brown dwarf – exoplanet connection 1157

Table 1. Parametres of the systems being polled for planet/brown dwarf status.

Companion(s) to Primary Companion Mass Ratio Separation Age (M /SpT) (MJup/SpT) (MB/MA) (AU) HR 8799 1.5 (A) 5–7 (late-L) 0.4% 15–70 30 Myr MOA-2010-BLG-073L 0.16 (M) 9–13 (...) 6.5% 1.2 few Gyr 2MASS 1207−39 0.03 (M) 3–10 (late-L) 20% 41 8-12 Myr κ And 2.5 (B) 12–15 (...) 0.5% 55 ≈30 Myr Corot-3 1.4 (F) 22 (...) 1.6% 0.06 few Gyr WD 0806-661 1.5 (DQ) 5–9 (Y) 0.4% 2500 1.5 Gyr σ Orionis members ... 3+ (L & T) ...... 2–4 Myr CFBDS 2149−04 ... 7–9 (T7) ...... ≈100 Myr? 2MASS 0103-55AB 0.36 (MM) 12–14 (L0) 4% 84 ≈30 Myr Ross 458AB 0.6 (MM) 6–11 (T8) 1–2% 1100 0.2–0.8 Gyr Gliese 229 0.6 (M) 20–50 (T7) 5–8% 44 <1 Gyr?

might not. It was also pointed out that two talks tematics exist. Furthermore, there was discus- at the conference suggested that the current sci- sion of the difficulty of measuring such signa- entific literature is biased by researchers prior- tures even within the planets in our own Solar itizing publishing planets over brown dwarfs. System, casting doubt on our ability to do so for much more distant objects. However, these signatures may represent one way of distin- 3.2. How should we differentiate guishing between brown dwarfs and planets if between brown dwarfs and planets a formation scenario definition is adopted. observationally? It was proposed that an observational sig- During the discussion, a number of observa- nature could be simply whether the object or- tional signatures or characteristics were posed bited another star or was free floating. The sug- that could potentially be used to distinguish be- gestion was made that perhaps four categories tween brown dwarfs and planets. First, spec- were necessary, with brown dwarfs orbiting an- troscopic signatures were brought up as a pos- other star being called something slightly dif- sible probe of formation scenario. It has been ferent than single brown dwarfs. This would suggested that objects forming in a circumstel- then be somewhat analogous to the planet vs. lar disk via core accretion would have mea- planetary mass object nomenclature that is of- surably enhanced metallicity (e.g. Fortney et ten used. However, there was discussion that al. 2008). Further, it has been proposed that such a scheme might be too confusing given the condensation of various species from gas that many of the “first” brown dwarfs were dis- phase into solid phase at different locations in covered orbiting other stars (e.g. Nakajima et a disk could lead to observational signatures al. 1995), and had become in some sense pro- such as an elevated carbon-to-oxygen ratio totypes for entire spectral sub-classes. (e.g. Oberg¨ et al. 2011). Indeed, such enhanced Somewhat along those lines, another obser- ratios have already had tentative detections vational signature suggested for distinguishing (e.g., Madhusudhan et al. 2011; Konopacky brown dwarfs and planets was mass ratio rather et al. 2013). However, there was some concern than mass. If the mass of the object is already about the fidelity of atmospheric models that considered via the deuterium burning limit to would be used to measure the abundances and be the demarking line between planets and ratios of species. These models have not been brown dwarfs, then it is fairly straightforward tested against the copious brown dwarf spec- to apply instead the mass ratio between the tra that are available to determine whether sys- substellar object and its host star. It was sug- 1158 Pinfield et al.: The brown dwarf – exoplanet connection

Table 2. Polling results for planet/brown dwarf sample.

Companion(s) to Planet Not Planet (%/#) (%/#) HR 8799 91% (50) 9% (5) MOA-2010-BLG-073L 83% (50) 17% (5) 2MASS 1207−39 67% (40) 33% (20) κ And 66% (41) 34% (21) Corot-3 58% (36) 42% (26) WD 0806-661 47% (29) 53% (33) σ Orionis members 42% (25) 58% (35) CFBDS 2149−04 41% (25) 59% (36) 2MASS 0103-55AB 40% (25) 60% (37) Ross 458AB 33% (20) 67% (40) Gliese 229 11% (6) 89% (50)

gested that 13 MJup was, though fairly straight- A final brief suggestion of an observable forward in terms of a workable definition, a signature was density. It was mentioned that very arbitrary limit because objects straddling COROT-3b, detected via the transit method, this border have very similar interior physics. has a density that necessitates it having a dense This is in contrast to the hydrogen burning core (Deleuil et al. 2008). Such measurements limit definition for a brown dwarf, where the are often possible in the case of transiting ob- interior physics is quite different. A mass ratio jects. This is another line of evidence for for- definition could therefore account for objects mation via a process like core accretion, and like 2M1207b (Chauvin et al. 2004), which perhaps planetary definition. likely formed more like a star but certainly has In summary, the general consensus seemed a planetary mass. Hence, mass ratio perhaps to favor a formation rather than a mass defi- also traces formation. nition, and the discussion focused on whether formation was observationally distinguishable. A final suggestion was that system architecture could help distinguish whether an ob- 3.3. What are the priorities for improving ject formed like a “planet” or more like a theory to differentiate brown dwarfs star. The question was raised whether someone and planets? looking at our Solar System would be able to determine whether Jupiter was a brown dwarf It was felt that this topic could be usefully or a planet. It was noted by the panel that the separated into two main areas; what would architecture of our Solar System, with multi- observers like in the way of a shopping list ple planets orbiting in the same direction with from the theorists, and how can the commu- very low eccentricity, was an extremely com- nity use benchmark systems (ultra-cool dwarfs pelling case for formation in a common disk. whose physical properties are constrained at It was also mentioned that this was why the some level in a relatively model-independent objects orbiting HR 8799 (Marois et al. 2008) way) to foster our ability to measure the prop- were overwhelming considered to be planets erties of brown dwarfs and giant exoplanets in rather than brown dwarfs in the vote (92%). general. Though such a criteria would be difficult for The measurement and study of individual cases where there are not multiple objects or- element abundances was felt to be of funda- biting a given star, it is at least a fairly obvious mental importance, in particular the possibil- signature of formation when available. ity of measuring atmospheric deuterium. A Pinfield et al.: The brown dwarf – exoplanet connection 1159 deuterium test could provide a physically ro- though many potentially suitable clusters do bust means of separating brown dwarfs and not have metallicity constraints at this time. exoplanets (akin in some sense to the lithium However, there could be a strong science case test at the other end of the brown dwarf mass for deep searches of many more young clus- regime; Nelson et al. 1993). This would re- ters for substellar members, though a note of quire some individual element abundances as caution was added that even an expanded set input to the models, and will be challeng- of young clusters may not actually span the re- ing to implement due to the relative under- quired metallicity range. abundance of deuterium. However, feedback from the modelling community was upbeat. They felt that the observational challenge of 3.4. Does the field population contain measuring deuterium rather out-weighed the free-floating planets, and what can theoretical challenge. It was also commented we learn from them? that wide binary systems containing a star and an ultra-cool companion could provide an ex- The existence of free-floating planets in the cellent means of making individual abundance field is backed by several independent lines studies. This has also been noted in connection of evidence. Firstly, their young counterparts with the discovery of BD+012920 (Pinfield et in young clusters and stellar formation regions al. 2012). have been identified by several imaging sur- This brought the discussion naturally onto veys, down to a few Jupiter masses (see for benchmark wide binary systems, and it was instance Zapatero Osorio et al. 2002; Burgess noted that young systems containing low grav- et al. 2009; Marsh et al. 2010; Haisch et al. ity ultra-cool companions would provide the 2010; Ram´ırez et al. 2011; Scholz et al. 2012). best giant planet analogues. Targeted and If free-floating planets are currently forming in database searches for ultra-cool companions stellar formation regions, it is reasonable to as- around young stars, particularly the (numer- sume they have been forming during the previ- ous) young M dwarfs, were advocated. And ous Gyrs and that a corresponding population it was commented on that in general what of field free-floating planets does exist. is required is a large number of benchmark Secondly, their is some direct evidence for objects, populating a grid of parameter-space the existence of field objects in the plane- (Teff/log g/[M/H]). The panel also noted that tary mass range, backed by a few observa- rotation could play an important role in the tions, notably of cool field brown dwarfs which formation and nature of atmospheric dust, would be of planetary mass for a significant and there was some concern expressed that L fraction of their estimated age range (see for dwarfs seen in young clusters seem to show a instance Burningham et al. 2009; Liu et al. range of properties (very red spectra, triangular 2011; Kirkpatrick et al. 2012). The case of H-band) even at the same age and composition. CFBDSIR2149 (Delorme et al. 2012) is par- However, possible contamination amongst the ticularly interesting because its probable mem- cluster sequence was suggested as a source of bership to a young moving group provides a confusion, and clearly more detailed studies of much tighter constraint on its age and therefore the young cluster populations are important. on its mass, robustly anchoring it in the plane- Another issue raised was that planet atmo- tary mass range. Wide field surveys for brown spheres are generally believed to be enriched dwarfs consequently provide strong evidence compared to those of brown dwarfs, leading that at least a few free-floating planets do exist to a desire for metal-rich benchmark systems in the field. as well as low gravity. One avenue towards Thirdly, microlensing surveys have uncov- this goal was considered in the form of an ex- ered what could be a huge population of field panded study of young open clusters. Figure 2 free-floating planets, up to 1.8 giant free- from Pinfield et al. (2006) was considered to floating planets per star in the Milky way illustrate the potential scope for such studies, (Sumi et al. 2011). It is to be noted this is not 1160 Pinfield et al.: The brown dwarf – exoplanet connection compatible with the relative scarcity of young – GAIA: The very large number of ultra- free floating planets uncovered in the wide field cool dwarfs now contained within large- surveys of the field and of stellar formation re- scale near- and mid-infrared surveys can gions. Microlensing observations also back the link up with the GAIA population in a existence of an old population of free-floating very important. This link-up will produce a planets, but with a different frequency com- vast collection of benchmark brown dwarfs pared to the young ones. However a fraction in wide binary systems whose properties of the microlensing detections could be caused can be known with unprecedented accu- by bound planets that are distant enough from racy. Ultra-cool dwarf distances will be in- their host star so that the signal of the host star ferred from the primary stars, with accura- does not appear in the microlensing event. On- cies of ∼0.1% limited only by the unknown going high angular resolution observations are separation of the companions. The com- underway to check whether or not these free prehensive SED-fit properties of the GAIA floating candidates indeed do not have a com- primaries can be used to constrain the panion star. benchmark companions, yielding metallic- In the case that most of the microlensing ity and age data for many thousand systems detections are caused by such distant planets, Indeed, this sample will be so large that results from microlensing and imaging surveys it will be important to be selective, study- could be reconciliated toward the existence ing only the systems with the best con- of a small population of free floating plan- straints on their properties (including e.g. ets. Conversely, the observed scarcity of plan- those with subgiants or high-mass white etary mass objects in imaging surveys could dwarfs as primaries, where the age and/or be compatible with a huge population of free- metallicity measurements are best). floating planets, if the atmosphere and evolu- – Euclid: Although Euclid has primarily tionary models (that constrain the mass of im- cosmological and extragalactic science aged objects) would significantly overestimate drivers, it will yield very important ultra- the luminosity of planetary mass objects. cool dwarf results as well. It will survey the sky to unprecedented depth ( 24.5 in 1 op- tical and 3 NIR bands) over a very large 3.5. How can future surveys help area (¿15000 sq degs), and provide diffrac- address these questions? tion limited spatial resolution. Results-wise this will increase the number of resolved The panel and audience had a variety of dis- ultra-cool binaries by more than an order of cussion about future surveys. In the following magnitude, providing stringent constraints we give a summary of the surveys considered, on models. It is also expected to find signif- and key points. icant numbers of rare free-floating objects (free-floating planets, and pop III ultra- – Pan-STARRS and LSST: These large-scale cool dwarfs; Mart´ın et al. 2013a). surveys will identify brown dwarfs through – JWST: The mid-infrared (∼5 micron) capa- parallax, and explore the time domain for bilities of Nirspec will provide the means the local substellar population. Many vari- to spectroscopically characterise a high able objects will be identified, and it is fraction of the emitted flux from a wide likely that field brown dwarf eclipsing bi- range of ultra-cool dwarfs. And its multi- naries will be discovered. These will yield object spectroscopic mode (with a 3x3 ar- mass-radius data providing superbly quan- cminute of field of view) will be particu- tified benchmark systems to test atmo- larly amenable to cluster populations. sphere and evolutionary models. Indeed, – EChO: The Exoplanet Characterisation there is some chance that the first Pan- Observatory (Tinetti et al. 2012) is a pro- STARRS telescope (PS1) may reveal the posed ESA mission to measure extremely first such eclipsing system. accurate and stable 0.5-16 micron spec- Pinfield et al.: The brown dwarf – exoplanet connection 1161

tra of exoplanet hosts (with mainly tran- regarding the requirement for planets to orbit siting planets) from L2. EChO will mea- stars or stellar remnants. And in the final anal- sure emission and transmission spectra, ysis there was fairly strong support for the con- during secondary and primary eclipse re- tinued use of the Deuterium burning mass limit spectively. Optimal interpretation of these as a separator between brown dwarfs and plan- observations will only be fully facilitated ets. by the kind of empirical studies and cali- bration offered by current and near-future ultra-cool dwarf populations. 5. Closing – New opportunities: Warm Spitzer is offer- It seems a crucial point that disk instability ing major opportunities for very large-scale can lead to a range of substellar mass ob- programmes, that could be survey-based jects, including masses reasonably close to that or targeted. Also there is an opportunity of Jupiter, with some of these objects being to search for wide eclipsing brown dwarf ejected into the field. It thus seems impractical companions in Kepler. Such long period to imagine a planet-brown dwarf definition that systems await discovery, and could pro- is linked to formation in a disk but is based on duce invaluable mass-radius-metallicity. measurements of mass, mass-ratio, separation Also very low-mass dwarfs have already and/or orbital properties. been identified in the Kepler field of view Using a mass-based definition does allow (e.g. Gizis et al. 2011; Mart´ın et al. 2013b). one to place a fence around certain types of objects, such as those that are not undergo- 4. Community view - IAU definition ing fusion. However, indications are that such badging, while good at identifying objects that After discussion of the main questions, the are currently close to the detectability limits panel posed four additional questions to the of the latest instruments, does not reflect an conference to gauge opinion on the existing improved understanding of substellar popula- IAU definition (separating planets from brown tions. The existing “planet” label has impli- dwarfs). We asked; cations for telescope time awards, grant fund- 1. Is the Deuterium burning mass limit useful ing, and press coverage, though we are still in for separating brown dwarfs and planets? search of more meaning in the name. 2. Should another mass limit be involved in If a means was available to deter- the definition (e.g. ∼40 MJup)? mine/constrain interior structure observation- 3. Should planets be required to orbit stars or ally, then this might offer a far more mean- stellar remnants? ingful way to differentiate between sub-stellar 4. Should free-floating planetary mass objects objects. Formation in a disk in itself would in clusters (or in the field) be given a sepa- not then be the requirement for planet-hood, rate label (e.g. sub-brown dwarfs)? but rather we could chose to separate objects with cores and layered structure, from those The votes cast were as follows: that are simply balls of gas (more akin to low- 1. Yes=46, No=16 mass stars). In this type of definition disk in- 2. Yes=6, No=55 stability would lead to brown dwarfs whatever 3. Yes=32, No=30 their mass and location, while core accretion 4. Yes=7, No=55 would lead to planets. Such a definition would be rooted in formation and interior structure, The clearest decision was to reject the idea and the challenge would be to find the obser- of using an additional or alternative mass limit vational means to make this separation. to separate brown dwarfs from planets. Feeling Developing rigorous methods to constrain was also strong that there should not be a sepa- the physical properties of ultra-cool dwarfs rate label for free-floating planetary mass ob- will be crucial in the coming years if such jects. The conference was fairly evenly split goals are to be achieved, and benchmark sys- 1162 Pinfield et al.: The brown dwarf – exoplanet connection tems will be vital in this respect. Also the study Kirkpatrick, J. D., et al. 2012, ApJ, 753, 156 of detailed atmospheric chemistry should shed Konopacky, Q. M., Barman, T. S., Macintosh, light on the provenance of substellar objects B. A., & Marois, C. 2013, Science, 339, and give many clues to their internal structure. 1398 Having said this, we note that at this time Liu, M., et al. 2011, ApJ, 740, 108 the vote on existing IAU criteria shows that as- Madhusudhan, N., et al. 2011, Nature, 469, 64 tronomers at this conference quite like a simple Marois, C., et al. 2008, Science, 322, 1348 separation based on mass. So perhaps this has Marsh, K. A., Kirkpatrick, J. D., & Plavchan, its place for the time being. P. 2010, ApJ, 709, L158 Acknowledgements. Thanks to the conference orga- Mart´ın, E. L. 2013, in Hot Planets and Cool nizers and all those who contributed to the discus- Stars, ed. R. Saglia, (EDP Sciences, Les Ulis sion in Fuerteventura. Thanks also to Joana Gomes Cedex A) EPJ Web of Conf., 47, 15003 for vote counting, and to James Canty for recording Mart´ın, E. L., et al. 2013, A&A, 555, 108 the discussion. Mayor, M., & Queloz, D. 1995, Nature, 378, 355 References Nakajima, T., et al. 1995, Nature, 378, 463 Nelson, L. A., Rappaport, S., & Chiang, E. Boss, A. P., et al. 2003, in Brown Dwarfs, 1993, ApJ, 413, 364 ed. E., Mart´ın, (ASP, San Francisco), IAU Oberg,¨ K. I., Murray-Clay, R., & Bergin, E. A. Symposium, 211, 529 2011, ApJ, 743, L16 Burgess, A. S. M., et al. 2009, A&A, 508, 823 Pinfield, D. J., et al. 2006, MNRAS, 368, 1281 Burningham, B., et al. 2009, MNRAS, 395, Pinfield, D. J., et al. 2012, MNRAS, 422, 1922 1237 Ram´ırez, I., et al. 2011, ApJ, 740, 76 Chauvin, G., et al. 2004, A&A, 425, L29 Rebolo, R., Zapatero Osorio, M. R., & Mart´ın, Deleuil, M., et al. 2008, A&A, 491, 889 E. L., 1995, Nature, 377, 129 Delorme, P., et al. 2012, A&A, 548, 26 Scholz, A., et al. 2012, ApJ, 756, 24 Fortney, J. J., Marley, M. S., Saumon, D., & Sumi, T., et al. 2011, Nature, 473, 349 Lodders, K. 2008, ApJ, 683, 1104 Tinetti, G., et al. 2012, Experimental Gizis, J. E., Troup, N. W., & Burgasser, A. J., Astronomy, 34, 311 2011, ApJ, 736, L34 Zapatero Osorio, M. R., et al. 2002, ApJ, 578, Haisch, K. E., Barsony, M., & Tinney, C., 536 2010, ApJ, 719, L90