# An Unusual Gas Giant in a Red Dwarf System

The gas giant GJ 3512 b does not particularly stand out at first glance. About 30 light years from the Sun, it orbits its host star in 204 days, discovered by radial velocity methods by the CARMENES collaboration, which is all about finding planets around small stars. But look more deeply and you discover what makes this find provocative. GJ 3512 b turns out to be a gas giant with about half the mass of Jupiter, and small red dwarfs like this one aren’t supposed to host such worlds.

In fact, GJ 3512 b is at least an order of magnitude more massive than what we would expect from current theoretical models, making it an interesting test case for planet formation. Core accretion models assume the gradual agglomeration of material in a circumstellar disk, with small bodies banging into each other and growing over time until their gravity is sufficient to draw in an atmosphere from the surrounding gas. This gas giant defies the model, evidently having formed directly from the disk through gravitational collapse.

Image: Comparison of GJ 3512 to the Solar System and other nearby red-dwarf planetary systems. Planets around solar-mass stars can grow until they start accreting gas and become giant planets such as Jupiter, in a few millions of years. However, up to now astronomers suspected that, except for some rare exceptions like GJ 876, small stars such as Proxima, TRAPPIST-1, Teegardern’s star, and GJ 3512 were not able to form Jupiter mass planets. Credit: Guillem Anglada-Escude – IEEC/Science Wave, using SpaceEngine.org (Creative Commons Attribution 4.0 International; CC BY 4.0).

As for the host star, GJ 3512 has about 12 percent the mass of the Sun. The disks of gas and dust that surround such low mass stars are assumed to contain insufficient material to form planets like this. Consider: The Sun is 1050 times heavier than Jupiter, while the mass ratio between GJ 3512 and GJ 3512 b is 270. A much more massive debris disk would be needed to build this planet the conventional way, but if a disk of more than a tenth of the stellar mass is present, the star’s gravity cannot keep the disk stable. Gravitational collapse can then occur, as in star formation, but no disks this massive have been found around young red dwarf stars. Is this new exoplanet evidence that such disks can indeed form and be productive?

Things get even trickier when we consider other planets in the same system. At least one other planet is thought to exist, and the elliptical orbit of GJ 3512 b offers evidence for the gravitational effect of a possible third planet just as massive, one that may have been ejected. So now we have a small red star that would have needed to produce multiple massive planets, taking us well beyond current models. In a paper on this work from researchers at the Max Planck Institute for Astronomy, the University of Lund in Sweden and the University of Bern, the authors argue for gravitational disk collapse as the only viable method of formation.

“Until now, the only planets whose formation was compatible with disk instabilities were a handful of young, hot and very massive planets far away from their host stars,” says Hubert Klahr, who heads a working group on the theory of planet formation at the MPIA. “With GJ 3512 b, we now have an extraordinary candidate for a planet that could have emerged from the instability of a disk around a star with very little mass. This find prompts us to review our models.”

Image: Visualisation of the radial velocity (RV) measurement time-series and residuals obtained with CARMENES. Panel a illustrates how the RV of GJ 3512 (vertical axis) changes with time indicated in days since 8 December 2014, 12:00 p.m. UT (Universal Time, horizontal axis). HJD stands for Heliocentric Julian Day. Both the visual (blue symbols) and the infrared (red symbols) channels agree well. The black solid curve is the best orbital fit to the data. After subtracting the contribution of GJ 3512 b, panel b shows the residual, which indicates the presence of a long-term period hinting to a second planet. Panels c and d depict the residuals of the best overall orbital fit for the two CARMENES channels. Credit: Morales et al. (2019)/MPIA.

Considering pebble accretion vs. gravitational instability of the disk around this star, the scientists must look at a time early in formation when the disk was still massive relative to the star. The authors were unable to model an accretion process that would explain this system. But the competing model of gravitational instability results in a disk that is gravitationally unstable in a range of viscosities and surface densities at distances below 100 AU. From the paper:

The estimated masses of the fragments formed are less than that of Jupiter, consistent with the mass of GJ 3512 b. Except for unrealistically low values of a [disk viscosity], fragmentation of the disk occurs at radii of ≳10 au, so the planets must have migrated a substantial distance from their formation locations to their present positions. This is possible given the large mass of the disk with respect to the planet, and is often seen in numerical simulations of disk fragmentation. For realistic viscosity a > 0.01, disk fragmentation typically occurs at radii of a few tens of au, and the total disk mass within this radius is ~30 MJ. Disks cannot extend too far beyond this fragmentation radius, because the total disk mass would become extremely large (up to 1 M within 100 au). Thus, the planetary system around GJ 3512 favors the gravitational instability scenario as the formation channel for giant planets around very-low-mass stars.

If you’re looking for a comparison, consider TRAPPIST-1, a star with many of the characteristics of GJ 3512. Here we have seven planets with masses equal to or less than the mass of the Earth, and the ‘bottom up’ accretion model fits with observation. But GJ 3512 b all but forces us to look at models where the planet forms directly from gravitational collapse in the disk. We’d still like to know why GJ 3512 b hasn’t migrated closer to its star, an indication that the mysteries of this system may prove fodder for a great deal of future analysis.

The paper is Morales, et al. “A giant exoplanet orbiting a very low-mass star challenges planet formation models”, Science 27 September 2019 (abstract).

• Michael C. Fidler September 30, 2019, 9:26

Binary brown dwarfs and red dwarfs are relatively common and the same process should also form this type of system. The other possibility is a rough planet capture, but I think we are depending too much on models and as in everything in nature chaos rules. As the engineer said bumble bees cannot fly, nonlinear systems have the ability to jump into other states and we will find planet formations and planets themselves will have the greatest variety of anything man has ever found. No manner how hard you try the square peg will not go in the round hole!

• Harry R Ray September 30, 2019, 10:03

MORE headaches for planetary formation theorists: ArXiv:1909.11246 “Jupiter’s composition suggests its core assembled exterior to the N2 snowline”. by I Karin, Robin Wadsworth Oberg.

• wdk September 30, 2019, 12:03

While the mass of the exoplanet is large relative to the primary, it appears that the theory of core accretion is “reasonable” and probably not in danger of being thrown out the astrophysical window. So, it would seem that we might look for reasons for encountering such an outlier, especially since it has an eccentric path.

Encounter is a key word. A hunch would be that this system was once part of a binary and when it broke up, GJ 3512 inherited a large planet from a larger star. But when I started looking at particulars for the planet ( e= 0.4356 and a = 0.3380 AUs), I believe I see some discrepancies with the illustration above. Or else I am inferring that planet is depicted at periastron rather than at semi major axis radius.
But with the eccentricity from the Science article, it would come in to 0.1907 AU and go out to 0.4852AU. One wild possibility: maybe GJ 3512 had a normal sized planet going in one direction and b was going in the other, their paths
crossed, momentum was exchanged, the original was ejected and B dropped in. Three or four body conjectures wait in the wings to solve all kinds of problems.

• Michael C. Fidler September 30, 2019, 16:41

Some variable stars exchange mass between the two stars, could this be the core of a red or brown dwarf companion that’s orbit has widen after exchanging most of its mass with the primary.

• Geoffrey Hillend September 30, 2019, 16:19

I don’t think that disk instability rules out pebble accretion since we don’t know how the pebbles are made. They might not be made from fragmentation, but built up from dust. If they are, then maybe it both pebble accretion and disk instability contribute to forming gas giants. The pebble accretion model is supposed to predict a fast birth of gas giants from the protoplanetary disk like disk instability. I like the gravitational disk instability idea since it would explain gas giant exoplanets being formed quickly without fragmentation and migration to a point near their stars. It does kind of rule out the core accretion model which needs a lot more time with collisions of bodies to form planets like our rocky, inner planets.

• Enzo September 30, 2019, 18:02

Off topic and not sure if Paul was going to report about the NIAC symposium. Agenda here :
https://www.nasa.gov/sites/default/files/atoms/files/niac_2019_symposium_agenda_huntsville.pdf

Videos here :
https://livestream.com/viewnow/NIAC2019

My preferred ones this year are :
1) DUET and MOST, Ditto’s unusual diffraction machines. Long, more detailed video about concepts related to the latter here :
2) The KST : 1 Km inflatable telescope and/or a 1,000 km array of 40 m inflatable, of lower precision, mylar telescopes using intensity interferometry.
3) PuFF a propulsion system that uses and electromagnetic pulse to implode uranium/deuterium pellets and cause fission/fusion.

• wdk October 1, 2019, 12:41

MCF,

Since we have discussed this type of thing before I think I follow you.
Or else can imagine a possible event. that being that the red dwarf was the result of a previous merger of two sub hydrogen fusion brown dwarfs.

But I’m not sure that would solve anything, save to say that another less than Jupiter mass object happened to be in the vicinity.
This would assume that a lot of brown dwarfs are out there and becoming engaged in complicated interactions, including some in our neighborhood. But proving or disproving such an idea is difficult due to their intrinsic dimness. If mergers of brown dwarfs occurred and some gained main sequence status, would that result in an observable flare?

Thinking about the other form of mass exchange, assuming that the red dwarf was stripped of outer layers, it would seem that the proximity of the larger star would have to be very close. And I suspect that the process would be dissipative of energy and that the smaller star would be drawn closer to the larger. So then why would it vanish from the scene and a jovian size planet remain?

Planets retrograde with respect to each other would be possible if there were once another star in a binary or a close passage. This event also lacks substantiating evidence. but attribution is easier.

• Michael C. Fidler October 2, 2019, 14:36

This would be a contact binary, they are usually found in more massive stellar systems.

http://www.vikdhillon.staff.shef.ac.uk/seminars/lives_of_binary_stars/types.gif

http://www.vikdhillon.staff.shef.ac.uk/seminars/lives_of_binary_stars/masstrans.html

But found this M dwarf contact binary article that may be what has actually happened in the gas giant M dwarf system!

SDSS J001641-000925: THE FIRST STABLE RED DWARF CONTACT BINARY WITH A CLOSE-IN STELLAR COMPANION.

Abstract
SDSS J001641-000925 is the first red dwarf contact binary star with an orbital period of 0.19856 days that is one of the shortest known periods among M-dwarf binary systems. The orbital period was detected to be decreasing rapidly at a rate of $\dot{P}\sim {8}\,{\rm s}\,{\rm yr}^{-1}$. This indicated that SDSS J001641-000925 was undergoing coalescence via a dynamical mass transfer or loss and thus this red dwarf contact binary is dynamically unstable. To understand the properties of the period change, we monitored the binary system photometrically from 2011 September 2 to 2014 October 1 by using several telescopes in the world and 25 eclipse times were determined. It is discovered that the rapid decrease of the orbital period is not true. This is contrary to the prediction that the system is merging driven by rapid mass transfer or loss. Our preliminary analysis suggests that the observed minus calculated (O–C) diagram shows a cyclic oscillation with an amplitude of 0.00255 days and a period of 5.7 yr. The cyclic variation can be explained by the light travel time effect via the presence of a cool stellar companion with a mass of M 3sin i’ ~ 0.14 M ☉. The orbital separation between the third body and the central binary is about 2.8 AU. These results reveal that the rarity of red dwarf contact binaries could not be explained by rapidly dynamical destruction and the presence of the third body helps to form the red dwarf contact binary.

https://iopscience.iop.org/article/10.1088/2041-8205/798/2/L42/meta

• wdk October 3, 2019, 11:53

MCF,

Who would have thought! As yet I don’t grasp all the ins and outs, but sounds like our deliberations might be on to something. At the very least, while elsewhere there is examination of circum-stellar disk mechanics, it might pay off to look at effects related to close binaries too.

In the paper you cited, if I understand it correctly, the jovian planet’s presence seems to slow the merger process. Best I can picture that would be if the center of mass of the 3 body system would be shifted ( e.g. as with the sun and Jupiter) keeping the wins from finishing off the merger. Now, in contrast, GJ3512 has not been reported to be a contact binary, so far as I know as yet, and I don’t think it has been suggested. In the original paper there might be an estimate of the age of the star. But if there were, we would probably have to examine implications of a merger event. Premerger, two brown dwarfs could sit on their original metals content interminably and then start to produce helium and other nuclei after the fusion process is switched on. If the system had an association with a known cohesive cluster, maybe that would be a clue too.

I’ll try to look into some of these leads.

Best regards,

• Laintal October 1, 2019, 21:42

Hi Paul

Very interesting, I’m glad you had the time to write this up.
It seems like a very interesting planet I had trouble following the link to the paper?

Cheers Laintal

• Paul Gilster October 2, 2019, 7:24

Sorry about the broken link, but I think I’ve got it fixed now.

• Laintal October 2, 2019, 17:53

Thanks

I Did find the link the paper is about 16MB in size! but only a few pages a quiet day in the office to I’m about to read,

Also enjoyed the latest post on the subject too

• Daniel October 2, 2019, 12:48

What about Gliese 876 planetary system? it’s more unusual and haven’t been talk a lot about it.

• Harry R Ray October 4, 2019, 9:24

Gliese 876 is an early(M0-M4)type M class star(M3)whereas Gliese 3512 is a late(M5-M10)type M class star(M5.5, like Proxima Centauri is). The current theory can be used to describe formation of gas giants around early type M dwarf stars, but NOT around late type M dwarf stars.

• andy October 2, 2019, 16:11

It’s not entirely clear to me how unusual this planet is. GJ 3512 b is in a similar regime to several of the candidate planets found by gravitational microlensing, although the latter have far larger uncertainties on the properties of the planets and the host stars.