≡ Menu

GEMSS: Hunting for Red Dwarf Exoplanets

If I were a betting man, I would put some money on this proposition: The first detection of a potentially habitable planet will be made before the end of this decade, and the planet will be found around an M-class red dwarf. The method will doubtless be photometry, picking up the slight drop in light caused by such a planet transiting its star. A planet the size of our Earth will block about one percent of the stellar flux, as a recent paper points out, and a one percent photometric dip is quite detectible.

Animation of a transiting exoplanet

Image: An animation of a stellar transit around HD 209458. Credit: Transits of Extrasolar Planets Network. Although TEP concluded its work in 2001, you can still read about it online.

So the key is to find the right M-dwarf, with its planetary system lined up so that the hypothetical terrestrial world passes between its star and us. That’s no small challenge, but Paul Shankland (U.S. Naval Observatory), who is lead author on the paper mentioned above, is working with colleagues around the world on a new collaboration to monitor M-dwarfs, enhancing our chances of finding transiting worlds or, just as valuable, proving that transits are not possible in the systems studied.

The collaboration, called Global Exoplanet M-dwarf Search-Survey (GEMSS) takes in James Cook University (Queensland) and includes Greg Laughlin’s TransitSearch effort based at UC Santa Cruz). Here’s how Shankland describes its method:

Our modus operandi as mentioned elsewhere, is to operate a continuous-coverage web of small telescopes – to characterize our prolific red dwarf population’s propensity to have transiting exoplanets… We want to employ observers across the Earth in longitude in particular, to keep constant surveillance underway on a given target. Our initial plan is to test-run our USNO-JCU collaboration – and once that proves to run well, we will post campaign bulletins to expand the observational cadre – so stay tuned!

The GEMSS Web site, now coming online, appears at a time of unprecendented interest in red dwarfs, especially now that prospects for stable ecosystems in such a star’s habitable zone have improved. Within the last ten years we’ve learned that even tidally locked worlds may retain an atmosphere and generate a flare-protective magnetic field. Moreover, the scarcity of gas giants around the M-dwarfs surveyed by the California & Carnegie Planet Search leads to an interesting prospect.

For the core accretion theory of planetary formation says that a large mass rapidly acquires gas from the material around it, resulting in a gas giant. Start with a small star and such a mass may not have time to form before the surrounding gases have dissipated. One scenario, then, has M-dwarfs housing smaller, rocky worlds but not often gas giants.

That would make stars like Gl 876, which Shankland’s team has studied extensively, something of an anomaly. But even here, the two gas giants already detected are accompanied by a much smaller planet 7.5 times the mass of Earth (though too close to its parent to be habitable, evidently).

The effort now gearing up to find M-dwarf transits is exciting indeed, and ponder this: The slow fusion burn of these small stars gives them exceedingly long lives. Quoting Shankland again:

These stars are pretty small – with a mass & radius below a third of our Sun (to 0.08 solar masses, which are then classified as brown dwarfs) and a surface temperature below 3,500K. M dwarfs incorporate a slow P-P fusion (due to the temperature) to turn hydrogen into helium. This slow burn is likely to last billions to trillions of years (!) – and so they also burn dimly (1/10,000th the sun). This is a detractor to photometric work by GEMSS… But c’est-la vie.

A stable star that shines for a trillion years and does not brighten with age seems to create a remarkable chance for the development of intelligent life. And although that dim burn keeps us working within a stellar population roughly 20 parsecs from Earth, the small size of a red dwarf means that the ratio in mass between a terrestrial planet and its star is far greater than what we would find in solar systems like our own. That makes GEMSS’ work easier. With as much as 80 percent of all stars being red dwarfs, GEMSS moves into a fertile hunting ground indeed.

Comments on this entry are closed.

  • Paul Dietz February 5, 2007, 13:22

    A star that burns for trillions of years wouldn’t necessarily mean the planet would stay habitable for that long. For example, recycling of volatiles by volcanism would run down after radioactive isotopes (uranium, thorium, potassium being the major ones) had mostly decayed.

  • Larry Palkovic February 5, 2007, 17:49

    Any habitable planet around a red dwarf would be subject to ferocious tidal stresses, which, like Io, would generate tidal volcanism.

  • Paul Shankland February 5, 2007, 18:41

    Hi Larry-Curious why you would say that. It would certainly depend on the orbital dynamics of the system whether or not the tidal g’s would be ferocious, and there are combinations which would certainly not be ferocious – even tidally locked scenarios have been demonstrated to be tenable for providing a working HZ. I’d agree that a planet like Gl 876d might be under ‘alot of pain’, but I cannot think that larger orbits (albeit a far greater detection problem) would be a concern. And you said “Any”… So can you elaborate on your assertion about the level of stresses? Thanks! Paul

  • JD February 6, 2007, 0:28

    Interesting, has anyone done work on tidal stresses of a roughly terrestrial sized planet, within the habitable zone for variouse masses of K throught M? There’s probably some area within that range where tidal stresses would have effect. It seems plausible that if tidal lock was in effect there would concievably be at least a low degree (not anywhere as great as Io) of warming induced by tidal stress.

  • andy February 6, 2007, 4:33

    Ordinarily, the tidal forces would work to synchronise the rotation and circularise the orbit. Once these have been achieved, the tidal forces no longer heat up the body. Io is volcanic because it is in a 1:2:4 resonance with Europa and Ganymede, which means that moon-moon interactions keep pumping its orbital eccentricity. A red dwarf planet will not experience severe tidal volcanism unless a similar mechanism keeps its orbit eccentric.

  • Edg Duveyoung February 6, 2007, 9:34

    Do planets come in bunches? It seems that if a proto-star has the material to spit out a planet, happenstance would favor this material gathering together in “various spots” instead of one spot.

    So, isn’t finding one planet — no matter how close it is to it’s star or whatever other aspects about it that preclude life — a hopeful sign that other planets will be in the system spaced out in such a way that the “life zone” would have a decent chance at having a planet in it?

    Only a few multiple-exoplanet systems discovered so far out of the couple hundred spotted, but even that small percentage bodes for Sagan-billions of possible-life planets in our galaxy alone, er, yes?


  • Adam February 6, 2007, 22:55

    Hi Edg

    Planets probably do come in bunches. One of the recent papers put out by Marcy & Butler’s crowd has hinted at more planets in known single-planet systems. The main reason, I’d suggest, why so many systems seem single is that the other planets are probably out in sensible gas giant orbits that take more than the 12 years we’ve been watching them continuously for. With 20-30 years of data I’m sure plenty of multiple planet systems will appear in the data. We’ve seen only the tip of an iceberg.

  • Stephen February 7, 2007, 18:54

    Even if i get a tracking mout for my 10 inch reflector, and a high speed photometer (or even a good CCD camera), i doubt i’m going to get usable transit data for M-Dwarfs. So, since we need a network of these, what’s the minimum sized scope? What’s the cost?

    Can COROT (now in space) detect M-Dwarf stars?

  • David February 7, 2007, 19:52


    This kind of photometry is doable for the brighter M dwarfs with a 10 inch telescope, a solid mount and a good CCD camera. One would also need a computer and the relevent software to do the data reduction. The needed precision is about 1%.

  • Administrator February 7, 2007, 20:18

    Good comment, David, and let me add for Stephen that he should check out Greg Laughlin’s TransitSearch site, where information about how amateurs are contributing to the hunt for exoplanets (and their equipment) can be found. The address is: http://www.transitsearch.org.

  • Administrator February 7, 2007, 20:19

    One thing I forgot to add in my last post: Stephen asks whether COROT can work with red dwarf stars, and the answer is that many of its targets are indeed M-dwarfs.

  • ljk March 14, 2007, 12:55

    Astrophysics, abstract

    From: Peter R. McCullough [view email]

    Date: Tue, 13 Mar 2007 18:25:29 GMT (140kb)

    The Life Cycle of an XO Planet and the Potential to Detect Transiting Planets of M Dwarfs

    Authors: Peter R. McCullough, Christopher J. Burke

    Comments: 8 pages, 2 figures. To appear in the ASP Conference Series: “Transiting Extrasolar Planets Workshop” MPIA Heidelberg Germany, 25-28 September 2006. Eds: Cristina Afonso, David Weldrake & Thomas Henning

    We describe strategies and tactics for detecting transiting planets, as learned from the experience of the XO Project. A key component is the web-enabled collaboration with a longitudinally-distributed Extended Team of dedicated volunteers operating small-aperture telescopes near their homes. We also quantify the (small) potential to discover transiting planets of M dwarfs from existing data such as that obtained by the XO Project.


  • Paul Shankland March 24, 2007, 14:10

    Also, word from Chile’s recent planet conference is that more than a few people are interested in small-aperture and/or array-ed m-dwarf exoplanet searches – which is good news – the time of the red dwarf is here!


  • marsbound2024 April 26, 2007, 2:27

    Congratulations! You should’ve placed several thousand dollars on that hunch. However it was through radial velocity that Gliese 581 c was discovered.

  • ljk January 21, 2008, 12:03

    Design Considerations for a Ground-based Transit Search for Habitable Planets Orbiting M dwarfs

    Authors: Philip Nutzman, David Charbonneau

    (Submitted on 18 Sep 2007 (v1), last revised 18 Jan 2008 (this version, v2))

    Abstract: By targeting nearby M dwarfs, a transit search using modest equipment is capable of discovering planets as small as 2 Earth radii in the habitable zones of their host stars. The MEarth Project, a future transit search, aims to employ a network of ground-based robotic telescopes to monitor M dwarfs in the northern hemisphere with sufficient precision and cadence to detect such planets. Here we investigate the design requirements for the MEarth Project. We evaluate the optimal bandpass, and the necessary field of view, telescope aperture, and telescope time allocation on a star-by-star basis, as is possible for the well-characterized nearby M dwarfs.

    Through these considerations, 1,976 late M dwarfs (R less than 0.33 Rsun) emerge as favorable targets for transit monitoring. Based on an observational cadence and on total telescope time allocation tailored to recover 90% of transit signals from planets in habitable zone orbits, we find that a network of ten 30 cm telescopes could survey these 1,976 M dwarfs in less than 3 years.

    A null result from this survey would set an upper limit (at 99% confidence) of 17% for the rate of occurrence of planets larger than 2 Earth radii in the habitable zones of late M dwarfs, and even stronger constraints for planets lying closer than the habitable zone. If the true occurrence rate of habitable planets is 10%, the expected yield would be 2.6 planets.

    Comments: accepted to PASP

    Subjects: Astrophysics (astro-ph)

    Cite as: arXiv:0709.2879v2 [astro-ph]

    Submission history

    From: Philip Nutzman [view email]

    [v1] Tue, 18 Sep 2007 16:59:57 GMT (177kb)

    [v2] Fri, 18 Jan 2008 19:56:55 GMT (178kb)