≡ Menu

Water World Around Gliese 581?

Gliese 581, the star that teased us a few years back with reports of a ‘super-Earth’ planet in the habitable zone, is back in the news. Michel Mayor’s Geneva team has located a fourth planet in the system, Gliese 581 e, which weighs in at a mere 1.9 Earth masses, making it the least massive exoplanet ever detected. Orbiting its primary in 3.15 days, the newly found world is too close to the star to be in the habitable zone, but the other shoe that drops here is that Gl 581 d may itself be.


Image: By refining the orbit of the planet Gliese 581 d, first discovered in 2007, a team of astronomers has shown that it lies well within the habitable zone, where liquid water oceans could exist. This diagram shows the distances of the planets in the Solar System (upper row) and in the Gliese 581 system (lower row), from their respective stars (left). The habitable zone is indicated as the blue area, showing that Gliese 581 d is located inside the habitable zone around its low-mass red star. Credit: Based on a diagram by Franck Selsis (University of Bordeaux).

Watching this story unfold has been instructive. Early indications that Gl 581 c was in the habitable zone were quickly challenged, with a consensus developing that the planet was too hot for liquid water to exist on its surface. Indeed, it might well resemble Venus more than Earth, a notion that was rarely picked up by the media. The belief that Gl 581 c is habitable has become more or less codified in some quarters — a National Geographic special not long ago depicted it as a green and blue living world — long before the scientific process could come to a resolution on the matter.

At the same time, Gl 581 d, a planet seven times as massive as the Earth, was thought to be on the very outer edge of the habitable zone, a longshot candidate for liquid surface water. The Geneva team, in the process of discovering the new planet, has now refined the orbit of Gl 581 d, the updated work indicating that it could be an interesting place indeed:

“Gliese 581 d is probably too massive to be made only of rocky material, but we can speculate that it is an icy planet that has migrated closer to the star,” says team member Stephane Udry. The new observations have revealed that this planet is in the habitable zone, where liquid water could exist. “‘d’ could even be covered by a large and deep ocean — it is the first serious ‘water world’ candidate.”

So now we have firmer indications of a planet in the habitable zone. And as this news release points out, the progress of our work in the Gliese system is nothing short of phenomenal. Michel Mayor, the man who, along with Didier Queloz, was behind the planetary discovery at 51 Pegasi just a decade ago, notes that the mass of Gl 581 e is fully eighty times less than that of 51 Pegasi b. That’s also a reminder that low-mass red dwarfs are good places to look for small worlds, since the gravitational tug of close-in planets is more pronounced.

Look for the paper on this find to run in Astronomy & Astrophysics. It’s Mayor et al., “The HARPS search for southern extra-solar planets: XVIII. An Earth-mass planet in the GJ 581 planetary system,” available here.

Comments on this entry are closed.

  • ljk April 21, 2009, 10:44

    Whether anybody with a radio telescope is there or not,
    the Gliese 581 system will be getting a message from a
    selection of Earth humans in the year 2029:


  • Vagueofgodalming April 21, 2009, 13:15

    Actually, there’s a link on that ESO page to the preprint:


    While I haven’t had a chance to read it in detail, it appears that their best solution for d gives its orbit an eccentricity of 0.38 ± 0.09, which has some interesting implications for habitability.

  • Administrator April 21, 2009, 14:10

    Several readers wrote in to identify the link, which I had missed — thanks to all! I’ve updated the post to reflect this, and also managed to get around a server problem (with some good tech support) that was preventing me from uploading the image for this story.

    Vagueofgodalming, the eccentricity is interesting, and so is the fact that Gl 581 d is likely to be a water world, which sets up all kinds of interesting questions about whether and how life might develop. The Gliese 581 system gets more interesting every day.

  • Didac April 21, 2009, 14:19

    Tidal locking and higher excentrities could be a problem for planet habitability in other systems. But I guess that in a red dwarf system, that it is not very important, because of the short orbit period of both c and d planets.

  • Christopher L. Bennett April 21, 2009, 15:31

    On Earth, oceanic life depends on mineral runoff from the land. Portions of the ocean without material flowing from the land are essentially deserts. So if a planet is covered completely in ocean, particularly if it’s a full-on Leger-type ocean planet with a mantle of allotropic ice beneath the sea, I doubt there’d be enough mineral content in the oceans to support life.

  • TheoA April 21, 2009, 16:26

    Does anyone know what the relationship is between atmospheric density and surface Temperature/Habitability?

    Seems to me the bigger planet = thicker atmosphere = better green house effect = higher surface temperature = expanded habitability zone.

  • Dave Moore April 21, 2009, 18:37

    Some first thoughts on the new discoveries.
    Gliese 581 e
    If the planet had a similar composition to Earth, it would have a diameter of 9800 miles. I’m guessing that its level of illumination is similar to Mercury’s so it would probably resemble a more hellish version of Venus, but unlike Venus, it would have a magnetic
    field because even though it is tidally locked, keeping one face to its parent star, the orbital period of 3.14 days would generate a magnetic field in its liquid-iron core.

    Gliese 581 d
    The orbital eccentricity of this planet 0.38 is close to the 0.412 eccentricity of a 2:1 orbital/spin resonance. This would give the planet a 33 day rotational period. That may be too slow to maintain a decent magnetic field.

    If we assume the planet has an Earthlike core with a greater volatile inventory (i.e., an ocean planet), then if the overlying atmosphere is below 2-3 atm. then the surface of the ocean will be frozen. At about that level, the adiabatic lapse rate will keep the surface of the planet above freezing point even though the planets level of illumination looks to be less than Mars’s.
    If, however, the overlying atmosphere reaches the 8-10 atm. level then the planet’s surface will approach the boiling point of water and you will have a steam/superfluid planet.

    The about figures are just wild guesses off the top of my head, but I was using them to make a point about the general principles involved.

    Incidentally, there’s a good little primer on atmosphere escape mechanisms in this months (May 09) Scientific America.

  • Dave Moore April 21, 2009, 19:19

    Another quick thought.

    The figures for Gliese 51 d look suspiciously like the combined signal of two planets orbiting in a 2:1 resonance. If this was the case then, the inner of two would have a 33 day orbital period, which would put it right in the middle of the habitable zone.

    Incidentally, if you do that, you get a system with a very regular set of orbital periods: 3.14, 5.37, 13, 33 & 66 days.

  • kurt9 April 21, 2009, 20:26

    So, if Gliese 581 d has an atmosphere with a pressure around 3-4 atmospheres, then the temperature should be nice and warm, but not too hot. Of course, if it has a planetary ocean, then the dissolved minerals might be too low for much life to exist in that ocean.

  • spaceman April 22, 2009, 2:05

    Does this discovery of an ~2 Earth-mass planet around a red dwarf change the pessimistic outlook existing among some in the exoplanet community when it comes to the ability of M-dwarf disks to give rise to earth-mass rocky planets? Afterall, I think in one of the news stories regarding this discovery one of the astronomers involved said that this new planet is “very likely a rocky planet.” Furthermore, the microlensed planet that was originally thought to be ~3 Earth masses, had its mass est. revised downward–closer to 1-2 Earth masses; thus, we now know, despite our impressive yet still limited capablities, of two examples of very likely terrestrial planets existing around low mass stars.

    Here is the title of the pessimistic outlook for Earth-sized terrestrial planets around M-dwarfs: “A decreased probability of habitable planet formation around low-mass stars” by Raymond et al (2007).

  • Adam April 22, 2009, 7:03

    Hi Chris, Dave & All

    Dave, two for the price of one. That’s cool. And a mass sum of 7 Earths… so they’d both be more Earth-like. Earth-like planets between 1 and 10 Earth masses have radii (earth = 1) = M^n, with n in the range 0.27-0.26. That’s according to most power-law fits to more detailed modelling using the best rheological & Equation of State data of the mantle and core. Ocean planets, 1/2 water & 1/2 silicates/iron, have radii about 1.26*M^n, with n in the same range as before. An ocean planet would thus be about 16,900 miles across at 7 Earth masses.

    Chris, yes the ‘ocean’ would probably be mostly ice VII and higher phases of ice, probably glowing red-hot at the bottom of the cryosphere. But convection might keep the cryosphere well churned, so I wouldn’t rule out mineral outflows from deep down just yet. There would be no cool, hard oceanbed of silicates, just churning high-pressure phases of ice. Hard to imagine the weird cryo-volcanism that might be possible with such a hell-brew of a mantle.

  • MaDeR April 22, 2009, 7:52

    “weighs in at a mere 1.9 Earth masses, making it the least massive exoplanet ever detected.”
    You apparently forgot about some certain pulsar planet. But yeah, 0.025 Me is somewhat small for a planet. :)

  • Ronald April 22, 2009, 10:00

    The whole system looks somewhat like a miniaturized version of our solar system, with ‘mini-giant’ planets not bigger than super-earths.

    I wonder whether that is due to the smaller star mass, the significantly lower metallicity, or both.

    Question: with what we know now (from existing RV surveys), are planets around red dwarfs relatively common, or rather the exceptions? My impression is the latter.

    In the Extrasolar Planets Encyclopedia I found about 20 M dwarfs with planets, out of a total of 293 stars with planets. So that would be only some 7% of all stars with planets found so far, very few in relation to the predominance of M dwarfs in our galaxy ánd considering the fact that these dwarfs would reveal their planets relatively easily. But I do not know the proportion of M dwarfs in the total survey.

  • ljk April 22, 2009, 12:12

    Planets circling pulsars may be a wee bit hard to live on,
    but just imagine if we can what kind of creatures could
    survive on a world that was first blasted by a stellar
    explosion then bathed in intense radiation for ages.

  • Dave Moore April 22, 2009, 15:41

    This is a re-post of Gerald Nordley’s comments to the Contact list (with his permission). I wish he would post on this blog when there is an exosolar discovery as he has a spread sheet set up that automatically produces supplementary information such as insolations for a given planet.

    The preprint of the GJ 541e discovery paper is (for now…) available at:

    GJ 542 (for Gliese-Jahreiss) of Gl 541 (for Gliese) is a M3 (“red dwarf) main sequence star 20 light years away. Its first planet was discovered in 2005; there are now four, b, c, d, and e

    Note that the minimum mass of GJ 541e is 1.94 Earth masses, Mayor et al. do not speculate on the size of the planet. They do estimate a maximum mass of 3.1 Earth masses, based on system dynamical considerations that put the minimum inclination of the system at about 40 degrees. They also note the following:

    Dopplerdetected super-Earths, planets with a minimum mass under10 M e.
    This planetary mass domain, as well as the Neptune mass range (<30 M ) has largely been populated by the HARPS (European-built spectroscopic equipment at a Chilean observatory) surveys. A few statistical properties have already emerged from these early discoveries (Mayor & Udry 2008):
    •• The full distribution of planetary masses is bimodal, with distinct peaks corresponding to gaseous giants and super-Earths. Despite the observational bias against low mass planets, the distribution below 30 M rises towards super-Earth planets (cf. Fig. 7 Mayor & Udry 2008).
    •• The majority of super-Earths and Neptunes are found in multiplanetary systems. 4 of the 6 planetary systems with a known super-Earth, (GJ 876, HD 40307, HD 7924, GJ 176, GJ 581,HD 181433) have multiple planets. 2 of these 4 multi-planet systems associate one super-Earth with one or two gaseous giant planets (GJ 876, HD 181433), and the other 2 have several low-mass planets on non resonant orbits (HD 40307, GJ581). GJ 176 and HD 7924 have only one detected planet, though they could obviously have more which haven’t been detected yet. Indeed, the periodogram of HD 7924 shows hints of possible additional planets (Howard et al. 2009).
    •• Contrary to gaseous giants, low mass planets seem no more frequent around metal-rich host stars (Udry et al. 2006).
    •• In a preliminary analysis, we have detected low mass close in planets (P< 50d and msin i 40 and therefore that the mass of each planet be no more than 1.6 times its minimum mass. For GJ 581e, b, c and d, those upper limits are 3.1, 30.4, 10.4 and 13.8 M .

    For comparison, Uranus is 14.6 Earth masses and Saturn is 95 Earth Masses.

    Based on an estimated bolometric correction* of -2.03 magnitudes, my estimated luminosity of GJ541 is 0.01222 suns (about 1/80th solar luminosity) and the insolations of its planets in terms of the Earth’s are:

    GJ 541 e: 15.1
    GJ 541 b: 7.4
    GJ 541 c: 2.3
    GJ 541 d: 0.26

    It’s difficult to place a conventionally habitable planet in this system. Planet c is at least a 30% the mass of Uranus. One might be able to fit a large moon in between 80,000 km (solar tidal perturbation problems beyond that) and 36,000 km (Roche limit) from the planet, but the insolation is Venus-like. The next planet out has more room for moons, but with only about a quarter of Earth’s insolation, they would need very massive atmospheres for liquid water surface temperatures. One might conceivably fit an Earth-mass planet between c and d, but that seems dynamically dubious. I’d suspect something like an asteroid belt between the two. If so, an astrotechnical civilization might find the system attractive to colonize; there would be plenty of raw materials and nearby hydrogen.

    –Best, Gerald

  • amphiox April 23, 2009, 11:36

    Everyone seems to be ignoring the pulsar planets lately? Have they been pluto’d somehow, and redefined as not being planets because they’re not circling a main sequence star?

  • Administrator April 23, 2009, 13:19

    amphiox, the pulsar planets are well known, but I think that although we can consider them planets, their environment is so extreme, and their hosts so far off the main sequence, that we don’t usually focus on them when talking about the prospects for life in their systems.

  • Administrator April 23, 2009, 13:20

    Dave Moore writes:

    I wish he would post on this blog when there is an exosolar discovery as he has a spread sheet set up that automatically produces supplementary information such as insolations for a given planet.

    I’m a great admirer of Gerald’s, and maybe between the two of us we can get him to repost his always interesting ideas here?

  • John Hunt April 23, 2009, 14:15

    If we are about two years from discovering an Earth-sized planet then how far are we from being able to find all such planets with liquid water within 20 light-years? Whenever that is, we’ll also be close to being able to do direct imaging of their atmospheres and then we may know whether there is a life-bearing planet within reach of an interstellar probe.

    If Earth-sized planets with liquid water are rare, and/or if it is rare for those planets to have life, then what determines our criteria for a good target for an interstellar mission?

    I too appreciate Gerald Nordley’s post and I especially appreciate his brining up the implications for colonization by a “astrotechnical” civilization. I can’t imagine that that means anyone else than us!

    I think that we interstellar mission advocates need to bear in mind as a possibility the other great motivation for an interstellar mission, namely colonization. If we could find a cost-effective and safe way of doing that then all sorts of planets become legitimate targets for an interstellar mission.

  • Ronald April 24, 2009, 4:57

    Ref. to John Hunt:
    I recently read (Improving Earth-like planets’ detection with an ELT: the differential radial velocity experiment; Riaud, P.; Schneider, J.
    Astronomy and Astrophysics, Volume 469, Issue 1, July 2007; and: Detectability of rocky planets with ELT infrared spectroscopy
    Martín, E. L.; Guenther, Eike, Cambridge University Press, 2006), that the planned European ELT will be able to detect earth-sized planets, both indirectly by RV (up to some 100 ly) ánd direct imaging and spectroscopic analysis (only for the nearest stars), using a coronagraph and adaptive optics (the segmented mirror will be 42 meters diameter!).

    “If Earth-sized planets with liquid water are rare, and/or if it is rare for those planets to have life, then what determines our criteria for a good target for an interstellar mission?”

    I think the main criteria for a target terrestrial (rocky) planet *is* liquid water, besides a breathable or at least tranformable atmosphere. Not life perse, although that would be fascinating for study of course.

    Rocky planets with liquid water and some kind of usable atmosphere should not be too uncommon. The near future will teach us.

  • amphiox April 25, 2009, 13:31

    Christopher L. Bennet: But even those “desert” ocean areas are not sterile. And there are many ecosystems on earth that do not rely on mineral run-off from land-masses.

    Since we still don’t now how life arose on earth, nor have any idea as to the diversity of chemical pathways and conditions that might have potential to produce life, I think it is rather premature to declare that this or that planetary environment cannot have life just because it lacks some feature of earth, even one we might think was important in the genesis of life of earth.

    Unless it is 100% water only, without any carbon at all. A pure ocean planet without landmasses might have an impoverished biosphere compared to earth, but completely sterile? Seems a little premature to conclude that.

  • Adam April 26, 2009, 19:17

    Just a quick note about the pulsar planets – they may not be regular matter. There’s a few theoretical reasons for thinking they might be quarkonium nuggets and an Earth mass would be mere metres across. Until we know more they’re not definitely planets-as-we-know-them.

  • ljk May 14, 2009, 9:13

    A Scientist’s Guide to Finding Alien Life: Where, When, and in What Universe

    A variety of new findings point to the “habitable zones” where we’re likely to find extraterrestrials.

    by Adam Frank

    From the May 2009 issue, published online May 11, 2009

    Massive stars give, but they also take away—and that puts the inner limit on the galactic habitable zone. The supernova explosions that create and spread heavy elements also unleash a torrent of high-energy radiation: gamma rays, X-rays, and ultraviolet light. Those stellar explosions can have lethal effects on planets orbiting stars even tens of light-years away. In the crowded central regions of the galaxy, home to large numbers of massive stars, supernovas are so common that the evolution of complex life-forms might be difficult if not impossible.

    The big question is how bad the supernova effect is. Lineweaver and his colleagues calculate that radiation poisoning could exclude the inner 20 percent of the Milky Way, which encompasses about half of all the stars in the galaxy. “You are looking for that sweet spot,” says Fred Adams of the University of Michigan, “where you are not so close to the center that conditions are hostile and not so far out that the metal abundance is too low.” But the Milky Way is huge, so Adams suggests putting things in perspective. “At worst the amount of galactic real estate favorable to life is reduced by a factor of two or three,” he says.

    Full article here:


  • ljk July 4, 2009, 9:27


    June 10, 2009 | Science

    New definition could further limit habitable zones around distant suns

    Vince Stricherz vinces@u.washington.edu

    As astronomers gaze toward nearby planetary systems in search of life, they are focusing their attention on each system’s habitable zone, where heat radiated from the star is just right to keep a planet’s water in liquid form.

    A number of planets have been discovered orbiting red dwarf stars, which make up about three-quarters of the stars close to our solar system. Potentially habitable planets must orbit close to those stars — perhaps one-fiftieth the distance of Earth to the sun — since those stars are smaller and generate less heat than our sun.

    But new calculations indicate that, with planets so close, tidal forces exerted on planets by the parent star’s gravity could limit what is regarded as a star’s habitable zone and change the criteria for planets where life could potentially take root.

    Scientists believe liquid water is essential for life. But a planet also must have plate tectonics to pull excess carbon from its atmosphere and confine it in rocks to prevent runaway greenhouse warming. Tectonics, or the movement of the plates that make up a planet’s surface, typically is driven by radioactive decay in the planet’s core, but a star’s gravity can cause tides in the planet, which creates more energy to drive plate tectonics.

    “If you have plate tectonics, then you can have long-term climate stability, which we think is a prerequisite for life,” said Rory Barnes, a University of Washington postdoctoral researcher in astronomy.

    However, tectonic forces cannot be so severe that geologic events quickly repave a planet’s surface and destroy life that might have gotten a foothold, he said. The planet must be at a distance where tugging from the star’s gravitational field generates tectonics without setting off extreme volcanic activity that resurfaces the planet in too short a time for life to prosper.

    Barnes is lead author of a paper to be published by The Astrophysical Journal Letters that uses new calculations from computer modeling to define a “tidal habitable zone.” Co-authors are Brian Jackson and Richard Greenberg from the University of Arizona and Sean Raymond from the University of Colorado. The research was funded by NASA.

    “Overall, the effect of this work is to reduce the number of habitable environments in the universe, or at least what we have thought of as habitable environments,” Barnes said. “The best places to look for habitability are where this new definition and the old definition overlap.”

    The new calculations have implications for planets previously considered too small for habitability. An example is Mars, which used to experience tectonics but that activity ceased as heat from the planet’s decaying inner core dissipated.

    But as planets get closer to their suns, the gravitational pull gets stronger, tidal forces increase and more energy is released. If Mars were to move closer to the sun, the sun’s tidal tugs could possibly restart the tectonics, releasing gases from the core to provide more atmosphere. If Mars harbors liquid water, at that point it could be habitable for life as we know it.

    Various moons of Jupiter have long been considered as potentially harboring life. But one of them, Io, has so much volcanic activity, the result of tidal forces from Jupiter, that it is not regarded as a good candidate. Tectonic activity remakes Io’s surface in less than 1 million years.

    “If that were to happen on Earth, it would be hard to imagine how life would develop,” Barnes said.

    A potential Earth-like planet, but eight times more massive, called Gliese 581d was discovered in 2007 about 20 light years away in the constellation Libra. At first it was thought the planet was too far from its sun, Gliese 581, to have liquid water, but recent observations have determined the orbit is within the habitable zone for liquid water. However, the planet is outside the habitable zone for its sun’s tidal forces, which the authors believe drastically limits the possibility of life.

    “Our model predicts that tides may contribute only one-quarter of the heating required to make the planet habitable, so a lot of heat from decay of radioactive isotopes may be required to make up the difference,” Jackson said.

    Barnes added, “The bottom line is that tidal forcing is an important factor that we are going to have to consider when looking for habitable planets.”


    For more information, contact Barnes at 206-543-8979 or rory@astro.washington.edu.

    The paper is available at


  • Vince September 17, 2011, 0:04

    Question, I notice that this planet is mostly water. and i can’t help but notice it is a red dwarf young small star. a sign of a new solar system and earth 500 million years ago was mostly water and a start of new life forms so is it possible that this new planet could be a new born planet at the very first stages of containing life? interesting theory to me but theres only one way to find out.