≡ Menu

Red Dwarf Planets and Habitability

The question of habitability on planets around M-dwarfs is compelling, and has been a preoccupation of mine ever since I began working on Centauri Dreams. After all, these dim red stars make up perhaps 75 percent of the stars in the galaxy (percentages vary, but the preponderance of M-dwarfs is clear). The problems of tidal lock, keeping one side of a planet always facing its star, and the potentially extreme radiation environment around young, flaring M-dwarfs have fueled an active debate about whether life could ever emerge here.

At Northwestern University, a team led by Howard Chen, in collaboration with researchers at the University of Colorado Boulder, NASA’s Virtual Planet Laboratory and the Massachusetts Institute of Technology, is tackling the problem by combining 3D climate modeling with atmospheric chemistry. The paper on this work, in press at the Astrophysical Journal, examines how general circulation models (GCM) have been able to simulate the large-scale circulation and climate system feedbacks on planets around red dwarfs, but these models have not accounted for atmospheric chemistry-driven interactions that the authors believe are critical for habitability. Thus so-called coupled chemistry-climate models (CCM) are needed to factor in how an atmosphere responds to the star’s radiation.

The study takes both ultraviolet radiation (UV) from the star and the rotation of the planet into consideration, noting how UV affects gases like water vapor and ozone. Says Chen:

“3D photochemistry plays a huge role because it provides heating or cooling, which can affect the thermodynamics and perhaps the atmospheric composition of a planetary system. These kinds of models have not really been used at all in the exoplanet literature studying rocky planets because they are so computationally expensive. Other photochemical models studying much larger planets, such as gas giants and hot Jupiters, already show that one cannot neglect chemistry when investigating climate.”

Image: An artist’s conception shows a hypothetical planet with two moons orbiting within the habitable zone of a red dwarf star. Credit: NASA/Harvard-Smithsonian Center for Astrophysics/D. Aguilar.

The researchers simulate the atmospheres of synchronously-rotating planets (i.e., with one side always facing the star) at the inner edge of the habitable zones of both K- and M-class stars. using numerical simulations of climate coupled with photochemistry and atmospheric chemistry through their 3D CCM. They find that the thin ozone layers produced on planets around active stars can render an otherwise habitable planet (in terms of surface temperatures) hazardous for complex life, as there is insufficient ozone to block UV radiation from reaching the surface.

Active photochemistry is a crucial issue, for according to Chen and team, planets can also lose significant amounts of water due to vaporization. Added to the ozone issue, we find boundaries beyond which a planet habitable in terms of liquid water on the surface is rendered lifeless. Understanding stellar activity becomes a predictive tool for gauging which M-dwarfs are most likely to merit precious telescope time for future missions looking for biosignatures. More active M-dwarfs appear far less likely to host life-bearing planets. From the paper:

…we find that only climates around active M-dwarfs enter the classical moist greenhouse regime, wherein hydrogen mixing ratios are sufficiently high such that water loss could evaporate the surface ocean within 5 Gyrs. For those around quiescent M-dwarfs, hydrogen mixing ratios do not exceed that of water vapor. As a consequence, we find that planets orbiting quiescent stars have much longer ocean survival timescales than those around active M-dwarfs. Thus, our results suggest that improved constraints on the UV activity of low-mass stars will be critical in understanding the long-term habitability of future discovered exoplanets (e.g., in the TESS sample…)

The effects of stellar UV radiation become a useful predictive tool as we narrow the target list. Vertical and horizontal winds in the upper atmosphere are strengthened as UV flux goes up. Moreover, the global distribution of ozone and hydrogen depends upon all these processes, which can affect the contrast between the dayside and nightside conditions under varying UV flux. The authors believe that only by bringing atmospheric chemistry into the picture of 3D modeling can we gauge whether a planet can attain true habitability and maintain it. Usefully, using their results, they show that both water vapor and ozone features could be detectable by instruments aboard the James Webb Space Telescope if we choose our targets carefully.

The paper is Chen et al., “Habitability and Spectroscopic Observability of Warm M-dwarf Exoplanets Evaluated with a 3D Chemistry-Climate Model,” in press at the Astrophysical JournaL (preprint).


Comments on this entry are closed.

  • T Pliska November 13, 2019, 13:40

    I was imagining a tidally locked world losing its water due to vaporization. I’d think most of the water would freeze on the dark side and build up. Could it build up so much the planet became unbalanced, and flip hemispheres facing their sun? It would be an amusing event if it could happen.

  • James M Essig November 13, 2019, 15:23

    I hope red dwarfs do have habitable zones. There is something romantic about a red sun and a civilization that could live, thrive, and evolve over a time frame of perhaps 30 trillion on a planet orbiting a small red dwarf. Since most stars are red dwarfs, establishing habitability around red dwarfs could be interesting.

  • Harry R Ray November 13, 2019, 16:26

    Andrew Lepage has a very interesting post on the two recently planets orbiting Wolf 359. The bad news is that neither one is habitable. The good news is that there is still room for a potentially habitable planet orbiting in between them. I posted a comment on his Drew Ex Machina website asking him if either Carmines or SPIRu has the sensitivity to detect a 1 Earth mass planet in Wolf 359’s habitable zone . No reply as of yet. Check his post out. It is an excellent read.

    • Andrew LePage November 17, 2019, 16:52

      I thought I felt my ears burning ;-) Here is the link to my latest piece on Wolf 359. While neither of the pair of planet candidates found in this system are potentially habitable, an Earth-size exoplanet orbiting in the HZ would have escaped detection and could still lurk in this system:


    • Andrew LePage November 17, 2019, 18:23

      Sorry about the delayed response but I checked into your question about the ability of CARMENES and SPIRou to make useful radial velocity measurements of Wolf 359. I found that with the J mag of 7.1 for Wolf 359, CARMENES is capable of a radial velocity accuracy of 1 m/s with an integration time of ~350 second. CARMENES easily can measure radial velocities with sufficient accuracy to detect Earth-size exoplanets in the HZ. I have yet to find the needed details on SPIRou but it appears it will also be capable of making useful observations of Wolf 359 as well. Of course the ultimate detection limits will depend on the magnitude and nature of the star’s jitter. Hope that answers the question!

      • Harry R Ray November 18, 2019, 10:56

        Great! Now. based on the assumption that Carmines has been observing Wolf 359 for an extended period of time, two possible scenarios emerge. One is obvious: no evidence for an Earth mass planet in the HZ. However, that leads to the question: Why haven’t they published a paper confirming or refuting Toumi et al’s planets, with the above possible non-detection mentioned in the paper? Because of this I favor the second scenario: There IS evidence for a small planet in the HZ, but, because it is in RESONANCE with one of the above mentioned planets, it is extremely hard to tease out strong enough evidence as of now to publish and more observations will be needed.

        • Andrew LePage November 18, 2019, 14:40

          I think it is dangerous to attempt to infer anything meaningful about the lack of any published CARMENES results on Wolf 359. While it MIGHT mean there are no Earth-size exoplanets in the HZ, they have yet to publish anything about Wolf 359c either which would be readily detectable by them. Having only been up and running for three years (compared to 13 years for the HARPS/HIRES data set), the CARMENES team might not have sufficient data yet to publish any results or maybe noise created from jitter is causing some issues (not unexpected given the youth of Wolf 359) or any number of other plausible issues one could contrive. I learned quite some time ago not to read too much into the lack of published, peer-reviewed results.

          • Harry R Ray November 22, 2019, 11:19

            Space Telescope Live tweeted this 22 hours ago:” I am looking at the star WOLF 359 with Space Telescope Imaging Spectrograph for Dr Christopher Michael Johns-Krull.” Dr Johns-Krull’s field of expertise is T Tauri stars, so I have absolutely no idea why he would target Wolf 359 with STIS. Any ideas?

  • Scott November 13, 2019, 16:50

    UV emissions and near red frequencies are the least problematic of the issues with M class stars. Even if the photochemistry works out, any atmosphere is going to escape dude to proton flux while any macromolecules are going to be disassociated by xrays.

    The real problem with M class stars is their inherent magnetic instability, and not planetary tidal locking in orbits where the water phase diagram looks familiar under earth like pressure and temperature conditions.

  • Geoffrey Hillend November 13, 2019, 17:02

    It depends on the thickness or amount of atmosphere. If the exoplanet has one Earth atmosphere, then the night side would be much colder. We would have one large Hadley cell with a tidally locked planet and no rotation with the winds blowing from the night side into the lower pressure, daylight side. If the atmosphere was two or several bars or Earth atmospheres, then a greenhouse effect, then the night side might be warmer.

    The problem with all tidally locked planets is that the Hill radius prevents it from having any moons and without a moon, there can be no magnetic field since the liquid iron core’s charged particles have to move in circles from a fast rotation to generate a magnetic field like Earth’s rotation does. There is still hope that maybe life might develop near the terminator between day and night were the radiation from the star is less intense and then the life might live near the terminator, but this is asking a lot from life. It might have to adapt in some way or live underground if it needed to leave that area.

    If we seeded one of those tidally locked planets with life it would survive, but if there is any native life and it was restricted it might be hard to detect biosignature gases unless life somehow adapted to the dark side of the planet and the atmosphere would reflect the large scale release of biogases. I still don’t think there is any life on such exoplanets, and a lack of biosignature gases in the atmospheric spectra would support a too hostile environment for life, but the only way to completely rule it out would be to go there and land on the planet and look for it with probes or human exploration.

    • Ashley Baldwin November 14, 2019, 14:34

      It is possible to have a magnetic field without a moon or even in a tidally locked slowly rotating planet. As long as you have a liquid outer iron:nickel core which is convective. Key to this in turn is a convective mantle which helps transport heat away from the core keeping it partially liquid. Also offering plate tectonics , vulcanism and secondary atmosphere outgassing to boot.

      By way of comparison Venus doesn’t have a significant magnetic moment. This is because it’s mantle is locked in by a ‘stagnant lid’ crust with no tectonics and no mantle convention – likely due to its runaway greenhouse past history giving rise to total desiccation. No water, no lubricated mantle and no convection. Mantle or core, despite the latter being similar in size and nature to
      Earth’s. See ‘Characterising Exoplanet Habitability’ , Kopparapu et al, 14th Nov 2019; Astro-ph. Great read all round.

      It is also possible for tidally locked planets to have moons though generally gas giants given the Hill radius gravitational sphere of influence you refer too ( as well as the Roche limit ). See ‘Exomoons in the habitable zones of M dwarfs’ – Martinez-Rodriguez et al ,26th Oct 2019 Astro-ph

      • H. Floyd November 16, 2019, 1:59

        Wondering why Venus still had an atmosphere without a magnetosphere, I’d read that friction between solar wind and a robust ionosphere can create a magnetic field, no convection or tectonics required.


        But that might not be a protective magnetic field. Runaway greenhouse conditions could just replenish atmosphere through non-tectonic volcanism.

        Venus-like might be a better exoplanet norm anyway. Easier to reform a hot rocky planet with too much atmosphere, than a cold one with not enough?

    • Ronald November 15, 2019, 8:20

      “without a moon, there can be no magnetic field ”

      I think that is not right: the magnetic field of the earth is generated by electric currents, in turn caused by the convective motion (convection currents) of molten iron in the Earth’s outer core, and these convection currents are caused by internal heat (geo-dynamo). The internal heat is caused both by residual heat from the Earth’s formation and, probably more important, from radioactive decay, foremost Uranium and Thorium.

      The moon seems to become less and less important anyway, over the years: formerly it was often suggested that the moon was particularly important to stabilize the axial tilt of the Earth and in fact this is still often quoted.
      However, even this has been seriously questioned in recent years, e.g. by Lissauer, Barnes and Chambers (Obliquity variations of a moonless Earth; 2011), quote: “We find that (…) the obliquity remains within a constrained range, typically 20–25 in extent,
      for timescales of hundreds of millions of years. (…) A large moon
      thus does not seem to be needed to stabilize the obliquity of an Earth-like planet on timescales relevant to the development of advanced life.”.

  • Brett November 13, 2019, 17:05

    I still think they’d all be Venus-style planets because of the long Pre-Main Sequence Phase with heightened luminosity, but maybe the water can come after that from comets, etc.

    • Ashley Baldwin November 14, 2019, 14:41

      Or from the planet’s core where it can have been sequestered since formation from a volatile rich accretion disk .

  • H. Floyd November 13, 2019, 17:16

    A normative overabundance of exposed, cold, dry, Andean mummy-planets, everywhere we look. It’s demoralizing — at least, to hopes of encountering other conversationally intelligent species.

    But I’m coming to appreciate what that could mean in terms of ethically unencumbered real estate. Native terrestrial magnetospheres may be rare, but for a spacefacing technological species magnetospheres will just be a school of architecture:



    We’re actually on a pretty tight clock here, circling this late middle-aged star. Even our magnificent natural magnetosphere won’t save us forever. And if we don’t thrive I doubt there’ll be time for Earth to raise another species like ours from scratch.

    Experience with Mars could equip Homo sapiens to breathe first life into a wealth of mummy planets comfortably situated around longer-lived stars. In what appears to be an otherwise depressingly sterile universe, that’s a silver lining I could live with.

  • Ashley.Baldwin November 13, 2019, 19:51

    You have to see this paper for what it is . A pragmatic approach to identifying and then targeting those planets orbiting in the habitable zones of less active ( at this current time and presumably for some billions of years before ) M dwarfs. Then using this to allocate available resource.

    Observing time on JWST is going to be at an absolute premium and whilst finding out that a planet has had its atmosphere stripped is scientifically useful it’s rather a dull outcome and WOULD NOT be able to be extrapolated to all M dwarf planets. The ELTs too, even with their huge capabilities and much longer lives . The kind of cross correlation high dispersion spectroscopy / high contrast imaging techniques they would employ to look for biosignatures will be both hugely labour intensive and expensive ON TOP of observing time on the scope.

    What this article does not do is rule out potentially habitable planets around M dwarfs . Even those orbiting active, atmosphere stripping stars , can keep a large volatile content sequestered in their mantles to replenish vulcanism driven secondary atmospheres in more favourable times. Or undergo volatile injection via late cometary bombardment . Or migrate in late from beyond the ice line as appears to have been the case with the Trappist-1 planets. Many of which have already been shown by Hubble spectrophotometry to have maintained a substantial volatile fraction ( which is of course not necessarily indicative of having atmospheres on its own – hence the prioritisation for JWST ) . Trappist-1, though much less active now, being a fully convective late M dwarf must have had an extended pre and most main sequence period of high activity. If some or any of its seven planets have significant atmospheres despite this, such results WOULD be 1/ hugely interesting 2/ highly indicative and 3/ able to be extrapolated .

    Keep a thick primordial or secondary atmosphere – perhaps with an ocean too – volatiles anyway – and you have two very efficient ways of transferring heat from star facing to non facing hemispheres . More than capable of preventing atmospheric collapse on the “cold” side even in a fully synchronised planet. Bearing in mind too that gravitational interactions with other planets , especially if closely packed as with the Trappist-1 system, might induce 3:2 or 2:1 resonances even in tidally locked planets. Gravitational tides induced by thick atmospheres have also been shown by Leconte et al ( 2015) to help M dwarf planets resist synchronicity over Giga years.

    I have to say that it is exciting to see 3-D atmospheric modelling coming into its own . Not just in screening potential targets , it represents a big step forward in describing various exoplanetary atmospheres . This will be central to the interpretation of the ground based high dispersion spectroscopy utilised to characterise exoplanet atmospheres by the ELTs.

  • Ronald November 14, 2019, 3:38

    Related interesting paper about planets with multiple host stars, a survey:

  • Ronald November 14, 2019, 4:00

    What is the definition of, or criteria for ‘active’ M dwarf?
    Aren’t most M dwarfs active, in UV output and flaring?

    This seems like another confirmation (plus flaring and tidal locking), that M dwarfs are not the most suitable class of stars for planets with (complex) life.

  • Michael Fidler November 14, 2019, 10:35

    Gee, how quickly the tide changes! Oh, as I said before in the last article on Ariel these planets in the compact M-dwarfs are constantly being bombarded by comets at a much higher rate then any thing in our solar system. Short orbit period – in close to the red dwarf makes for many more comet encounters, think about it!
    Now, just to make it more confusing, the M-dwarfs on the low end M5-M9 are fully convective but have a mass 80 or more times Jupiter’s mass but are close to the same size as Jupiter. This means their density is also much higher, but they are also having many more large impacts from comets and asteroids. Remember the impacts on Jupiter in 1994 from a huge comet?
    These stars are a different beast compared to our Sun and the dynamics and flaring work differently then ours. One of the unusual aspects of low end red dwarfs is the carbon based molecules like Methane in their atmosphere. Plus Titanium-oxide compounds, with other metals that would fuel magneto-hydrodynamic (MHD) processes in flares in the powerful magnetic fields around and in M-dwarfs. Remember the ink black impact spots from when the chain of comets crashed into Jupiter?
    Well take a guess, the comets and asteroids are seeding the red dwarfs and causing the active flaring rates. So as these system age there is less and less comets and asteroids impacts just as in our solar system. So the older stars have more or less stopped flaring. Like I said these late red stars are a different beast. O’ don’t forget that brown dwarfs also give out huge UV flares, so go figure!

    • Michael Fidler November 15, 2019, 23:37

      “The depletion of the Ozone layer from UV flaring around active M-dwarfs may not be as bad as indicated in the article, though. Recently an article mentions the higher impact rates from comets around these tightly wound M-dwarf planetary systems. The early active period of M-dwarfs may be countered in later life by the addition of large quantities of water from the much higher cometary impact rates. This would resupply the Ozone layer along with the surface based water supply. It will not be long till we find out if this question is so, the JWT should show if the Ozone layers are present around the numerous nearby M-dwarfs planets.

      Cometary impactors on the TRAPPIST-1 planets can destroy all planetary atmospheres and rebuild secondary atmospheres on planets f, g, h.
      “The TRAPPIST-1 system is unique in that it has a chain of seven terrestrial Earth-like planets located close to or in its habitable zone. In this paper, we study the effect of potential cometary impacts on the TRAPPIST-1 planets and how they would affect the primordial atmospheres of these planets. We consider both atmospheric mass loss and volatile delivery with a view to assessing whether any sort of life has a chance to develop. We ran N-body simulations to investigate the orbital evolution of potential impacting comets, to determine which planets are more likely to be impacted and the distributions of impact velocities. We consider three scenarios that could potentially throw comets into the inner region (i.e within 0.1au where the seven planets are located) from an (as yet undetected) outer belt similar to the Kuiper belt or an Oort cloud: Planet scattering, the Kozai-Lidov mechanism and Galactic tides. For the different scenarios, we quantify, for each planet, how much atmospheric mass is lost and what mass of volatiles can be delivered over the age of the system depending on the mass scattered out of the outer belt. We find that the resulting high velocity impacts can easily destroy the primordial atmospheres of all seven planets, even if the mass scattered from the outer belt is as low as that of the Kuiper belt. However, we find that the atmospheres of the outermost planets f, g and h can also easily be replenished with cometary volatiles (e.g. ∼ an Earth ocean mass of water could be delivered). These scenarios would thus imply that the atmospheres of these outermost planets could be more massive than those of the innermost planets, and have volatiles-enriched composition.”

      Susceptibility of planetary atmospheres to mass loss and growth by planetesimal impacts: the impact shoreline.
      “This paper considers how planetesimal impacts affect planetary atmospheres. Atmosphere evolution depends on the ratio of gain from volatiles to loss from atmosphere stripping f_v; for constant bombardment, atmospheres with f_v1. An impact outcome prescription is used to characterise how f_v depends on planetesimal impact velocities, size distribution and composition. Planets that are low mass and/or close to the star have atmospheres that deplete in impacts, while high mass and/or distant planets grow secondary atmospheres. Dividing these outcomes is an fv=1 impact shoreline analogous to Zahnle & Catling’s cosmic shoreline. The impact shoreline’s location depends on assumed impacting planetesimal properties, so conclusions for the atmospheric evolution of a planet like Earth with f_v~1 are only as strong as those assumptions. Application to the exoplanet population shows the gap in the planet radius distribution at ~1.5R_earth is coincident with the impact shoreline, which has a similar dependence on orbital period and stellar mass to the observed gap. Given sufficient bombardment, planets below the gap would be expected to lose their atmospheres, while those above could have atmospheres enhanced in volatiles. The level of atmosphere alteration depends on the total bombardment a planet experiences, and so on the system’s (usually unknown) other planets and planetesimals, though massive distant planets would have low accretion efficiency. Habitable zone planets around lower luminosity stars are more susceptible to atmosphere stripping, disfavouring M stars as hosts of life-bearing planets if Earth-like bombardment is conducive to the development of life.”

      The question is whether the super-Earths and sub-Neptunes would have a thick organic goo covering it. Trappist 1c may be an example of this, with the hydrogen atmosphere depleted. The other factor that would be effecting all planets in the M-dwarf family would be a much faster moving Lithosphere similar to our oceanic crust. The higher impact rates from comets and asteroid would cause higher mixing rates for the Asthenosphere and upper Mantle with a higher content of water.
      This could indeed make for a completely different type of crust with much higher concentration of carbon and organic matter. The late impacts could also give rise to seamounts and volcanic eruptions, see this; https://www.gns.cri.nz/Home/News-and-Events/Media-Releases/largest-caldera

  • ljk November 14, 2019, 11:25

    While the possibility for *native* ETI in red dwarf systems may be problematic, there is nothing saying that non-native species with interstellar capabilities could not be inhabiting such systems for settlement, resources, etc. Especially since they may otherwise be unoccupied. I wish these astronomers would seriously consider that possibility.


  • Geoffrey Hillend November 14, 2019, 19:29

    Most M dwarf exoplanets might have considerable atmospheres, but their surfaces also might be sterile due to the continual EUV and X rays. A magnetic field could block or deflect some solar wind, but a tidally locked exoplanet will never have a magnetic field since there is no rotation to spin the liquid core and make charged particles move in circles. Tidal forces could certainly help with volcanism and replenish an atmosphere. If there are not any biosignature gases like oxygen and methane I won’t be surprised. We still have to study their spectra and also we are only scratching the surface of how many potential exoplanets are out there.

    Maybe with the extremely large telescope and other missions like JWST we could look at only G class stars or design a search to look at them and locate a good portion of those stars if we can’t find any biosignature gases. In other words keep long term watch of only G class star systems instead of just looking for what we can see in short term.

  • Laintal November 18, 2019, 21:34

    An excellent post and comments

    I have enjoyed reading, its one of my top areas of interest too

    Cheers Laintal