≡ Menu

The Survival of M-Dwarf Planet Atmospheres

I was interested in yesterday’s story about the two super-Earths around nearby M-dwarfs — TOI-1634b and TOI-1685b — partly because of the research that follows. In both cases there is the question of atmospheres. The two TESS planets are so numbingly close to their host stars that they may have lost their original hydrogen/helium atmospheres in favor of an atmosphere sustained by emissions from within. Hearteningly, we should be able to find out more with the James Webb Space Telescope, on which ride the hopes of so many exoplanet researchers.

Today’s system is the intriguing L 98-59, only 35 light years from Earth and possessed of at least four planets, with a fifth as yet unconfirmed. Here we have two rocky inner worlds, a possible ocean planet (L 98-59 d) and another likely rocky world to the inside of the habitable zone boundary. Perhaps within the habitable zone, if it exists, is L 98-59f, so this is a system to keep an eye on, an obvious candidate as a JWST target.

At UC Riverside, Daria Pidhorodetska wants to know whether small, rocky planets orbiting M-dwarfs like this one have atmospheres. The paper is devoted to the question of whether either Hubble or JWST, perhaps working in tandem, could detect atmospheres in this system. The authors proceeded to model four different types of atmospheres to answer the question with regard to the three inner planets.

Image: This is an infographic from the European Southern Observatory showing a comparison between the L 98-59 exoplanet system (top) with part of the inner Solar System (Mercury, Venus and Earth), highlighting the similarities between the two. L 98-59 contains four confirmed rocky planets (marked in color in the top panel), orbiting a red-dwarf star 35 light-years away. The planet closest to the star is around half the mass of Venus, making it the lightest exoplanet ever detected using the radial velocity technique. Up to 30% of the third planet’s mass could be water, making it an ocean world. The existence of the fourth planet has been confirmed, but scientists don’t yet know its mass and radius (its possible size is indicated by a dotted line). The team also found hints of a potential fifth planet, the furthest from the star, though the team knows little about it. If confirmed, it would sit in the system’s habitable zone where liquid water could exist on its surface. Credit: ESO/L. Calçada/M. Kornmesser (Acknowledgment: O. Demangeon). [Note: The distances from the stars and between the planets in the infographic are not to scale. The diagram has been scaled to make the habitable zone in both the Solar System and in L 98-59 coincide].

A major problem for M-dwarf planets, to go along with tidal lock, is the fact that during their formation, they are bathed in intense ultraviolet radiation. Enough so that the potential is there to cause any water at the surface to evaporate, while their atmospheres would be under a fierce barrage and might not survive. The question for Pidhorodetska and team is, then, whether the two inner rocky planets have lost their atmospheres completely, if they had one, or if they have been able to replenish them.

The range of atmospheric scenarios takes in planets with atmospheres dominated by water, hydrogen, carbon dioxide, or oxygen and ozone (remaining after loss of hydrogen). The authors argue that an oxygen-dominated atmosphere is the most likely. For each of these scenarios, the authors simulated transmission spectroscopy. L 98-59’s proximity to Earth as well as the fast orbits (less than a week) of its planets speeds up the process of discovery. In fact, says Edward Schwieterman (UC Riverside):

“It would only take a few transits with Hubble to detect or rule out a hydrogen- or steam-dominated atmosphere without clouds. With as few as 20 transits, Webb would allow us to characterize gases in heavy carbon dioxide or oxygen-dominated atmospheres.”

I’m interested, though, specifically in that question of atmosphere loss, with hydrogen escape leaving oxygen and ozone behind. The paper explains:

Highly irradiated planets such as those of the L 98-59 system could have a desiccated atmospheric composition, such as one that is dominated by O2, as a result of major ocean loss during an extended runaway greenhouse phase. A desiccated planet that is rich in abiotic O2 would be expected to form O3 from the photochemical processing of O2, meaning that the direct detection of O3 absorption could be another key indicator of this planetary state.

And the authors point out in their conclusion that we can learn a great deal about the evolution of these planets depending on whether we detect water in their atmospheres. An atmosphere high in oxygen due to the loss of hydrogen during the star’s pre-main sequence phase — in other words, an atmosphere that survives utter desiccation — should have no oceans to detect. Water, or the lack of it, is another marker for this early stage of planetary evolution, and our instruments should be able to make the call.

The paper is Pidhorodetska et al., “L 98-59: A Benchmark System of Small Planets for Future Atmospheric Characterization,” Astronomical Journal Vol. 162, No. 4 (29 September 2021), 169 (full text).


Comments on this entry are closed.

  • Alex Tolley October 1, 2021, 12:15

    Speculatively, a world with an O2 atmosphere of sufficient pressure might enable an explorer to travel on the surface with suitable protection – an A/C rover with a recycled inert gas as a diluent. The explorer could even make short trips outside, rather like breathing in a hyperbaric O2 chamber. Tidal locking is your friend on such worlds, allowing for a more temperate region near the terminator, perhaps situated behind a range of mountains to hide from the direct rays of the star. Surface habitats could be situated there, but with sub-surface shelters as CME storm refuges. The habitats would need to be designed as faraday cages to protect any electronics that had not been replaced with optical systems.

    A harsh world to be sure, but possibly livable, given a subsurface water source.

    • Mike Serfas October 2, 2021, 14:14

      Well, they modelled 1 and 10 bar atmospheres of O2. A ship’s crew may rejoice that the planet for their emergency landing has an oxygen atmosphere, but as they ride their heat shield down into 10 bars of it this is likely to be a very short story. B)

  • Dave Moore October 1, 2021, 16:22

    This is an interesting system. One thing the authors don’t seen to have taken into account is the degree of tidal heating these planets are subjected to. They are big planets with orbital periods of 2+, 3+ and 7+ days, so I’m pretty sure they have a level of volcanism that makes Io’s look mild. In fact, I would think of b & c, the two innermost planets as Ios, but large enough to retain an atmosphere. They would be like Io, completely desiccated throughout–volcanism having turn their mantles inside out–so oxygen left over from the disassociation of water would, like Venus, have been drawn into the mantle. I would guess the main atmospheric gas you would have would be CO2 and SO2.

    d is interesting in that it gives some indication of the desiccation line for red dwarfs of this size. I think of it as Europa, only larger and hotter.
    A look at my trusty H2O phase diagram chart would indicate that it has a deep steam atmosphere that is a couple hundred kilobars at its base, probably directly over molten mantle, but there made be a layer of high pressure ice in-between.

    Just a note, steam at several hundred degrees Celsius and at a hundred kilobars is a superfluid that has a density of about 1.4 grams/cc–denser than water at 1 bar–so when you hear of hot ocean planets like d, you can account for their density with steam.

    • Andrei October 2, 2021, 12:30

      Yes it is indeed an interesting system.
      While the original paper note that the central star indeed would dry out these planets to a significant degree as the energy output is two orders of magnitude larger during the early days.
      But an oxygen atmosphere is only likely to happen on a water covered world, where water is split into the components while the hydrogen escape to space.
      On these worlds vulcanism would indeed bring new unoxidised material in contact with the atmosphere, so it would end up with CO2, SO2 but also a few other ones such as NOx.

  • andy October 1, 2021, 17:05

    And the authors point out in their conclusion that we can learn a great deal about the evolution of these planets depending on whether we detect water in their atmospheres.

    Determining whether the identified features in exoplanet transmission spectra are actually from planetary atmospheres is quite tricky in the face of stellar activity, particularly for the small stars that offer the most favourable planet:star radius ratios. Here’s a cautionary tale from the temperate sub-Neptune K2-18b, previously announced to have water in its atmosphere: Barclay et al., “Stellar surface inhomogeneities as a potential source of the atmospheric signal detected in the K2-18 b transmission spectrum”

    • Paul Gilster October 1, 2021, 19:05

      Excellent point!

  • Geoffrey Hillend October 1, 2021, 20:56

    I don’t get the paper posted above Stellar surface inhomogeneities. For one thing, there are not any water absorption lines in the photosphere of a red dwarf or any star with hydrogen burning, photosphere and plasma for that mater since it is above temperature needed to break water, H2O into hydrogen and oxygen. Consequently, we know that any water vapor spectra must be only from the exoplanet which will be seen as infra red absorption lines with transmission spectroscopy.

    I also am skeptical of the idea of false positives of abiotic oxygen because Earth had oceans, but it’s atmosphere had very little oxygen before life. Whether any abiotic oxygen has a high enough percentage in the atmosphere from Earth is still debatable.

    • Alex Tolley October 2, 2021, 12:04

      Isn’t the build up of O2 going to be a result of the amount of insolation and UV radiation received by the planet? Eniugh insolation to saturate the atmosphere with water vapor, and enough UV for photolysis of the water. As long as the hydrogen can escape, then O2 should build up once any sinks such as iron become saturated. One countervailing effect is the creation of ozone that will block UV and hence reduce photolysis.

      AFAIK, Earth never reached the needed conditions for abiotic O2 accumulation, even though it does constitute a minor source of O2 today, it pales into insignificance compared to oxygenic photosynthesis.

      What we really want to see is the presence of CH4 with O2. In an abiotic environment, the CH4 is quickly oxidized at a rate that exceeds its release, whilst in a biotic environment, the release is greater than its oxidation.

      On a desert world with relatively little surface water, it may be possible that the serpentinization reaction of olivine rocks and mantle water may exceed the capacity of a low pressure O2 atmosphere to oxidize the generated CH4, leaving both present in the atmosphere. Whether this is actually thermodynamically possible IDK, but I throw it out as a speculation that might confuse this hoped for biosignature.

      • Geoffrey Hillend October 3, 2021, 16:28

        No. UV radiation does not make O2. UV makes O3 or ozone. O3 blocks UV radiation. There has to be a lot of O2 in the air in order to make O3. Life is what gave us the O2. The photolysis of molecular oxygen or O2 into atomic oxygen or single atoms of O1. O plus O equals O2. O2 plus O1 is O3. The photo lysis of H2O into H and O can produce some oxygen, but not any where near what life makes. Also can we see that amount in an Exoplanet from Earth? The JWST should remove all doubt. The JWST date was moved from October 31, 2021 to December 18, but that is only a small delay.

        • Alex Tolley October 4, 2021, 12:04

          While you are technically correct that UV creates o3, the O3 is favored to revert to O2. Therefore the net result is that O2 builds up. The equilibrium is determined by the level of UV and temperature, but AFAICS, the equilibrium is many orders of magnitude ion favor of O2. Therefore, effectively, photolysis generates the O2 in the atmosphere, once the hydrogen is lost.

          Is that your understanding, or do you disagree with this?

          • Geoffrey Hillend October 4, 2021, 15:38

            We agree Alex Tolley. I read more into your first sentence which I though did not make any sense, since I thought that your wrote the O2 was dependent on the UV, but you clearly don’t mean that. Sorry about that. It’s obvious that one has to have some O2 first in order for the UV to make O3.

    • andy October 4, 2021, 13:33

      I think the issue is not so much “we detected water: is it in the planet or the star?” but more “is this dip in the spectrum actually caused by water or was the planet simply crossing a brighter region of the stellar surface when we were taking measurements at the required wavelengths, or in our out-of-transit ‘control spectrum’?”. As the paper puts it on page 2:

      Transmission spectroscopy is inherently a differential measurement – for each wavelength bin, the transit depth is measured from the difference between the brightness of the star in transit compared with out of transit. Spots and faculae can cause the stellar spectrum to provide an imperfect representation of the source of light occulted by the planets, primarily because the spectrum of the stellar cord that the planet transits differs from the average stellar spectrum (Rackham et al. 2018). This is further complicated by the fact that the star is rotating and the stellar photosphere visible to an observer changes on the timescale of a transit. If stellar inhomogenities are not correctly accounted for, residuals from an imperfectly subtracted stellar spectrum can be left in the observed transmission spectrum. Under or over corrected stellar features can confuse the interpretation of absorption features in exoplanet spectra.

    • Geoffrey Hillend October 4, 2021, 15:55

      Sorry about the mistake, but red dwarfs evidently do have water in the spectra. I was looking at one of my books on spectroscopy by Jonathan Tennyson called Astronomical Spectroscopy, 3rd edition and I found the spectra of an L-subdwarf star with water absorption spectrum in the near infra red on p. 103. I made an assumption that since the surface temperature of a red dwarf was above the temperature to break the molecular bonds of water molecule that there would be not be water in the spectra.

      As far as differentiating the exoplanet infra red absorption lines of water from the red dwarf, I guess that there is a difference in brightness between the star alone when the exoplanet is out of sight behind it and the star plus the exoplanet or there are two different spectral lines one for the star and one for the planet? I do know that there is line broadening caused by pressure and higher temperature.

  • James Essig October 3, 2021, 1:24

    As a guy who likes every thing big including scenarios for starships able to achieve extreme Lorentz factors, I find the above article on habitability of red dwarfs a whimsical detour from my normal focus.

    Red dwarfs if habitable or supportable of orbiting space colonies likely can be a way of the cosmically deep future.

    Red and orange stars will be with us a very long time.

    For example, orange or K-class stars live from about 20 billion years to about one trillion years.

    Red dwarfs or M-class stars live for between one trillion and 30 trillion years. Most stars are red dwarfs and orange stars.

    A red dwarf set on a planet in the habitable zone would likely appear about ten times the diameter of our Sun at sunset, but would look a deep red-brown: a perfect backdrop for a Thanksgiving turkey meal.

    Now, the Sun is a typical G-2 class star or yellow dwarf and has radiative power output of about [4 x (10 EXP 26)] watts.

    This is enough power to accelerate 40 starships to about 87 percent of the speed of light every second background time where each starship has invariant mass of about 100,000 metric tons. Today’s modern Aircraft Carriers in the U.S. Navy have mass of about 100,000 metric tons each.

    A long lived red dwarf puts out only about [4 x (10 EXP 22)] watts but this is still enough power to accelerate 40 starships to about 87 percent of the speed of light every approximately 3 hours background time where each starship has invariant mass of about 100,000 metric tons.

    However, low-end red dwarfs live about 3,000 times longer than our Sun will live so they will be available for spacecraft propulsion over the next tens of trillions of years.

    Light sails and beamed energy stations around red dwarfs will facilitate their use in powering cosmic light-cone infrastructure for tens of thousands of eons to come. An eon is one billion years.

    We can travel just about anywhere in our cosmic light-cone with velocities less than light-speed, say at about 98 percent of the speed of light which corresponds to a Lorentz factor of about 5. Cryogenic sleep will be a key technology for such enablement.

    Within our cosmic light-cone alone, which most astronomers now believe is a tiny fraction of our universe, there are approximately 10 EXP 24 stars. This is about as much as the number of fine grains of table sugar that would cover the entire United States about 100 meters deep. The number of planets orbiting stars in our cosmic light-cone is about ten times greater or about equal to the number of fine grains of table sugar which would cover the entire United States 1,000 meters deep. The number of moons orbiting planets orbiting stars in our cosmic light-cone is likely ten times greater yet or equal to the number of fine grains of table sugar that would cover the entire United States 10,000 meters deep.

    Given that most stars are red dwarfs and orange dwarfs, I believe our out-look over the next 30 trillion years is good. By then, hopefully we will have learned how to convert dark energy to hydrogen and helium to construct red dwarfs in a perpetual manner.

    • Alex Tolley October 3, 2021, 8:27

      Interesting calculations.

      Using your figures for the sun, and consuming Jupiter to make those 100,000 starships, launching 40 every second would use up Jupiter in 15 billion years, 3x longer than the sun has yet to burn before becoming a red giant.

      The red dwarf calculation, also consuming a Jupiter-sized planet would take 10,000x longer – 150 trillion years, also many times the age of such a star.

      It is mind-boggling to think of 40x 100,000 MT starships rolling off the shipyards every second and being propelled to a large fraction of c. This rate of production could be maintained for every G-type star, with appropriate production rates for different star types.

      It almost makes one thankful that there is a Fermi Question, as this suggests that there could be starships everywhere. Those shipyards are equivalent to those Von Neumann replicators turning planetary systems into copies of themselves, in this case starships at a frightening rate.

      What would we see? Stars going dark as their energy output is focused on propelling these starships out of their systems? Like Clarke’s short story “The Nine Billion Names of God”, – “Overhead, without any fuss, the stars were quietly going out.” Perhaps we might be lucky and catch a glimpse of the energies propelling those ships?

      As for human passengers and crew, that might be difficult. The earth currently has a population growth of 83 million/yr – about 2.5/s. So each starship will have just 2 or 3 people unless there is some serious effort to increase teh population growth rate to populate these ships. But robots – that would be much easier to crank out the numbers to fill the ships if needed. Perhaps each ship uses its human passengers to generate the many fertilized eggs and offspring in flight. Veritable population bombs to meet the need for new humans at the next star to populate the next generation of starships.

      • James Essig October 4, 2021, 12:13

        It definitely is mind boggling.

        I like the idea of habs on red dwarf orbiting planets or out in space orbiting red dwarfs. Somehow, I think I would enjoy the beautiful red light.

        It is amazing that, as you mentioned, Jupiter could provide resources for such a vast starship program for 15 billion years.

        Another awesome resource would include harvesting brown dwarfs.