≡ Menu

On White Dwarf Planets as Biosignature Targets

So often a discovery sets off a follow-up study that strikes me as even more significant in practical terms. This is not for a moment to downplay the accomplishment of Andrew Vanderburg (University of Wisconsin – Madison) and team that discovered a planet in close orbit around a white dwarf. This is the first time we’ve found a planet that has survived its star’s red giant phase and remains in orbit around the remnant, and quite a tight orbit at that. Previously, we’ve had good evidence only of atmospheric pollution in such stars, indicating infalling material from possible asteroids or other objects during the primary’s cataclysmic re-configuration.

The white dwarf planet, found via data gathered from TESS (Transiting Exoplanet Survey Satellite) and the Spitzer Space Telescope, makes for quite a discovery. But coming out of this work, I also love the idea of studying such a world with tools we’re likely to have soon, such as the James Webb Space Telescope, and on that score, Lisa Kaltenegger (Carl Sagan Institute, Cornell University), working with Ryan MacDonald and including Vanderburg in the team, have shown us how JWST can identify chemical signatures in the atmospheres of possible Earth-like planets around white dwarf stars. Assuming we find such, and I suspect we will.

The planet at the white dwarf WD 1856+534 is anything but Earth-like. It’s running around the star every 34 hours, which means it’s on a pace 60 times faster than Mercury orbits the Sun. The planet here is also the size of Jupiter, and what a system we’ve uncovered — the new world orbits a star that is itself only 40 percent larger than Earth (imagine the transit depth possible with white dwarfs transited by a gas giant!) In this planetary system, the planet we’ve detected is about deven times larger than its primary. Says Vanderburg:

“WD 1856 b somehow got very close to its white dwarf and managed to stay in one piece. The white dwarf creation process destroys nearby planets, and anything that later gets too close is usually torn apart by the star’s immense gravity. We still have many questions about how WD 1856 b arrived at its current location without meeting one of those fates.”

Image: In this illustration, WD 1856b, a potential Jupiter-size planet, orbits its dim white dwarf star every day-and-a-half. WD 1856 b is nearly seven times larger than the white dwarf it orbits. Astronomers discovered it using data from NASA’s Transiting Exoplanet Survey Satellite (TESS) and now-retired Spitzer Space Telescope. Credit: NASA GSFC.

So on the immediate question of WD 1856 b, let’s note that we have a serious issue with explaining how the planet got to be this close to the white dwarf in the first place. White dwarfs form when stars like the Sun swell into red giant status as they run out of fuel, a phase in which 80 percent of the star’s mass is ejected, leaving a hot core — the white dwarf — behind. Anything on relatively close orbit would be presumably swallowed up in the stellar expansion phase.

Which is why Vanderburg’s team believes the planet probably formed fully 50 times farther away from its present location, later moving inward perhaps through interactions with other large bodies close to the planet’s original orbit, with its orbit circularizing as tidal forces dissipated. Such instabilities could bring a planet inward, as could other scenarios involving the red dwarfs G229-20 A and B in this triple star system, although the paper plays down this idea, as well as the notion of a rogue star acting as a perturber. Other Jupiter-like planets, presumably long gone, seem to be the best bet to explain this configuration.

From the paper:

…a more probable formation history is that WD 1856 b was a planet that underwent dynamical instability. It is well established that when stars evolve into white dwarfs, their previously stable planetary systems can undergo violent dynamical interactions that excite high orbital eccentricities. We have confirmed with our own simulations that WD 1856 b-like objects in multi-planet systems can be thrown onto orbits with very close periastron distances. If WD 1856 b were on such an orbit, the orbital energy would have rapidly dissipated, owing to tides raised on the planet by the white dwarf. The final state of minimum energy would be a circular, short-period orbit. The advanced age of WD 1856 (around 5.85 Gyr) gives plenty of time for these relatively slow (of the order of Gyr) dynamical processes to take place. In this case, it is no coincidence that WD 1856 is one of the oldest white dwarfs observed by TESS.

Did you catch that reference to the white dwarf’s age? The 5.85 billion year frame gives ample opportunity for such orbital adjustments to take place, winding up with the observed orbit. Or perhaps we’re dealing with interactions with a debris disk around the star, as co-author Stephen Kane (UC-Riverside, and a member of the TESS science team) hypothesizes:

“In this case, it’s possible that a debris disc could have formed from ejected material as the star changed from red giant to white dwarf. Or, on a more cannibalistic note, the disc could have formed from the debris of other planets that were torn apart by powerful gravitational tides from the white dwarf. The disc itself may have long since dissipated.”

But back to Lisa Kaltenegger, lead author of a paper in Astrophysical Journal Letters probing whether an exposed stellar core — a white dwarf — would be workable as a target for the JWST, in which case we would like to look at planetary atmospheres to probe for the possibility of biosignatures. Here the news is good, for Kaltenegger believes that such detections would be possible, assuming rocky planets exist around these stars. WD 1856 b gives hope that such a world could exist in the white dwarf’s habitable zone for a period longer than the time it took for life to develop on Earth. The implications are intriguing:

“What if the death of the star is not the end for life?” Kaltenegger said. “Could life go on, even once our sun has died? Signs of life on planets orbiting white dwarfs would not only show the incredible tenacity of life, but perhaps also a glimpse into our future.”

Image: In newly published research, Cornell researchers show how NASA’s upcoming James Webb Space Telescope could find signatures of life on Earth-like planets orbiting burned-out stars, known as white dwarfs. Credit: Jack Madden/Carl Sagan Institute.

The Kaltenegger team used methods developed to study gas giant atmospheres and combined them with computer models configured to apply the technique to small, rocky white dwarf planets. The researchers found that JWST, when observing an Earth-class planet around a white dwarf, could detect carbon dioxide and water with data from as few as 5 transits. According to co-lead author Ryan MacDonald, it would take a scant two days of observing time with JWST to probe for the classic biosignature gases ozone and methane. Adds MacDonald:

“We know now that giant planets can exist around white dwarfs, and evidence stretches back over 100 years showing rocky material polluting light from white dwarfs. There are certainly small rocks in white dwarf systems. It’s a logical leap to imagine a rocky planet like the Earth orbiting a white dwarf.”

So we have a possible target we’ll want to add into the exoplanet mix when it comes to nearby white dwarf systems. WD 1856 is about 80 light years out in the direction of Draco. The white dwarf formed over 5 billion years ago, as noted in the paper, but the age of the original Sun-like star may take us back as much as 10 billion years. The post red giant phase allows plenty of time for orbital adjustment, drawing rocky worlds inward and circularizing their orbit. Will we find such planets in this setting in the near future? The hunt for such will surely intensify.

The paper is Vanderburg et al., “A giant planet candidate transiting a white dwarf,” Nature 585 (16 September 2020), 363-367 (abstract). The Kaltenegger paper is “The White Dwarf Opportunity: Robust Detections of Molecules in Earth-like Exoplanet Atmospheres with the James Webb Space Telescope,” Astrophysical Journal Letters Vol. 901, No. 1 (16 September 2020). Abstract.


Comments on this entry are closed.

  • Alex Tolley September 21, 2020, 14:15

    While a natural explanation is the most likely one, this also strikes me as possibly artificial.
    I can imagine ETI constructing this with the thought that something so obviously “unnatural” would be a clear signal. ;)

  • wdk September 21, 2020, 20:09

    Maybe it’s just filling in some details described above. If there were a Neptune or Jupiter sized planet in orbit around the main sequence pr0genitor of the white dwarf, then we could imagine a body with a large gravitational sphere of influence embedded in another gravity field with a large net outflow. There might be some significance in the shape of the outflow field from the star that’s releasing its outer envelope: it could be spherical, toroidal or some other shape, but the Jovian planet encounters a lot of flow across its bow even though it does not capture a large fraction of it. Whether you interpret this as drag, deflection and exchange of angular momentum or actual impacts, the net effect would have to be spiral inward, for some reason. A remaining issue would be how fast the envelope rips away. If it were a rapid event and the “Jupiter” were close in, then a lot of the planetary body would be sheared away. Would not be surprised that each of these bodies have significant magnetic fields. Now considering that the planet has been drawn into a tight orbit, do you suppose that the two bodies would have some sort of magnetic co-alignment?

  • Robin Datta September 21, 2020, 22:52

    A resort to panspermia and its variants and/or abiogenesis may support a wide variety of conjectures suggesting or explaining biosignatures, their ircidence, prevalence, and distribution. Small molecules may survive where megastructures don’t.

    For intelligence to survive incineration in a red giant in a form of biology based on large complex, heat-disruptable molecules, that biology has to move away, either within the same stellar system or beyond it. Either as the same biology or after transitioning to a post-biologic state.

    Unless, of course some postulate an almost-impossible xenobiology. The closer we get to the limits of detectability and conceptual grasp, the closer it is to implausibility.

  • John walker September 21, 2020, 23:10

    When considering the prospects of the habitability of rocky planets on large multi billion year time scales such as in white dwarf or red dwarf systems it’s important to consider that terrestrial planets also evolve and won’t have the same longevity that their host stars do. Core solidification, collapse of magnetic field, termination of plate tectonics, end of vulcanism, terminal erosion of atmosphere following cessation of mantle outgassing. All of these aspects are likely to occur on time scales that are a tiny fraction of the white or red dwarf stellar life times.

  • Geoffrey Hillend September 22, 2020, 0:19

    As usual, I have to reject the migration theory. The tides of a white dwarf are only strong when the gas giant is already in a close orbit. The gas giant around the white dwarf WD 1856+534 formed close to the star when it was born from the protoplanetary gas cloud. WD 1856b survived it’s parent star’s red giant phase and was pulled a short distance into a tighter orbit after the gases of the planetary nebula floated away. Some gas was recaptured by the gas giant.

    The helium burning core is where the fusion of hydrogen into helium is still small compared to the outer layers of the star and changes into a carbon core and does not swallow the gas giant, but only the outer hot plasma of hydrogen.

    The gravity of the white dwarf is clearly not strong enough to tear apart the gas giant nor the formation of the white dwarf.

  • Geoffrey Hillend September 22, 2020, 0:23

    On the other hand, Maybe there is a lot of drag on the gas giant by the star’s hot plasma once it is inside the star, so it is forced to migrate inwards from a further orbit close to the carbon core of the star until it becomes a white dwarf?

  • torque_xtr September 22, 2020, 4:36

    A skimming transit by almost a point source with continuous emission – what a wonderful target for transit spectroscopy and atmospheric study! Though the atmosphere seems not to be extended, no rises more than 0.5…1% from refraction lensing and/or forward scattering on full transit curves.

  • Douglas Muir September 22, 2020, 7:59

    It’s very unlikely that any sort of biological life could survive the transition to a red giant.

    That transition involves an increase in stellar luminosity of around three orders of magnitude, sustained over tens of millions of years. So, for instance, in the Sun’s case? Peak luminosity will reach around 2000 times its current value. That’s enough to heat the surface of the Earth to around 1900 Kelvin. That’s hot enough to melt steel and boil lead into vapor. The Earth’s surface will become a magma ocean many miles deep.

    Mars, same-same. Even the moons of Jupiter will have surface temperatures around 800 K — hotter than Mercury is today. Europa’s icy cover will first melt and then boil away, leaving behind only a tiny, desiccated rocky core. Further out, Titan will lose its atmosphere, its seas, and about half of its mass; what remains will be a hot, airless rocky body rather similar to Earth’s Moon.

    The dramatic surge in luminosity will be pretty much impossible for any form of biological life to survive. But even if it somehow could, there’s another problem.

    When a star becomes a red giant, it begins to emit a “superwind” — a solar wind millions of times more intense than the relatively mild wind that comes off the Sun. The superwind is both denser and moving much faster. Come the day, it will blast right through the Earth’s magnetic field and will start ablating the atmosphere. At full power, the Sun’s superwind would strip the Earth of its atmosphere in a geological eyeblink — a few tens of thousands of years. That’ll be true of any remotely terrestrial planet.

    If you want to imagine intelligent life somehow migrating outwards, knock yourself out. But without intelligence? The passage through red giant status will kill pretty much every plausible form of biological life dead, dead, dead.

    Doug M.

    • Paul Gilster September 22, 2020, 9:05

      In the case of WD 1856+534, we’re dealing with a white dwarf that formed almost 6 billion years ago, following the star’s red giant phase, long enough to allow a rocky planet in a circularized orbit to spawn its own version of life. The original star is thought to have formed about 10 billion years ago.

      • Douglas Muir September 22, 2020, 14:19

        Ah, okay.

        Well… life around such a white dwarf wouldn’t be flat-out impossible, but it would face its own set of challenges.

        1) A star this old is likely to be pretty metal-poor. [googles] Holy socks, this poor old star has an Fe/H of -8.8 dex! That means it has barely one *billionth* as much iron as the Sun! Okay, in the case of this particular star system, that’s probably the end of the conversation right there. But for white dwarfs generally…

        2) Because of their low luminosity, white dwarves have a tiny, narrow habitable zone. Even very modest perturbations could knock a planet out, and of course any hypothetical planet would be tidally locked.

        3) White dwarves evolve — they get yellower and dimmer with time. With a Teff of around 4700 K, this star is already yellower than the Sun, and it’s a lot less luminous than it was a few billion years ago. Any life that evolves on a white dwarf planet has a limited time before the dwarf cools to the point where the planet is no longer habitable.

        It’s not unthinkable, I guess, and goodness knows the universe is full of surprises. But a white dwarf system wouldn’t be the first place to look for biosignatures.

        Doug M.

        • Alex Tolley September 22, 2020, 18:16

          I take it you mean that the low metallicity of the WD means that the primordial material that formed the planets in its earlier stage would also be metal poor. But if the planet or its moons have a rocky core wouldn’t that have been enough to support life?

          As it seems to be a Jupiter, I would be skeptical about life on/in such a world. We are only consiering life as a maybe in its moons if it has any.

          Supposing this world has a retinue of icy moons and migrated in after the formation of the WD, if the moons are still icy, we have no way to detect life in any subsurface ocean. If the ices melted, then we have a possibility that life is now emerging into the surface layers of these ocean worlds.

          The other possibility is that there are other worlds more conducive to life that have migrated in to the new HZ and that is where we would look for biosignatures.

          • Douglas Muir September 23, 2020, 5:47

            At dex -8.8? I’m skeptical that rocky planets could even exist.

            Fe/H is a logarithmic scale using log 10. So that means Fe/H -8.8 is 10 ^ -8.8 the ratio of the Sun. That is around one *billionth* the Sun’s iron.

            There was a flurry of publicity a couple of years back about rocky planets being common around “low metallicity” stars. But they were talking about stars with dexes between -0.5 and -1 — between 1/3 and 1/10 solar Fe/H. Rocky planets are exceedingly rare below dex -2, and I’m not aware of any rocky planet yet found around a star with a dex lower than -4.

            It’s true that the Fe/H ratio is just a crude approximation of general metallicity. You could have low Fe/H and still be relatively rich in, say, silicon. But when FE/H is this low, we can be pretty sure that we’re looking at an ancient Population II star that is just very metal-poor generally. This star is a survivor of the early days of the galaxy, before there’d been time for a lot of supernovas to enrich the interstellar medium. So it probably never had a lot of silicon or magnesium or other lithic elements to work with.

            I won’t say it absolutely couldn’t possibly have a rocky planet. One could imagine a scenario where a small gas giant planet grabbed enough metals to build a small rocky core, and then the giant’s atmosphere got boiled and ablated away during the red giant phase. It’s a long shot, but one could imagine it.

            Alternately… if you really want to get weird? This star is a carbon dwarf, rich in carbon formed by the fusion of helium during its red giant phase. We know that dying red giants can dredge up carbon from their cores and then expel it into space. That’s where most of the carbon in your body comes from. So one could imagine a scenario where secondary planet formation takes place in a carbon enriched medium around the star, and you get “carbon planets”. Theory suggests these should be possible, though we haven’t had one confirmed yet afaik.

            But even if so, you still have the problem that a lot of key elements for life as we know it — sulfur, potassium, phosphorus, iron — are going to be vanishingly rare, as rare as gold and uranium on Earth. Could you have a biochemistry based almost entirely on CHON, with just the occasional stray atom of heavier stuff? Who knows, but it does seem like another hurdle.

            Finally, note that *most* white dwarfs will be low-metallicity — maybe not as bad as dex -8.8, but still metal-poor compared to the Sun. That’s because most white dwarfs are old. It takes 11 billion years for a solar mass star to become a white dwarf, and 11 billion years ago the interstellar medium had a lot less metal. Now, more massive stars age faster, so white dwarfs formed from them can be younger and more metallic. However, the stellar mass distribution function means that more massive stars are a lot rarer. So, most white dwarfs are relatively old stars that are relatively metal-poor.

            Doug M.

            • Alex Tolley September 23, 2020, 14:38

              Interesting points. Given this, what would the composition of this “Jupiter” be like? All H/He with no rocky elements for a core, or as you suggest, just a very small one? It would also not have any icy moons (no oxygen for water) or silicates for the core. There would be no rocky planets around this star either, now or in the past. It would have no comet cloud nor icy/silicate “Kuiper Belt” objects either, if I understand your argument correctly and take it to its logical conclusion.

              Are there other similar WDs or very low metallicity stars of a similar age? Does the Kepler data support your argument of an absence of rocky worlds. either Venus-like or water worlds compared to stars born later with higher metallicities?

              • Douglas Muir September 24, 2020, 3:55

                “Given this, what would the composition of this “Jupiter” be like? All H/He with no rocky elements for a core, or as you suggest, just a very small one?”

                We don’t know! But those are reasonable guesses.

                “It would also not have any icy moons (no oxygen for water) or silicates for the core. There would be no rocky planets around this star either, now or in the past. It would have no comet cloud nor icy/silicate “Kuiper Belt” objects either, if I understand your argument correctly and take it to its logical conclusion.”

                That’s broadly correct. An Fe/H ratio of -8.8 is very, very metal poor. It doesn’t have *no* metals, but it has them in very tiny amounts.

                Now, a circumstellar disk can act like a gigantic mass spectrometer, concentrating certain elements and dispersing others. So it’s conceivable that a disk could concentrate those trace metals enough to form… something. I’m not aware of anyone modeling this, though, so I can’t really say more.

                “Are there other similar WDs or very low metallicity stars of a similar age?”

                Fe/H -8.8 is actually really, really low. It’s so low that I wonder if it’s a typo or error. I almost think it has to be, because if it’s correct, this would be the lowest metallicity star found with any sort of planet.

                But there are lots of stars with metallicities between -2 and -5.

                “Does the Kepler data support your argument of an absence of rocky worlds. either Venus-like or water worlds compared to stars born later with higher metallicities?”

                The Kepler data shows *no* planets around stars with metallicities below -2. In fact, the lowest confirmed metallicity in the Kepler data is around -1.1, or about 1/12 that of the Sun. See for yourself:


                Now, to be fair, that’s almost certainly a data artifact. For reasons beyond the scope of this comment, Kepler probably undersampled low-metallicity stars. We should get a better, more balanced survey with TESS.

                Nevertheless, there is a clear dropoff in planet abundance at low (>-0.5) observed metallicities. Counter-intuitively, this dropoff is much sharper for giant planets than for rocky ones!
                There are lots of rocky planets around stars with Fe/H between -0.5 and -1 — between a third and a tenth of Solar metallicity.
                But there are very few gas giants in that range.

                Why? You’d expect the opposite, right? Well, current thinking is that gas giants need dense cores to get their formation started. And that’s harder when there’s just not much dense stuff to work with.

                IIUC, one theory is that some of the rocky planets we see around low-metallicity stars are bodies that would have formed the cores of giants, if there had been more metals to work with. However, less metal –> these bodies took much longer to form –> by the time they formed, the circumstellar disk had dispersed and there was no more material for accretion into a giant.

                Meanwhile, metal rich stars may have slightly *fewer* rocky planets. Why? Well, one theory is that the abundance of metals -> gas giants form very quickly and easily -> there’s more time for them to migrate and otherwise interact with other planets -> they’re more likely to devour or perturb small rocky planets early on.

                The topic is a complex one, and IANA an astrophysicist. I’m an interested amateur, and I try to keep up with the literature, but it’s all very much a moving target. So take all the above with a grain of salt. But if you take a few minutes to google, you’ll find lots and lots of papers — it’s a very hot topic.

                Doug M.

  • Henry Cordova September 22, 2020, 8:02

    We have to remind ourselves that even though white dwarfs are very tiny stars, they are still very dense. According to the wikipedia entry on this star, WD 1856+534 is no lightweight red dwarf, it weighs in at about half a solar mass. Although this is nowhere near the upper limit of what a star can mass, its still heavier than most of the stars in the galaxy. The vector field of a star’s gravitational acceleration depends only on its mass and the distance from its center. A star’s density and its size are irrelevant to the contours of its gravitational surroundings.

    Does anyone know anything about the overall geometry of this system? In other words, what are the dimensions of both the star and its companion? What is the orbital radius of the planet? Is it actually “skimming” the surface? Did it ever?

    Also, the white dwarf is a member of a triple system, it orbits the center of mass between itself and a pair of closely orbiting red dwarfs nearby. How close are they, could they be influencing the gravitational dynamics of the system? Could they have influenced the evolution of WD 1856+534 and its planet?

    If the planet did indeed once orbit inside the surface of a highly evolved red giant, perhaps even in its convective zone, how would this have affected the evolution of the primary? Regardless of what the circumstances of this particular system are, I don’t think it is very representative of other stellar systems at all. It may be unique, an interesting laboratory to study unusual physical properties, but probably not much of a guide to what may be happening elsewhere.

  • Michael Fidler September 22, 2020, 10:02

    Well, heck 6 billion years, it only took us 1/10 that time to get from itty bitty one celled animals to humans. So what may be on the dark side of the fried tidally locked planet? Hard shelled, molecular acid for blood aliens!

  • charlie September 22, 2020, 13:13

    “WD 1856b, a potential Jupiter-size planet, orbits its dim white dwarf star every day-and-a-half. ”
    I myself look forward to the day they find a planet named WD-40 …

    • Alex Tolley September 22, 2020, 18:19

      …helps unfreeze frozen states of evolution allowing it to slip into high gear with a cornucopia of new life forms emerging…

  • Michael September 22, 2020, 16:11

    A Neptune object with a strong magnetic field could have been dragged inwards, the magnetic field would have a very large area of capture or influence in raising tides on the RG’s surface.

  • Mike Serfas September 22, 2020, 16:59

    The NASA image seems off with a ratio of radii of about 2.3. Lazily, I checked the numbers against Wikipedia ( https://en.wikipedia.org/wiki/WD_1856%2B534 ) – which says the star is 0.0131 solar radii and the planet is 10.4 Earth radii (about the same as Jupiter), giving a ratio of radii of 7.28, consistent with your text. This transit must be as amazing as you describe. How deeply can any EM wavelength shine into a Jupiter anyway?

  • Ryan MacDonald September 23, 2020, 0:21

    Another great article as always, thanks Paul. Minor correction: should read 5 transits for a detection of CO2 and H2O (not 25 transits). I’ve also produced a video covering our paper – and exoplanet biosignatures more generally – which some of you may find interesting: https://youtu.be/0zAo_CXPncY

    • Paul Gilster September 23, 2020, 6:30

      Thanks for catching the transit number, Ryan. Much appreciated — I’ll fix that in the text right away. Thanks also for the video link!

  • Michael September 24, 2020, 12:57

    It always pains me when they say the these stars are dying, they are not dying but giving birth too enormous amounts water and chemicals essential to life ! If a techalien species could capture those chemicals with say magnetic fields countless world’s could be created.

    • Alex Tolley September 24, 2020, 23:29

      Stars as natural [artificial?] factories to produce the needed elements for living worlds. I suspect it would be easier to just collect the material in Oort clouds to seed dry worlds with the needed water and chemicals to create petri dish conditions for life to be created or seeded. Artificial habits like Dyson swarms might offer the most living space for life compared to planets, if that was the aim.

      I could imagine countless such swarms being used to evolve life which is then sent to target worlds to start the process of evolution there. These worlds may even be prepared with infusions of water and essential compounds to create the optimal conditions for life to take hold.