≡ Menu

A White Dwarf Proposal for Kepler

With four years of collected data at hand, Kepler scientists will remain busy even with their spacecraft hobbled. We now know that we’re not going to get Kepler back to full working order following the degradation of two of its reaction wheels, but as this report noted on August 19, possibilities remain for scientific studies using the two remaining reaction wheels aided by thrusters to control the spacecraft’s attitude. And as we’re finding out, a ‘two-wheel’ Kepler mission may still offer opportunities, one of the more fascinating of which is our subject today.

The proposed target is white dwarf stars, the remnants of stars whose mass is not high enough to produce a neutron star as they evolve past the red giant phase. A typical white dwarf has a mass similar to that of the Sun, but a volume close to that of the Earth. While Sirius B, at 8.6 light years out, is the closest known white dwarf, eight white dwarfs are believed to be present among the one hundred closest star systems to the Sun. And while we don’t normally think of white dwarfs as capable of sustaining life-bearing planets, maybe we should take another look. A new paper points out that stars like these can provide an energy source for billions of years.


Image: A white dwarf as compared with the Earth. Credit: Ohio State University/Richard Pogge.

To orbit in a white dwarf’s habitable zone requires an orbit in the range of 0.01 AU for temperatures that could support liquid water on the surface to exist. This is a habitable zone that evolves with time, starting off too hot for liquid water and eventually becoming too cold to sustain it, but surprisingly, a white dwarf planet in this kind of orbit could have a maximum of eight billion years of habitability to support whatever life might form there. Lead author Mukremin Kilic (University of Oklahoma) and team calculate an overall habitable zone extending from 0.005 AU to 0.02 AU.

Could such planets exist? Clearly, an expanding red giant will consume its inner planets before contracting into a white dwarf, so planets within 1 AU or less will presumably have to arrive after the red giant phase. But possibilities exist: We’ve found planets orbiting close to the exposed core of a red giant (KOI 55.01 and KOI 55.02) and we’ve even found planets around pulsars. There are models that produce short period planets in billion-year timescales that seem to be applicable to white dwarfs, with planet formation from nearby gas being one scenario and the capture or migration of planets from much further out in the system being another.

Add delivery of water through cometary impacts and the presence of a habitable world in either scenario seems a bit less unlikely. We can also throw into the mix the fact that 30 percent of the white dwarfs near the Sun show metal-polluted atmospheres perhaps caused by the accretion of rocky debris. Indeed, some 4.3 percent have known debris disks, the latter a demonstration that interactions within the system can send asteroids, moons or small planets close to the white dwarf. But if short-period planets around these stars do exist, we have yet to find them, a fact the paper attributes to our lack of observational data for a sufficient number of stars.

Kilic and team argue that Kepler in its two-wheel mode offers an opportunity to run the kind of survey that would find the first exoplanets in a white dwarf habitable zone. From the paper:

If the history of exoplanet science has taught us anything, it is that planets are ubiquitous and they exist in the most unusual places, including very close to their host stars and even around pulsars (Wolszczan & Frail 1992). Currently there are no known planets around WDs, but we have never looked at a sufficient number of WDs at high cadence to find them through transit observations. It is essentially impossible to find Earth-Jupiter size planets around WDs by any other method (Gould & Kilic 2008). If habitable planets exist around WDs, the proposed Kepler imaging survey will find them.

The proposed survey would require 200 total days of observing time examining 10000 white dwarfs in the Sloan Digital Sky Survey imaging area, the great benefit being that Kepler’s wide field of view would allow a large number of white dwarfs to be observed at the same time. The researchers believe up to 100 planets will be identified in the habitable zone, an extension of the Kepler planet-hunting charter extended to a new set of targets. The paper continues:

Biomarkers, including O2, on such planets can be detected with the JWST [James Webb Space Telescope]. Hence, even though this is a completely unexplored search area for transiting planets, the scientific yield of the proposed survey will be enormous.

Exactly so. Remember this about white dwarfs as transit targets. The stars are about the same size as the Earth, so Earth-sized and smaller planets should be easy to detect as they pass in front of the primary. And once a planet in the habitable zone has been identified, the high contrast ratio between the planet and the host white dwarf means that future telescopes should be able to run the biomarker searches mentioned above. Is it possible that the first evidence of life on an exoplanet may come not from a G- or even an M-class system, but a white dwarf?

The paper is Kilic et al., “Habitable Planets Around White Dwarfs: an Alternate Mission for the Kepler Spacecraft,” a Kepler white paper available as a preprint. For more on white dwarf planets, see Habitable Worlds around White Dwarf Stars. Thanks to Antonio Tavani for the pointer to this work.


Comments on this entry are closed.

  • Antonio Tavani September 4, 2013, 10:21

    In questo interessante articolo, viene descritta la possibilità della presenza di pianeti, in “zone abitabili” molto vicino a “nane bianche”.

    Però, mi chiedo, se la presenza di una stella così massiccia(rispetto al vicino pianeta)non possa innescare un calore da “effetto marea” simile a quello(per esempio)che rende il satellite di Giove “Io” un luogo sottoposto a continue eruzioni vulcaniche.

    Questo, è un dubbio che mi viene.

    Saluti da Antonio Tavani

    As rendered by Google Translate:

    In this interesting article, describes the possibility of the presence of planets in “habitable zones” very close to “white dwarfs.”

    However, I wonder if the presence of a star so massive (compared to the nearby planet) can not trigger a heat “tidal effect” similar to that (for example) that makes the satellite of Jupiter Io a place subjected to continuous volcanic eruptions.

    This is a question that I get.

    Greetings from Antonio Tavani

  • Roger September 4, 2013, 10:21

    Wouldn’t the planet have to be orbiting exactly edge on to the Earth around the White Dwarf in order to detect using the transit method? Since the White Dwarf is so small any deviation from edge on and the planet does not block the light to produce a transit when viewed from the Earth. Orbiting a large G or even K-class star a planet transit viewed from the Earth would be much more likely, if relatively harder to detect with less % light blocked. I wonder if the paper took this into account when coming up with its figure of 100 planets being able to be found in the habitable zone with a re-tasked Kepler?

  • Eric September 4, 2013, 11:02

    Let’s say your standing on a non-tidal locked Earth analog with a 24 hour clockwise rotation period, which is itself orbiting a white dwarf in its habitable zone, making a complete clockwise orbit every 10 hours. What would the sky look like? Would the stars (and sun) be making loops?

  • Rob Flores September 4, 2013, 14:10

    Re: 24hr, sky.

    From what I’ve read, your sky would be infused with volcanic ash.
    If a planet did manage to “Aquire” a 24 hour spin, the white dwarf would
    be trying it’s darnest to tide lock it. (angular momentum is a bitch)
    The crust would be an unstable mess for the time it takes the planet
    to tide lock.

    As far finding habitable enviroments.
    Once again, a two body system is probably a best best, with a sizeable moon having something like a day/night cycle. (probably 48-72hrs, being
    desirable and within the realm of the possible.)
    While a body close to the White dwarf will attract comets and other
    matter valuable to life, the stuff will be screaming in to the .05 AU orbit.
    If planet/moon survives such a deluge then once the skies are clear I grant
    a life form could arise there, especially with such long stable period.

  • Michael September 4, 2013, 14:12

    The gravitation fields around White Dwarfs are very strong. Couple this with a fast orbital velocity of a planet close in around it and you could get an average mass of a car impacting a planet in orbit around it with more energy than 50 or more atomic bombs. No wonder it is dusty around these stars, you wouldn’t want to get any of that dust in your eye that’s for sure!

  • David September 4, 2013, 14:43

    What you say is roughly correct. More precisely, the probability that
    a planet will transit its star is given by the ratio of the stellar radius
    to the orbital radius assuming circular orbits. For a Hot Jupiter around a
    Solar -type star, that probability is around 10 percent and for a
    planet in the HZ around around a white-dwarf, that probability is
    around 1 percent. That is where the estimate that up to a 100 transiting White dwarf
    systems will be found, that is if all of the 10,000 white dwarfs that they observe have a planet
    in the HZ, than 1 percent of 10,000 is 100. If no white dwarfs have a planet in its HZ, then
    of course the probability is zero. Assuming some white dwarfs have planets in its HZ means
    that the number that is found will be greater than zero, but less than a 100. Thus making the
    reasonable assumption that the orientation of the orbits of planets around white dwarfs are
    random, means that we also get a good idea of how often white dwarfs have planets in their HZ.

    The other important factor is that the transit depth, or the fraction of light from the white dwarf is blocked, is given to first order,by the ratio of the radius of the planet squared, divided by the radius squared of the white dwarf. Therefore even a terrestrial size planet the size of Mars will produce an obvious drop in light from the white dwarf, and an Earth size planet will completely block the light. Also because of the size of a white dwarf, transits will only last on the order of a few minutes so high cadence observing is needed.

  • David September 4, 2013, 15:50

    I should also add what I wrote above assumes that there is only one planet orbiting the
    white dwarf. There of course could be more in which case one should say that the number of
    white dwarfs found with a transiting planet tells us how frequently do white dwarfs
    have at least one planet in its HZ.

  • Ronald September 5, 2013, 5:38

    Slightly off-topic, but very interesting:
    A new research report confirms that GJ 1214b has a lot of water vapor in its atmosphere and may therefore indeed be a water world/ocean planet (the same may be the case for Kepler-22b).

    If this is the case then this seems to justify the distinction between largely gaseous mini-Neptunes and the rocky core/water mantle/thin atmosphere ‘real’ super-Earths.

  • David Cummings September 5, 2013, 7:15


    This article predicts we won’t find life on these planets, and that we might even be fooled by oxygen signatures from atmospheres where all the hydrogen has been cooked out. The article nevertheless reaches the same conclusion as reached here at Centauri Dreams: planets around dwarf stars deserve a closer look.

  • wajai September 5, 2013, 8:49

    In the same vein (white paper proposals for Kepler) there is another one for searching the habitable zone of bright stars: http://astrobiology.com/2013/09/kepler-searching-the-habitable-zones-of-the-brightest-stars.html

  • wajai September 5, 2013, 10:04

    There is a new model to sort out mini neptunes from big terrestrials
    take a look at it :http://www.astrobio.net/exclusive/5664/a-super-time-for-superearths
    If you put the estimated radius and mass of Gj 1214b on the mass-radius diagram it indeed agrees with a ocean world.
    That was strongly suspected since Hubble infrared measurments of the atmosphere in 2012.

  • andy September 5, 2013, 14:32

    As David Cummings points out, the initial hot state of the white dwarf is nasty. Worse still if you form the planets via a process that leaves behind a helium-burning subdwarf (e.g. KOI-55 = Kepler-70)… we know that planets can end up in short-period orbits around post-red giant stars, but this is often not a very nice place to be.

    My personal suspicion is that the only way to get habitable planets around a white dwarf is to “cheat”: have a second star in the system that also goes through a red giant phase. This would potentially allow for a new epoch of planet formation, from material ejected from the giant star companion and captured into orbit around the white dwarf. The delay between the formation of the white dwarf and the formation of the planets would help to avoid the problems associated with the hot young white dwarf.

    I have to wonder what kind of planet formation theory any inhabitants of such a system would come up with, given the potential existence of multiple generations of planets of very different ages!

  • torque_xtr September 5, 2013, 15:12

    Possibly it’s raher too early to speculate about possible WD systems… Just everything there is so different! No water by comets: any impact involving WDHZ body and a comet on elliptical orbit would have delta-V on the order of hundreds of km/s, much more than planetary escape velocity. All the water will be dispersed for good, and great deal of target’s material – too. (For the super-earth-class worlds on wide orbits it would be possible to hold some…) Then, the tidal effects would have great influence in all cases except the single planet on a circular orbit. If there are other bodies, the tidal heating could become comparable to the stellar irradiation, or even a dominant heat source, and a very stable one, since the energy comes from the orbital motion, and specific kinetic+potential energy is huge. Anyway, it would be hellish – earthlike world with oceans and continents would not be habitable. But imagine a super-Europa five times more distant than outer HZ edge, slowly spiralling inwards on the course of 10 billion years, with a 100 km-deep ocean kept liquid and nutrient-saturated by the raging volcanism on the floor, and a plenty of UV from the WD at the surface, to provide biogenesis…

  • JB September 5, 2013, 20:39

    There around 2000 mostly amateur astronomers in the American Association of Variable Star Observers who devote their time to finding variable stars and plotting their light curves – including pairs of eclipsing (transiting) stars. If a white dwarf that is near enough and bright enough to be visible in an amateur telescope exhibits frequent transits where it’s light is totally or almost totally blocked out by a planet it would be pretty easy to detect even with amateur equipment. Someone ought to get some of them working on this.

  • coolstar September 6, 2013, 18:27

    Lots of interesting proposals in these new white papers. This one is unlikely to make the cut simply because it can be done well from the ground. First rule of astronomy in space: don’t do it unless you can’t do it from the ground! Spacecraft observing time is just much too valuable to use it for anything that can be done from the ground. And for all the reasons mentioned in the comments, it’s extremely unlikely that a planet in the habitable zone of a wd will in fact be habitable (unless some quite advanced ETs decide to settle there from another star system!).

  • NS September 10, 2013, 3:55

    In general I agree with coolstar that we shouldn’t do space missions for things that can be done from the ground, but in this case the spacecraft is already in place and may not be usable for anything else. Provided the tracking network time etc. is available it seems wasteful not to utilize the spacecraft however we can.

  • Aleksandar Volta February 3, 2014, 22:46

    Considering white dwarfs are a more rarer type of cosmic object, any findings from Kepler regarding, if not HZ Gaian sized planets, then at least the fraction of Gaian sized planets around them in any orbital arrangement whatsoever (enough to know that they exist there), would be of extreme value for planetology. If White dwarf show high enough fraction of such planets, then any other star, particularly main sequence stars, should be hosts to Gaian planets a few orders of magnitude more likely than stellar remnants like pulsars, neutron stars and white dwarfs. So I’m very happy that Kepler can still function and has turned its research in a different, but no less interesting direction. Same about those bright stars and HZ planets suggestions. Bright stars are also more rare than main sequence ones, so any findings there too would speak volumes on the statistics we wish to create for planetology.

    While on the subject of white dwarfs, I can’t find exact references, but there seems to have been a couple of unconfirmed detections of companions around Van Maanen’s Star, ranging from brown dwarf estimates to massive jovian in year or two orbit. I believe the paper was by Makarov, 2004. It had the same controversial odyssey of confirm and denial the type Gliese 581 g is experiencing today (only less press, and virtually no media exposure/promotion). I wonder if it has been anything since? I believe that even if the detections were faulty, this system, considering its the closest white dwarf after Sirius & Procyon B, deserves closer investigation.

    Thanks for another awesome article Paul! :)