≡ Menu

A Cometary Solution for KIC 8462852?

KIC 8462852 is back in the news. And despite a new paper dealing with the unusual star, I suspect it will be in the news for some time to come, for we’re a long way from finding out what is causing the unusual light curves the Planet Hunters group found in Kepler data. KIC 8462, you’ll recall, clearly showed something moving between us and the star, with options explored by Tabetha Boyajian, a Yale University postdoc, in a paper we examined here in October (see KIC 8462852: Cometary Origin of an Unusual Light Curve? and a series of follow-up articles).

To recap, we’re seeing a light curve around this F3-class star that doesn’t look anything like a planetary transit, but is much more suggestive of debris. Finding a debris disk around a star is not in itself unusual, since we’ve found many such around young stars, but KIC 8462 doesn’t appear to be a young star when looked at kinematically. In other words, it’s not moving the way we would expect from a star that has recently formed. Moreover, the star shows us none of the emissions at mid-infrared wavelengths we would expect from a young, dusty disk.

Jason Wright and the team at the Glimpsing Heat from Alien Technologies project at Penn State have taken a hard look at KIC 8462 and discussed it briefly in a recent paper (citations for both the Boyajian and Wright papers are at the end of this entry). It seems entirely reasonable to do what Wright did in referencing the fact that the light curve we see around the star is what we would expect to see if an advanced civilization were building something. That ‘something’ might be a project along the lines of a ‘Dyson swarm,’ in which huge collectors gather solar energy, or it could be a kind of structure beyond our current thinking.

We all know that the media reaction was swift, and we saw some outlets acting as if Wright had declared KIC 8462 an alien outpost. He had done no such thing, nor has he or the Penn State team ever suggested anything more than continuing investigation of this strange star. What seems to bother others, who have scoffed at the idea of extraterrestrial engineering, is that Wright and company have not explicitly ruled it out as a matter of course. The assumption there is that no other civilizations exist, and therefore we could not possibly be seeing one.

I come down on the side of keeping our options open and studying the data in front of us. We have a lot of work ahead to figure out what is causing a light curve so unusual that at least one of the objects briefly occulting this star caused a 22 percent dip in its flux. That implies a huge object, evidently transiting in company with many smaller ones. There seems to be no evidence that the objects are spherically symmetric. What’s going on around KIC 8462?

A new paper from Massimo Marengo (Iowa State) and colleagues looks at what Tabetha Boyajian identified as the most likely natural cause of the KIC 8462 light curves. All I have at this point is the JPL news release and a release from Iowa State — the paper has not yet appeared online — describing evidence for a swarm of comets as the culprit. The study, which has been accepted at Astrophysical Journal Letters relies on Spitzer data dating from 2015, five years later than the WISE data that found no signs of an infrared excess.


Image: This illustration shows a star behind a shattered comet. Is this the explanation for the unusual light curves found at KIC 8462852? Image credit: NASA/JPL-Caltech.

If there had been a collision between planets or asteroids in this system, it was possible that the WISE (Wide-field Infrared Survey Explorer) data, taken in 2010, reflected conditions just before the collision occurred. Now, however, we can rule that out, because Spitzer, like WISE, finds no excess of infrared light from warm dust around KIC 8462. So the idea of planet or asteroid collisions seems even less likely. Marengo, according to the JPL document, falls back on the idea of a family of comets on an eccentric orbit. He’s also aware of just how odd KIC 8462 is:

“This is a very strange star,” [Marengo] said. “It reminds me of when we first discovered pulsars. They were emitting odd signals nobody had ever seen before, and the first one discovered was named LGM-1 after ‘Little Green Men… We may not know yet what’s going on around this star, but that’s what makes it so interesting.”

It would take a very large comet indeed to account for the drop in flux we’ve already seen, but a swarm of comets and fragments can’t be ruled out because we just don’t have enough data to make the call. I assume Marengo also gets into the fact that a nearby M-dwarf (less than 900 AU from KIC 8462, is a possible influence in disrupting the system. The comet explanation would be striking if confirmed because we have no other instances of transiting events like these, and we would have found these comets by just happening to see them at the right time in their presumably long and eccentric orbit around the star.


Image: Left: a deep, isolated, asymmetric event in the Kepler data for KIC 8462. The deepest portion of the event is a couple of days long, but the long “tails” extend for over 10 days. Right: a complex series of events. The deepest event extends below 0.8, off the bottom of the figure. After Figure 1 of Boyajian et al. (2015). Credit: Wright et al.

So, despite PR headlines like Strange Star Likely Swarmed by Comets, I think we have to take a more cautious view. We’re dealing with a curious star whose changes in flux we don’t yet understand, and we have candidate theories to explain them. We’re no more ready to declare comets the cause of KIC 8462’s anomalies than we are to confirm alien megastructures. At this point we should leave both natural and artificial causes in the mix and recognize how long it’s going to take to work out a viable solution through careful, unbiased analysis.

The Marengo paper is Marengo, Hulsebus and Willis, ”KIC 8462852: The Infrared Flux,” Astrophysical Journal Letters, Vol. 814, No. 1 (abstract). I write about it this morning only because it is getting so much media attention — more later when I can go through the actual paper. The Boyajian paper is Boyajian et al., “Planet Hunters X. KIC 8462852 – Where’s the flux?” submitted to Monthly Notices of the Royal Astronomical Society (preprint). The Wright paper is Wright et al., “The Ĝ Search for Extraterrestrial Civilizations with Large Energy Supplies. IV. The Signatures and Information Content of Transiting Megastructures,” submitted to The Astrophysical Journal (preprint).


Comments on this entry are closed.

  • Isaac Kuo November 25, 2015, 11:20

    Could it be a combination of large comet comas to explain the scale of the dimming and gravity darkening to explain the asymmetry of the curves?

  • Harry R Ray November 25, 2015, 11:41

    I just read a excerpt of the excerpt on http://www.phys.org. This also apparrently rules out fragmentet COMETS in near circular orbits! I believe that by assuming that BOTH the 15% event AND the 22% event are the result of the SAME GROUP OF comets, that we can do some real science by analysing the different SHAPES of the two light curves! We may even be able to put CONSTRAINTS on the eccentricity of the orbits of the comet fragments! Logic dictates that if the orbits are TOO ECCENTRIC, that the great majority of the fragments would suffer the same fate as Comet ISON on their first passage near to the star. Conversely, a less excentric orbit would mean that apastron would be closer to the star than the star’s snow line. Logic also dictates a very RAPID evolution of the comet fragment configuration, indicating a VERY RECENT CAPTURE by the star(I RE-ITERATE that a dedicated search of old photographic plates of that region of the sky is a VERY HIGH PRIORITY), something akin to the capture of Shoemaker-Levy 9. The odds of seeing these comet fragments in just the right orbital configuration AND at just the right time is ASTRONOMICAL, so, at present, I do NOT buy into the theory that this is the BEST solution, but that it is just one of many PLAUSIBLE ones.

  • Black SciFi November 25, 2015, 11:48

    Let’s hope the principal investigators get enough funding to make a scientific determination.

  • Harry R Ray November 25, 2015, 12:00

    BACK AGAIN: I just read on http://www.universetoday.com that the authors FAVOR two SEPARATE swarms, one trailing toe other by approximately two years. To make THEIR scenario work, the lead comet of the first swarm would have had to be GARGANTUAN! This seems less likely to me than my above theory, although it is plausible. I cannot WAIT to read Dr Wright’s take on this on this website VERY SOON!

  • Jim Strom November 25, 2015, 14:29

    Black SciFi:

    To get funding, they need only title their proposal: “Request for funding to show that extra-solar comet swarms, not combustion of carbonaceous fuels, are causing global warming”.

  • Gerry November 25, 2015, 14:36

    Even allowing for the limited information available, it would be interesting to run some simulations to see what kinds matches can be found with regard to shapes and diameters, orbital speed and eccentricity of objects.

    There are some images in the papers that appear to have been created along these lines, portraying eclipsing objects of various sizes and trajectories against possible variations in the surface brightness of the star (I think).

  • ljk November 25, 2015, 16:32

    Well, even if there isn’t anyone building a Dyson Swarm around KIC 8675309, there is still a good chance somebody is doing something advanced somewhere in the galaxy or Universe:


  • Black SciFi November 25, 2015, 17:09

    Does the name “MorninLightMountain” ring a bell..??

    • SamChancer January 26, 2016, 15:59

      Yes, yes it does. ;-)

  • Jim Strom November 25, 2015, 17:24


    It wouldn’t be hard to craft a simple optimization algorithm, with:
    -3D model of:
    – Tabby’s star
    -observation point (Earth)
    -the time and shape of the the transit profiles
    -a candidate object permuter that permutes:
    -object shape
    -object orbit
    -object mass/type
    -a simulator that simulates gravity and motion to model transits for each permutation, decides if the permutation has moved closer to the constraints, if not, downweights the current permutation, if closer, upweights the current permutation.

    If I understood how to map between Kepler signal data and the 3d model, it wouldn’t be too difficult. There are probably more facts that would further constrain the solution, such as some of the non-Kepler observational data, etc. A really simple algorithm would only look for 3d shapes that are consistent with the transit profile. Anyone have a supercomputer handy?

    Also, anyone aware of a paper that discusses the resolution limits of a Kepler signal, with respect to profile shape?

  • Ashley Baldwin November 25, 2015, 18:03

    Great post. Thanks Paul. All the elements of scientific process coming together. Variance from the norm , atypicality , hypothesis formation and hypothesis challenge and testing m before acceptance.
    I would love to think this was some indication of alien intelligence . However being a suspicion academic before accepting this I would want alternative explanations ruled out . So here’s my alternative hypothesis .

    In the original post on this subject there was mention of a review of all the Kepler photometric false positive “signals” . The commonest causes of these was found to be either eclipsing binaries or star spots, in all cases except for KIC8462. Given it was only “exception” it’s obviously import at to know why , especially given the large implications.

    This star is in the early phases of evolving off the main sequence and has a spectral class of F3III/IV . Mid way between the main sequence and Red Giant phase . A volatile time in a star’s life and a cause of variability as its hydrogen core is used up and fusion is driven outwards as a shell bloating the outer layers of the star which increases in luminosity as it does in size . Procyon is a nearby similar example . Unlike Procyon, however , KIC8462 rotates at 84 Km/ sec ( versus a sedate 3.16 km/sec for Procyon ) , just 0.88 of a day. Such a rapid rotation would be expected to give rise to powerful internal magnetic fields , magnetic fields which if “surfacing ” create big star spots . Rapid rotation rates also make spectroscopy difficult as it tends to push individual lines closer together on a spectrum , like contour layers on a steep mountain , and in doing so makes accurate interpretation much harder . This is particularly the case in some B spectrum stars whose speedy rotation throws off star stuff and shows up as an emission both in a spectrum and the star’s class BVe.

    Differential rotation in the different pulsating layers of red Giants ( especially as they evolve off the main spectrum ) has also been seen to lead to large magnetic fields , fields that in turn fluctuate along with the star and give rise to large star spots. We know from the post earlier in the week that late M dwarf stars can have potent magnetic fields , fields we know cause huge star spots and are frequently seen in M dwarfs as well as no doubt being a common cause of a false positive ,as with close binaries, in other photometry programmes and not just Kepler . So we can be confident that large ,erratic magnetic fields can be a “smoking gun” for Star spots as the final ‘common pathway’.

    Interferometry has actually shown huge spots on Betelgeuse and the new technique of Doppler tomography recently showed even larger star spots on another red giant, XX Triangulum . These occupied over a third of the star’s disk. Like KIC8462 this star is about 1500 light years away.

    So that’s two potential causes of star spots coming together in one star found to be one of Kepler’s false positives ,Kepler whose two most likely causes of false positives included star spots . Coincidental perhaps but I would want star spots ruled out more definitively before accepting they were not the cause of this “anomaly” . Occam’s razor , common things are commonest and all that but above all the principle of falsifiability , by which a hypothesis or theory is subjected to challenge and testing to assess its accuracy before acceptance .

    So, is it possible that the somewhat unusual coming together of these two factors ( stellar evolution and rapid rotation leading to erratic magnetic fields) has created optimal conditions for star spot formation , large star spot formation?
    That’s my alternative hypothesis . Test and challenge as you should . Pass the Doppler Tomograph ? Happy to be wrong .

  • Daniel Suggs November 25, 2015, 18:32

    Black SciFi, I hope we don’t disturb the Dyson Pair. MorningLightMountain and his motils are no one to mess with. We don’t yet have those quantum nova bombs ready.
    (ordering the Void trilogy for winter reading)

  • JoeG November 25, 2015, 21:04

    “Look at him, he’s heading for that large comet.”
    “That’s no comet… Its a space station.”

  • Dmitry Novoseltsev November 26, 2015, 5:21

    In this context, it is interesting figure 1 ( https://sites.psu.edu/astrowright/2015/10/15/kic-8462852wheres-the-flux/, or in original paper “Planet Hunters X. KIC 8462852 – Where’s the flux?”): to the right of the mark 1500, which is clearly marked by a local increase in brightness relative to the average level , which could be interpreted as a reflection of the solar sail .
    Because the nuclei of comets are quite dark, the comet hypothesis does not explain this phenomenon.
    All of the observed pattern, with certain assumptions, could represent the initial phase of construction of the object as Shkadov thruster.

  • Richard November 26, 2015, 7:24

    I’ve still got some serious considerations on this whole cometary explanation. I mean we’re talking about a drop in flux of 20 percent. Over interstellar distances! When talking about comets or comet-swarms we’re talking about quite some dust that needs to block the light. It obviously couldn’t be just a regular joe comet because than we’d see similar light curves all over the galaxy since comets aren’t that strange of a phenomenon to begin with. So we do need a *huge amount of cometary dust to block 20 percent of starlight. And this dust doesn’t disappear all of a sudden. Shouldn’t we then pickup infrared emissions of the left-behind dust cloud??

  • Scott Gordon November 26, 2015, 7:49

    Responding to Harry’s comment : the star is 1480 light years away. No optical telescope currently available could hope to see it as anything but a single pixel at that distance – and the fact that any comet is drastically dimmer than the star would wash out comet images even if the comet was enough from the star. It’s going to be light curves and spectra only.

  • EricSECT November 26, 2015, 7:52

    Need to start ruling things out. Evidence of AT may be the least convoluted explanation? Can someone please explain why we should NOT expect to detect excess IR from a swarm of comets? Because it is so recent? Not enough time for it to warm up and re-radiate in IR? So in say 5 years, we WILL (may?) see a strong IR pulse? Neither hypothesis (comet swarm or gravity darkening) seems to adequately explain the long duration of the transients (up to 15 days) or the depth of the transients (up to 22% darkening). The duration implies a slow, distant orbit (think Uranus) and the darkening implies an immense object that if solid, is several times the diameter of Jupiter. And if diffuse, even larger. If we are reading the data right! Not totally convinced that Tabby’s companion dwarf is gravitationally influencing the primary (and not just a foreground star) and thus the cause of the comet swarm. Also, seems to be a transit signature of a ringed planet at day 1540? Need to observe this star with everything we can for another 5 years.

  • Michael November 26, 2015, 13:10

    I still feel it is planet and/or moon around it in a very eccentric orbit boiling off which forms these huge features. I tried loading some server software, Xming, from the net that would allow me to use Jason’s gravity darkening program which would have allowed me to take into consideration gravity darkening but no luck and I don’t have enough knowledge on Linux so I will have to wait like everyone else.

  • RussG123 November 26, 2015, 14:03

    EricSECT: I asked Dr. Wright that question during a Reddit AMA session and the answer was that we should immediately be expecting some IR signature from a swarm of comets/comet fragments, especially if the “event” causing the fragments is “new” to the system. Either way, I’d like to read this new paper and what the specific explanations are that lead to their conclusion. Based on some of the explanations in the WTF paper, if I recall correctly, some of the scenarios evaluated by the Boyajian team data suggested a calculated orbital distance of 1.6 AU…certainly close enough to exhibit some kind of IR signature…? One would think for a F3 IV/V star? Also, deep down I’ve been very interested in that number myself, only because that distance would appear to correspond with something at least close to the habitable zone for that star (not implying anything other than…”that’s interesting”).

  • Sara November 26, 2015, 21:36

    Well, I reread the first article and then followed by a second read of this one, and the one thing that sticks out in the whole ‘alien technology’ idea is that is no apparent transmission signal of any kind. You’d expect that, wouldn’t you? Stray radio emissions leave Earth all the time, and that includes broadcast TV signals from the pre-cable/Wi-Fi era, never mind early radio broadcasts starting with Marconi’s use of it, and possibly even the 19th century use of the telegraph.
    I haven’t seen anything in either of these articles that indicates an artificial signal of any kind.
    I would speculate, then, that the drop in the light signal is most likely caused by:
    A)- a massive planetary body with a huge satellite distribution;
    B)- a whole solar system that somehow got locked into the nearly the same orbital periods;
    C)- it WAS an alien civilization and something went terribly wrong; or
    D)- it’s the colonists on LB247, and they’ve run out of donuts and coffee, so they’re on strike.

  • Steven White November 27, 2015, 10:03

    “What seems to bother others, who have scoffed at the idea of extraterrestrial engineering, is that Wright and company have not explicitly ruled it out as a matter of course. The assumption there is that no other civilizations exist, and therefore we could not possibly be seeing one.”

    The HEIGHT of hubris, an inconceivable arrogance in anyone calling themselves a “scientist”. Yet, I believe in the end, they may well carry the day for just the reason I gave before: The last bastion of human uniquness, of human superiority, is what they will NOT surrender…our EXISTENCE. In the minds of many we must be alone, or we will be drastically lessened if we’re not.

    And it may well blind us…

  • Harry R Ray November 27, 2015, 10:40

    Dmitry: Interesting idea regarding Shkadov Thruster CONSTRUCTION(in past comments, I FAVORED a Shkadov Thruster REMINANT which was ORIGINALLY constructed around the M dwarf companion to move it into orbit around KIC8462852. However, if the M dwarf is NOT in orbit around KIC8462852, but; instead, poses some kind of DANGER to it as it makes a close pass, your theory makes a lot of sense. We’ll have to wait for some GAIA data on the two stars(assuming that the M dwarf has a magnitude greater than 20) to resolve this. NOW: A question for Dr Wright(or ANY reader who might know the answer): In the abstract, the authors stated “we found no evidence of infrared excess at 3.6 micron and small excess of 0.43 +/-0.18 mJy at 4.5 micron…”. Does this mean that there is no GAS around the star(comets have BOTH dust tails AND gas tails)? Also, should follow-up observations with JWST resolve the POTENTIAL 4.5 micron emmission to 3 sigma or above, what would THAT mean? Any megastructure that we can concieve(i.e. NOT “Black Monoliths”, which we can immagine, but not concieve of how they REALLY work) of MUST emit SOME KIND of waste heat, bot maybe at JUST ONE WAVELENGTH instead of several, which NATURAL objects like comets would.

  • Harry R Ray November 27, 2015, 12:17

    Dmitry: Did you happen to notice the only OTHER signifigant INCREASE in brightness than your post day 1500 event. This happened just AFTER the completion of quarter 3(day 280-300) and, to my unaided eye, it looks EXACTLY like your “spike”. The only difference is that it occurred amidst a period of quiescence(i.e a very “flat” baseline, as opposed to the chaotic episode your “spike appears in. What is your take on this. RSVP!

  • Michael November 27, 2015, 14:21

    If you look at the symmetrical feature in the middle of the distorted ones you can see it has the telltale diming of a circular object with the light dimming effect of a very wide object but optically variable. No planet gets that size unless very hot but if you look at Saturn which has a large ring system from the pole you can see just how big the light dimming effect can be.


    I see in my mind a planet with a large dense ring system, produced by an evaporating moon, thinning all the way to the cloud tops that is orientated more or less pole on towards us. Now what is causing the no IR excess, SO2 is relatively reflective in the near UV part of the spectrum but there would have to be a lot of it. Is it an Io type moon in total meltdown? if this moon is orbiting the planet in an eccentric orbit tidal effects would be very strong as the planet and Star flex it.

  • andy November 27, 2015, 21:18

    The paper was posted on the arXiv today. The eccentric orbit from the comet hypothesis would allow the bulk of the debris to move out to distances from the star where the infrared excess would be at longer wavelengths than the observations so far have been probing: remember the infrared observations are being made several years after the optical events.

    In any case, this shows the importance of getting multi-wavelength coverage of these kind of transient events.

  • S.Sadasivan November 27, 2015, 22:52

    The paper KIC 8462852: THE INFRARED FLUX by Massimo Marengo is in Arxiv Pl see link : http://arxiv.org/abs/1511.07908

    The comet swarm conclusion is what the Boyajian paper arrived at in despair !
    Thus the Spitzer observations are not that ground (comet ? ) breaking towards explaining the bizarre light curve !
    The conclusion about comet fragments to blame for the famous dip in the light curve of KIC 8462852 is at best still tenuous.

  • Eniac November 29, 2015, 1:22

    I think of a comet tail as a cloud of smoke. When a comet is right in front of the star, its tail will stream directly towards us. That puts it in the optimal position to occlude the star, using a minimum of material, and thus a minimum of IR emission. If there are a number of comets irregularly distributed around an eccentric orbit (i.e fragments from the break-up of a larger object) , you’d see random occlusions just like the ones that are observed. The total amount of material would still be low, and there would be far less IR emitted than expected from a full disk or ring. There would be a periodicity as the same comets come back around, but it could be very long. Best to keep looking, for that.

    I have to say this explanation makes a lot of sense. No need to evoke ET, here, in my opinion.

  • Michael November 29, 2015, 11:53

    @Eniac November 29, 2015 at 1:22

    ‘I think of a comet tail as a cloud of smoke. When a comet is right in front of the star, its tail will stream directly towards us…’

    This is not always the case, this Star rotates fast and most likely has a powerful magnetic field thus would tend to smear out the tail making it much larger with its stellar wind, possibly creating this distorted light curve. The greatest issue is the length of time it takes too move across the star, 20 -60 days! If the object was in a very eccentric orbit getting close to the star evaporating and with the Aphelion towards us it would look to move quite slowly across the disc.

    If it is an ET civilisation they are bad time keepers and messy, perhaps they contracted British rail to build their Dyson sphere!

  • Eniac November 29, 2015, 13:06


    It would take a very large comet indeed to account for the drop in flux we’ve already seen, but a swarm of comets and fragments can’t be ruled out because we just don’t have enough data to make the call.

    I am curious what prompts you to say it would take a large comet to achieve the observed drop in flux. Has anyone actually attempted to quantify the amount of starlight an active comet, seen tail-on, could block?

    Also, looking at the data, I would say it has to be a swarm, as we can clearly see the individual components passing through. Why do you say we do not have enough data for that call?

  • Paul Gilster November 29, 2015, 17:02

    Eniac, I am not aware that the amount of starlight that would be blocked by an active comet has been quantified in this way. I may well be wrong about this, and maybe one of the readers knows of a reference. But that’s one reason I think we don’t have enough data — or enough analysis — to make the call.

  • kzb November 30, 2015, 8:41

    I’m sure comets must’ve transited the sun many times in recorded history?
    I don’t recall any claims that the flux from the sun was noticeably reduced.

    Volcanic eruptions can certainly be seen in the climate records, but I don’t recall anyone pointing out the affect of a comet on such records. Maybe this should be looked at?

  • Eniac November 30, 2015, 22:17

    kzb: Good point. Very interesting, too!

    Good comets do not come often, and the chance that they would transit between the sun and the Earth seems really small. So, in my estimation, it is quite plausible that this has not happened in recorded history, but would look spectacular if it did happen.