≡ Menu

KIC 8462852: Enter ‘Gravity Darkening’

Back from my break, I have to explain to those who asked about what exotic destination I was headed for that I didn’t actually go anywhere (the South Pacific will have to wait). The break was from writing Centauri Dreams posts in order to concentrate on some other pressing matters that I had neglected for too long. Happily, I managed to get most of these taken care of, all the while keeping an eye on interstellar news and especially the interesting case of KIC 8462852 (for those just joining us, start with KIC 8462852: Cometary Origin of an Unusual Light Curve? and track the story through the next two entries).

Whatever the explanation for what can only be described as a bizarre light curve from this star, KIC 8462852 is a significant object. While Dysonian SETI has been percolating along, ably studied by projects like Glimpsing Heat from Alien Technologies, the public has continued to see SETI largely in terms of radio and deliberate attempts to communicate. Tabetha Boyajian and team, who produced the first paper on KIC 8462852, have put an end to that, ensuring wide coverage of the object as well as the notion that detection of an extraterrestrial intelligence might occur through observing large artificial structures in our astronomical data.

Meanwhile, the delightful ‘Tabby’s Star’ is beginning to emerge as a replacement for the star’s unwieldy designation. Coverage of the story has been all over the map. The term ‘alien megastructures’ has appeared in various headlines, while others have focused on the natural explanations that could mimic the ETI effect. The tension between natural and artificial is going to persist, and it’s the subject of Jason Wright and colleagues in their recent paper (submitted to The Astrophysical Journal), which asks that kinds of signatures an alien civilization’s activities could create, and what natural phenomenon could explain such signatures.

I think the Wright paper hits exactly the right note in its conclusion:

Invoking alien engineering to explain an anomalous astronomical phenomenon can be a perilous approach to science because it can lead to an “aliens of the gaps” fallacy (as discussed in §2.3 of Wright et al. 2014b) and unfalsifiable hypotheses. The conservative approach is therefore to initially ascribe all anomalies to natural sources, and only entertain the ETI hypothesis in cases where even the most contrived natural explanations fail to adequately explain the data. Nonetheless, invoking the ETI hypothesis can be a perfectly reasonable way to enrich the search space of communication SETI efforts with extraordinary targets, even while natural explanations are pursued.

Just so, and the lengthy discussions in the comments section here on the previous three articles on KIC 8462852 are much in that spirit. We do have the cometary hypothesis suggested in the original Boyajian paper as what had been considered the leading candidate, and Michael Million, a regular in these pages, has pointed to a paper from Jason Barnes (University of Idaho) and colleagues that looks at the phenomenon of gravity darkening and spin-orbit misalignment.

In this scenario, we have a star that is spinning fast enough to become oblate; i.e., it has a larger radius at the equator than it does at the poles, producing higher temperatures and ‘brightening’ at the poles, while the equator is consequently darkened. The transits of a planet in this scenario can produce asymmetrical light curves, a process the Wright paper notes, and one that Million began to discuss as early as the 17th in the comments here. That discussion was picked up in Did the Kepler space telescope discover alien megastructures? The mystery of Tabby’s star solved, which appeared in a blog called Desdemona Despair. The author sees the case as clear-cut: “There are four discrete events in the Kepler data for KIC 8462852, and planetary transits across a gravity-darkened disk are plausible causes for all of them.”

Screenshot from 2015-10-26 09:29:52

Image: Effects of rapid rotation on the shape of stars. Credit: Ming Zhao (Penn State).

Meanwhile, Centauri Dreams reader Jim Galasyn uncovered a paper by a team led by Shoya Kamiaka (University of Tokyo) studying gravity darkening of the light curves for the transiting system PTFO 8-8695, also studied by Barnes, which involves a ‘hot Jupiter’ orbiting a rapidly rotating pre-main-sequence star. Gravity darkening appears to be very much in play, and we can, as the Desdemona Despair blog does, cite the Barnes paper: “An oblique transit path across a gravity-darkened, oblate star leads to the long transit duration and asymmetric lightcurve evident in the photometric data [for the PTFO 8-8695 system].”

In Wright et al.’s “Signatures and Information Content of Transiting Megastructures” paper, which looks in depth at the natural sources of unusual light curves, these possibilities are discussed in relation to non-spherical stars, and this is worth quoting:

The dominant effect of a non-disk-like stellar aspect on transit light curves is to potentially generate an anomalous transit duration; the effects on ingress and egress shape are small. Gravity darkening, which makes the lower-gravity portions of the stellar disk dimmer than the other parts, can have a large effect on the transit curves of planets and stars with large spin-orbit misalignment, potentially producing transit light curves with large asymmetries and other in-transit features (first seen in the KOI-13 system, Barnes 2009; Barnes et al. 2011).

Another effect of a non-spherical star is to induce precession in an eccentric orbit. Wright also takes note of PTFO 8-8695, “which exhibits asymmetric transits of variable depth, variable duration, and variable in-transit shape.” Here astronomers were helped by the star’s age, which was soundly established by its association with the Orion star forming region. Wright adds that effects of this magnitude would not be expected for older, more slowly rotating objects.

The work on KIC 8462852 continues, and I also need to mention that the Allen Telescope Array focused in on this fascinating target beginning on October 16, even as the American Association of Variable Star Observers (AAVSO) published an Alert Notice requesting that astronomers begin observing the system. For more on this, see SETI Institute Undertakes Search for Alien Signal from Kepler Star KIC 8462852. Universe Today quotes the SETI Institute’s Gerald Harp as saying: “This is a special target. We’re using the scope to look at transmissions that would produce excess power over a range of wavelengths.” I’ll obviously be reporting on the paper that comes out of the ATA search.

The papers discussed today are 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); Barnes et al., “Measurement of Spin-Orbit Misalignment and Nodal Precession for the Planet around Pre-Main-Sequence Star PTFO 8-8695 From Gravity Darkening,” accepted at The Astrophysical Journal (preprint) and Kamiaka et al., “Revisiting a gravity-darkened and precessing planetary system PTFO 8-8695: spin-orbit non-synchronous case,” accepted at Publications of the Astronomical Society of Japan (preprint).


Comments on this entry are closed.

  • Jason Wright October 26, 2015, 10:08

    Gravity darkening might be a small part of the puzzle, but it does not explain the features of this star. Tabby’s star does not rotate fast enough to experience significant gravity darkening. That post also suggests that planets could be responsible, but planets are not large enough to produce the observed events, and there are too many events to explain with planets or stars.

  • Michael October 26, 2015, 11:00

    If this is a planet and it is in a near polar orbit getting close to the Star it may be possible to pick up radio waves from the interaction of the magnetic fields. I would love to know how it got into that orbit in the first place but after seeing some other planets behaving around their stars maybe it would not surprise me.

  • Terry Moseley October 26, 2015, 11:00

    The whole focus of this approach is wrong! Quote – “…. natural explanations that could mimic the ETI effect.” There IS NO “ETI effect”! To say that these odd light curves are an “ETI effect” is a highly dubious assumption in the first place. What grounds are there for ascribing such randomly variable changes in observed luminosity to ETI?
    Just because we can’t give an astrophysical explanation at present does not mean that we call it an “ETI effect”. It present it’s just an anomalous light curve, and while the article does say that an ETI explanation should not be a first choice; it should not then describe it as an “ETI effect”.

  • Alex Tolley October 26, 2015, 11:32

    Whatever the nature of the star, and the orbit the object[s] is making, perhaps more important for resolving this signal is what the object;s] is. One object is huge in relation to the star and any conceivable planet. So the explanation must contend with explaining its size and the apparent lack of IR signal. Is a comet or comets still the leading contender for the object? If so, is it sufficiently rare to account for a single sighting in all the Kepler data to date?

  • Jim Strom October 26, 2015, 11:37

    The fast-spinning star intuitively seems more plausible than a comet swarm. What conditions lead to fast-spin stars? Is it simply rapidly rotating protoplanetary disks? Aren’t fast spinning stars less likely to have significant planetry? Supposedly, accretion disks and planets slow down stars . Didn’t Spitzer data suggest something about spinning and planets, such as fast spinning stars are much less likely to have planets?

    If the fast-spinning star assumption is consistent with the Tabby signal, then the assumption should be applied to all aspects of the star, including its formation and likelihood of planets and their features such as size and orbit.

    Dos anyone know if star spots have been identified in Kepler data, and if so, has that been used to correlate spin rate with planet accompaniment?

  • Leo Serrano October 26, 2015, 11:56

    is there an explanation why the original authors of the paper did not mention the gravity darkening as a possible explanation? I assume they are aware of the literature on this matter. Why the comet and near-by star explanation ? Would gravity darkening have periodicity in its patter, if a planet is responsible for the signal observed?

  • Paul Gilster October 26, 2015, 12:13

    Terry Moseley writes:

    The whole focus of this approach is wrong! Quote – “…. natural explanations that could mimic the ETI effect.” There IS NO “ETI effect”! To say that these odd light curves are an “ETI effect” is a highly dubious assumption in the first place. What grounds are there for ascribing such randomly variable changes in observed luminosity to ETI?

    Terry, I think you are tripping over my clumsy phrasing. The quote means nothing more than “natural explanations that could mimic the appearance of an ETI signature.” The article does not claim we are seeing ETI. As the Wright paper stresses, we need to have an idea what an artificial construct might look like in our data, and we also need to screen for all the possible natural causes of the same signature.

  • Jason Wright October 26, 2015, 12:49

    Leo Serrano wrote:
    “is there an explanation why the original authors of the paper did not mention the gravity darkening as a possible explanation? I assume they are aware of the literature on this matter. ”

    Indeed, the lead author Tabetha Boyajian, is an experienced user of the instrument that made the images of rapid rotators in the figure, and the person who designed the figure, Ming Zhao, is a co-author of my paper about the ETI hypothesis (which, as Paul notes, actually does discuss gravity darkening, but not for Tabby’s star).

    The hypothesis was rejected because Tabby’s star does not rotate fast enough to have severe gravity darkening, and the gravity darkening hypothesis does not explain any of the most puzzling aspects of this star’s light curve. See my post above.

  • Björn Larsson October 26, 2015, 13:08

    I have already pointed out here, and in other places, that this obviously is an instrument failure. The anomalies occur only immediately after the telescope has been oriented in a certain identical way.

    Why are these sci fi fantasies continuing although the problem has been pointed out? Is it wishful thinking? Why didn’t the two papers published on the subject, one by Jason Wright here, mention the periodic correlation with the telescope’s orbit around the Sun? I would think that periodicity is the first thing to be looked at when it comes to a planet hunting telescope.

  • Michael October 26, 2015, 13:36

    Here is a light curve for a planet comet, similar to the first very large dip.


    If the eccentricity is high enough and the Aphelion towards us the planet will appear to undergo a longer transit as at aphelion it is travelling at its slowest. Add a cloud around it not much bigger than Saturn’s rings and you have a long transit and a deep too. I am wondering if the planet has trapped some of the ionised gas as it moved close to the star.

    Day 792 is close to day 1540 implying a ~25 day period.


    Tabby’s Firefly

  • andy October 26, 2015, 13:49

    Regarding PTFO 8-8695, apparently it doesn’t exhibit the expected periodicity, there is no signature of the expected emission from an orbiting planet, and there is no detection of the Rossiter-McLaughlin effect during the “transits”, so the orbiting planet hypothesis doesn’t seem to do a good job of explaining the observed behaviour of the system.

    Yu et al. (2015) “Tests of the planetary hypothesis for PTFO 8-8695b

  • Terry Moseley October 26, 2015, 13:50

    It would only be “huge in relation to the star and any conceivable planet” if it’s something close to the star, which has yet to be established. It could equally be some interstellar obscuring material which happens to lie approximately in the line of sight to the star.

  • Harry R Ray October 26, 2015, 14:31

    Jason Wright: Gerald Hart of the SETI institute apparantly discards the “comet” hypothesis entirely. He is discussing a signal that is “periodic” in nature(did he really mean to say “quasi-periodic”?). He contends that as a result, this is either almost certainly a recent planet break-up, with the only other plausable solution being alien intelligence. What I am NOT sure of is whether the “signal” he mentioned(which he reffered to as being produced by very large CLUMPS of material) is the signal detected by Kepler, or by the ATA(almost certainly the former, because any information LEAKED vis a vis the RAW DATA derived by last weeks observations may prove to be spurious when the ENTIRE data set id reduced). Have you heard anything YOURSELF regarding this?

  • Jim Strom October 26, 2015, 14:49

    If I’m clogging up this low-SNR blog with too much layman posting, someone please say so. Until then, here is another question. Does the Tabby signal indicate that there are no traditional planets present? Is there any way that scattered remnants of a demolished planet (in a mature and settled system with no accretion disk) could exist and transit without there also being a large IR signal?

    If the signal indicates a destroyed planet, could that extrapolate to a general galactic rate of planet destruction?

  • CharlesJQuarra October 26, 2015, 17:08

    If you had a standard gas giant with 2% – 3% area occlusion, in order for 2% of the area to affect the overall luminosity by 15% and 20% demands that the star luminosity be mostly confined within the narrow equatorial rim that is precisely on top of the planet trajectory. But obviously there is no planet since there are no regular orbits

    If it is the star spinning off the principal axis of inertia things can get nonlinear and irregular, but if that is true, eventually the dark side should face us eventually

  • TLDR October 26, 2015, 19:28

    You might have a point there. If it was a recently demolished planet, then the pieces wouldn’t be as round as expected, and would orbit in clumps of debris.

  • TLDR October 26, 2015, 19:30

    This one’s just for fun.
    If a star’s fleet were to be advancing on our pretty planet, certainly their warships would tend to obscure the light of their home planet, wouldn’t they?

  • Erik Landahl October 26, 2015, 19:35

    Layman here. Read the paper by Jason Wright et al. 4 times. Regarding planets, rings, dust, ionized gas, gravity darkening . . . per the Wright paper, there is strong evidence against such explanations for the KIC 8462852 light curve, is there not? And @ Bjorn Larsson – where is the evidence for instrument failure as an explanation?

    I agree with Paul – the Wright paper sounds “just the right note.” Balanced, calm, no definitive explanations, natural options kept open.

    What has been very interesting from a social observation is that in so many of the posts on this topic over the past week, many people have confidently and assuredly mentioned planets, rings, dust, ionized gas and gravity darkening as probable explanations for the KIC 8462852 light curve – despite the evidence to the contrary given in Wright’s paper. Even an unsubstantiated claim of instrument failure has been invoked in the posts here.

    This reaction by many posters – bending natural explanations for KIC 8462852 in ways a gymnast would envy, or simply ignoring the evidence and making conclusory assumptions – is a fascinating mirror to the human condition. There seems to be a strong, instinctive need for many – most? – humans to avoid an ETI explanation. Why?

  • David S. F. Portree October 26, 2015, 20:52

    Welcome back, Paul. You’re doing a great job of interpreting these papers for interested liberal arts (history) types like me. Thanks.

  • Terry Moseley October 26, 2015, 20:53

    Only if there was no relative motion between their star and the Earth (EXTREMELY unlikely), and if they were travelling in a dead straight line from that star to us! Imagine a car going past you while you stand in a field at the side of the road, and someone throws an apple out the car window at you. As you watch the apple approach you, the car will be moving off to the side, and very soon the apple will not appear to have come from the car at all.

  • Terry Moseley October 26, 2015, 20:55

    Why? Look up Occam’s Razor.

  • Milt October 26, 2015, 21:35

    @ Bjorn Larsson – that’s actually a good point, each Kepler quarter lasted 93 days so 8 quarters – i.e. 2 ‘years’ – would be 744 days, very close to the predicted period. But surely an instrument problem would have been manifested on more than one star.

  • Björn Larsson October 26, 2015, 21:46

    @Erik Landahl
    The evidence is in the raw data. The anomalies only occur the first days after the telescope is oriented the same way. Neither James Wright, nor “Tabby” mention this in their papers. They fail to observe this very strict periodicity in the Kepler data. I suppose it is too simple for their fantasy.

    The sensationalist astrophysical explanation with a gigantic cloud of comets, or a Dyson sphere, is unfounded and obviously impossible. It is just an instrument failure. It is so obvious that even I can see it.

  • Björn Larsson October 26, 2015, 21:53

    “Day 792 is close to day 1540 implying a ~25 day period.”
    No, 792 is not close to 25. It is close to two orbits of the telescope around the Sun. THe anomaly is correlated with the orbital period of the telescope. And looking more closely at it, the anomaly only occurs when the telescope is oriented the same way. And it escalates at the end when that reaction wheel fails. It is an instrument failure.

  • craig October 26, 2015, 23:53

    I’m just excited that we have found something so mysterious. Pulsars were an amazing discovery. Maybe they didn’t turn out to be as cool as alien signals, but they have been amazing for science nonetheless. I’m hopeful something pretty amazing will come out of this and I’m enjoying the speculation for its own sake.

    I agree with the paper cited here. This isn’t an either/or moment and there is no need to brush away alien hypotheses or to cling to them as probable causes. It’s just a moment when we get to glimpse the mysteries surrounding a scientific discovery. We already have resources devoted to SETI and this is a perfect target for those resources. Meanwhile, the rest of the scientific community can continue looking for the natural explanation that they are used to looking for.

  • H. Floyd October 27, 2015, 2:18

    Smart media coverage is not coming fast enough for the lay crowd following this story, and the scientific quality of this thread in particular is extremely helpful. Pardon a more sociological sequitur. I’m feeling awe that we’re commenting on something here that could prove to be the single most significant discovery in the entire history of our species to date.

    In what seems like the decreasingly unlikely event that Kepler and its scientists have spotted a 1,480 year old artifact in another solar system, the item itself would not just have unprecedented importance. It would say something huge – huge – that in billions of years of cell division on Earth, homo sapiens were the ones who spotted it.

    Our heritage as a species -might- expand suddenly in the weeks and months ahead. It can’t hurt to prepare for that awareness, just in case.

  • Michael October 27, 2015, 3:28

    When I look at each quarter I see some dips at 140, 180, 215, 260, 360, 425, 500, 790, 1145, 1205, 1515, 1540, 1565, 1590 days.

    They are awfully close to 25 days, the difference between the first dip at 140 and 1590 is ~5 days (series shows 1585, that’s ~12.5 seconds per day). The other ‘no shows’ may be due to reduced size of the eruptive ‘planet/moon’ at the time or a slight precession.

    @Erik Landahl October 26, 2015 at 19:35

    ‘…This reaction by many posters – bending natural explanations for KIC 8462852 in ways a gymnast would envy, or simply ignoring the evidence and making conclusory assumptions – is a fascinating mirror to the human condition. There seems to be a strong, instinctive need for many – most? – humans to avoid an ETI explanation. Why?’

    Because the probability of an advanced Alien civilisation so close to us is very small compared to other natural explanations.

    If they are an Alien civilisation they look to be very messy, bad constructors and have poor time keeping. Also if the Star is middle aged it would be around 1.5 to 2 billions years old. Our genetic linage was only at the level of the scum around a bath tub water mark at that age, they must be fast learners.

  • OhYeah October 27, 2015, 5:55

    I just wanted to touch on the subject on instrument failure or malfunction. This was my initial reaction when I read about this object (“must be this!”), however, the people responsible for the equipment have repeatedly assured us that after multiple checks the data is correct and everything is working just fine.

    After such a length of time has passed I would trust them on this. So no, it’s probably not instrument failure.

  • Michael Spencer October 27, 2015, 6:57

    Björn Larsson asks

    “Why are these sci fi fantasies continuing although the problem has been pointed out? Is it wishful thinking? ”

    Lots more fun than dusty or flattened stars, that’s why! (and, a bit of wishful thinking).

  • Daniel Högberg October 27, 2015, 7:36

    Theres no “instinctive need” to avoid an ETI explanation. Most people wish and hope for it to be real, but we cant start confusing it with fact before we are 100% sure. Imagine huge disappointment if it turns out to be a natural after all, the universe would look even coldier and lonlier then it was before .

  • kzb October 27, 2015, 8:20

    Bjorn Larsson:
    Another possible reason for the correlation with the length of (our) year could be that the obscuring objects are local to the solar system? Perhaps we have a small gas cloud passing by or in distant orbit?

  • beermotor October 27, 2015, 9:19

    The papers all say they confirmed with Kepler that there was no problem with the data, which I understood to mean nothing wrong with the instrument. If Bjorn Larsson is right and it’s an instrument failure, you would expect such anomalous readings across more than, you know, _a single star in the field of 180k (or whatever it was)_.

    But what do I know.

  • Jim Galasyn October 27, 2015, 11:30

    Thank you for the great thread, Paul.

    @Andy, the Yu paper is very enlightening, thanks for posting. Although it looks like gravity darkening doesn’t work for PTFO 8-8695, it appears to work spectacularly well for KIC 8462852, which shows the light-curve morphology predicted in Kamiaka, et al. (2015). Even the small in-transit features predicted by Kamiaka appear in the KIC 8462852 data. Coincidence?

    I’ve posted a comparison here: http://1drv.ms/1N2wyTN.

  • Björn Larsson October 27, 2015, 11:41

    But Oort Cloud objects passing by far away objects are used (attempted anyway) for microlensing or refraction. That means that the far away object is magnified, not dimmed.

    The papers don’t mention the strict timing with how the telescope was turned. Several other potential issues were investigated, but not this, which actually is so obvious that even I notice it by glancing the charts and looking up the telescope spinning schedule (if I understand it correctly).

  • Michael October 27, 2015, 14:31

    A Great thread Paul, I think I can speak for everyone in that we have ALL had a full mental workout or breakdown for that matter!

    All I can see now when I go to sleep is this planet skimming a searing surface…maybe I need one more cup of coffee…

  • andy October 27, 2015, 18:53

    Regarding the timing of the anomalies with respect to the Kepler quarters, figure 1a in Boyajian et al. (2015) indicates the boundaries between Kepler quarters with red ticks. The first dip occurs near the end of quarter 8 (i.e. not in the first days of that quarter), while the second group occurs in quarters 16-17 (the final ~8% drop looks like it is in quarter 17).

  • kzb October 28, 2015, 8:48

    @Bjorn Larsson
    We are told that one of the key checks carried out by the Kepler team was to examine other stars imaged at the same time as KIC 8462852. These images exhibit no anomaly.

    Surely if it were an instrumental fault that occurs in the way you hypothesise, all star images taken at the same time would show similar anomalies to KIC 8462852?

  • Björn Larsson October 28, 2015, 9:27

    The idea is that the pixels on the CCD array, which were pointed to the star, were defect. Just a local part of the image would show anomalies. When the telescope was turned in the same way (it was always pointed in the same direction), then the same pixels would be at or close to the same star. So a defect pixel would cause an anomaly for that star’s light curve.

    As far as I can determine, the anomalies occurred immediately after the telescope had been oriented the same way (every other time actually, for some reason). The first anomaly (which is <1% compared to 15% and 22% of the later two) happened a couple of weeks before the first quarter. But since the telescope was quite new in space by then (about 40 days), maybe it was turned in the same way then anyway?

  • beermotor October 28, 2015, 9:44


    If that was the case, and I don’t think it was, you would be able to test that hypothesis by looking for defective pixel data in the K2 data set. But, to me, this sounds like quite a stretch. Defective pixels receive less light from the (single) star but don’t fail entirely? Any time I’ve ever seen a defect in camera hardware, it’s been a wide-spread decline to a total failure.

  • Björn Larsson October 28, 2015, 9:57

    I think that these dates of quarterly turning of the telescope match up perfectly with the three biggest anomalies. But I might misinterpret or mismatch something here. I just don’t see what.

    95,000,000 pixels and 160,000 stars, in Kepler’s original mission, means that most pixels are pointed to darkness. K2 might never point that dead pixel to any light source again. And the star in question seems to have occupied several pixels. If a fifth of them were dead then a 22% dimming would occur. Maybe the pixels fade and come alive again over time. Or maybe the dead pixels rotated in and out of the starlight. That might explain why the 15% anomaly is so clean, just in and out. And why the last anomaly is so messy, because the second reaction wheel was gradually degrading, pushing that or those bad pixels in and out of the starlight.

    I think it is much more likely than some humongous IR-protected comet cloud with half of the star’s diameter. I think that the mirror makers are able to simulate how well such a defect would match the data. A big problem for astrophysical explanations, I think, is that the “transits” don’t have flat bottoms.

  • andy October 28, 2015, 14:27

    Ok I’m not seeing this “perfect” match up.

    Going from the light curve files, the biggest dips are:
    BJD 2455625.7199054 (2011-Mar-05, Q8)
    BJD 2456352.5225639 (2013-Mar-01, Q16)
    BJD 2456401.4820219 (2013-Apr-18, Q17)

    The actual start and end dates of these quarters are given on the page I linked, day = BJD-2454833, so converting the relevant quarters:

    Q8: BJD 2455568.35345625 – 2455635.35445643 (2011-Jan-06 – 2011-Mar-14)
    Q16: BJD 2456305.08713018 – 2456390.96870717 (2013-Jan-12 – 2013-Apr-08)
    Q17: BJD 2456392.21513782 – 2456424.01084956 (2013-Apr-09 – 2013-May-11)

    Note that the Q8 dates are substantially different from those listed on the Archive Schedule. So of the three largest anomalies, the first occurs near the end of a quarter, the second occurs near the middle, and the third isn’t even on the same detector as the other two. Also Q12 appears to be unaffected, which seems odd if it were an instrumental effect.

    Eyeballing figure 1a, the other large anomalies are at the start of Q1, at the Q2/Q3 boundary, and in Q13. So while the two largest dips do indeed occur on the same detector, it does not look to be only that detector that is affected.

  • Björn Larsson October 28, 2015, 15:30

    I don’t have perfect documentation of how to fit dates to the number of days in the diagrams of the papers, so there might be a problem with matching it correctly. My dates and numbered days(! :-)) below are just my simple estimates based on googled knowledge.

    The two gigantic light dips seem to me to occur soon after 28 June 2011 (day 787) and 8 May 2013 (day 1467). Both times as the telescope was turning into the same quarter (with each time one in between without anomalies). Those were the 15% and 22% ones, huge. The first anomaly of less than 1% occurred 20 June 2009, two years earlier, just before the beginning of the same quarter as the two big anomalies. Maybe it isn’t part of the whole, but maybe the telescope was turned the same way then because it had gotten into station just a few weeks earlier?

    The second of the four largest anomalies (<1%) maybe doesn't fit this idea, but it might be any kind of disturbance, or even a real exoplanet. Nothing to do with a huge comet cloud or superstructure anyway.

  • Milt October 28, 2015, 16:02

    No, the Julian Day used by the control team equates to January 1 2009. The probe launched on May 12 2009 so ‘the real day 1’ is actually ‘day 132’ in mission-speak. You’ll notice the graphs etc. are blank for the first 130 days or so. Therefore your dates are 4 months out.

  • andy October 28, 2015, 18:24

    In fact the light curve files list both Kepler day number and BJD. So just to be explicit here, the three dips I mentioned in my previous post are:

    day 792.7199054 = BJD 2455625.7199054 (2011-Mar-05, Q8)
    day 1519.5225639 = BJD 2456352.5225639 (2013-Mar-01, Q16)
    day 1568.4820219 = BJD 2456401.4820219 (2013-Apr-18, Q17)

    Furthermore, as I have mentioned, figure 1a in the Boyajian et al. paper (which is freely available to download from the arXiv) actually shows where the quarter transitions occurred. It is immediately obvious from this graph that the dips do not occur on the quarter transitions. The graphs also include in the axis labels the correct conversion between day number and BJD: the unit is given as “Time (BJD-2454833)”.

  • Eric Hughes October 28, 2015, 19:04

    @Björn Larsson – Gravitational lensing distorts a light field, brightening at some points and darkening at others. If lensing only brightened you’d violate conservation of energy.

  • Jim Galasyn October 28, 2015, 19:30

    I emailed Jason Barnes for his thoughts on KIC 8462852, and here’s his reply:

    “Indeed, the individual transit events in the KIC 8462852 lightcurve definitely show the signature of a misaligned object transiting a gravity-darkened stellar disk. Even if PTFO 8-8695b looks like it probably isn’t a planet, we have seen this kind of gravity-darkening signature at Kepler-13b, for instance. But WHAT might be transiting at such odd cadences and with such differing durations is far less clear.

    “The Kepler-13 gravity-darkened transit signature paper can be found here: http://barnesos.net/publications/papers/2011.11.ApJS.Barnes.KOI13.SpinOrbit.pdf

  • Björn Larsson October 29, 2015, 1:09

    The two big anomalies occur in quarters 8 and 16. When the telescope was oriented the same way (according to figure 1a in the paper).

    @Eric Hughes
    Okay, but the light curve in this case never brightens.

  • Michael October 29, 2015, 3:43

    If there is a Planet that is behaving as a comet due to the close passes of the Star we should see increases in brightness, due to reflection, as the cloud approaches and leaves from behind the Star with a minimum in between two ‘hills’ when the cloud is behind the Star. The period should be around half for the minimum (behind the star). Now I see some of that in the plot but not all the time although it is hard to read sometimes, messy plomet!


    Do you have Jason’s e-mail to pass on, I am interested in the gravity darkening bit which would smear out the light curve?

  • Jim Galasyn October 29, 2015, 9:49

    @Michael, you can find Jason’s email here: http://www.uidaho.edu/sci/physics/people/faculty/jwbarnes

    Please let us know what you find out!