I hadn’t expected a new paper on KIC 8462852 quite this fast, but hard on the heels of yesterday’s article on the star comes “KIC 8462852 Faded at an Average Rate of 0.165±0.013 Magnitudes Per Century From 1890 To 1989,” from Bradley Schaefer (Louisiana State University). Schaefer takes a hard look at this F3 main sequence star in the original Kepler field not only via the Kepler data but by using a collection of roughly 500,000 sky photographs in the archives of Harvard College Observatory, covering the period from 1890 to 1989.
The Harvard collection is vast, but Schaefer could take advantage of a program called Digital Access to a Sky Century@Harvard (DASCH), which has currently digitized about 15 percent of the archives. Fortunately for us, this 15 percent covers all the plates containing the Cygnus/Lyra starfield, which is what the Kepler instrument focused on. Some 1581 of these plates cover the region of sky where KIC 8462852 is found. What Schaefer discovers is a secular dimming at an average rate of 0.165±0.013 magnitudes per century. For the period in question, ending in the late 1980s, KIC 8462852 has faded by 0.193±0.030 mag. From the paper:
The KIC 8462852 light curve from 1890 to 1989 shows a highly significant secular trend in fading over 100 years, with this being completely unprecedented for any F-type main sequence star. Such stars should be very stable in brightness, with evolution making for changes only on time scales of many millions of years. So the Harvard data alone prove that KIC 8462852 has unique and large-amplitude photometric variations.
That’s useful information, especially given the possible objection to the Kepler findings that they might be traceable to a problem with the Kepler spacecraft itself. Evidently not:
Previously, the only evidence that KIC 8462852 was unusual in any way was a few dips in magnitude as observed by one satellite, so inevitably we have to wonder whether the whole story is just some problem with Kepler. Boyajian et al. (2015) had already made a convincing case that the dips were not caused by any data or analysis artifacts, and their case is strong. Nevertheless, it is comforting to know from two independent sources that KIC 8462852 is displaying unique and inexplicable photometric variations.
As Schaefer notes, KIC 8462852 can now be seen to show two unique episodes involving dimming — the dips described here yesterday for the Kepler spacecraft, and the fading in the Harvard data. The assumption that both come from the same cause is reasonable, as it would be hard to see how the same star could experience two distinct mechanisms that make its starlight dim by amounts like these. The timescales of the dimming obviously vary, and the assumption would be that if the day-long dips are caused by circumstellar dust, then the much longer fading that Schaefer has detected would be caused by the same mechanism.
Image: KIC 8462852 as photographed from Aguadilla, Puerto Rico by Efraín Morales, of the Astronomical Society of the Caribbean (SAC).
Thus we come to the comet hypothesis as a way of explaining the KIC 8462852 light curves. Incorporating the fading Schaefer has discovered, a cometary solution would require some mind-boggling numbers, as derived in the paper. From the summary:
With 36 giant-comets required to make the one 20% Kepler dip, and all of these along one orbit, we would need 648,000 giant-comets to create the century-long fading. For these 200 km diameter giant-comets having a density of 1 gm cm?3, each will have a mass of 4 × 1021 gm, and the total will have a mass of 0.4 M?. This can be compared to the largest known comet in our own Solar System (Comet Hale-Bopp) with a diameter of 60 km. This can also be compared to the entire mass of the Kuiper Belt at around 0.1 M? (Gladman et al. 2001). I do not see how it is possible for something like 648,000 giant-comets to exist around one star, nor to have their orbits orchestrated so as to all pass in front of the star within the last century. So I take this century-long dimming as a strong argument against the comet-family hypothesis to explain the Kepler dips.
If Schaefer’s work holds up, the cometary hypothesis to explain KIC 8462852 is deeply compromised. We seem to be looking at the author calls “an ongoing process with continuous effects” around the star. Moreover, it is a process that requires 104 to 107 times as much dust as would be required for the deepest of the Kepler light dips. And you can see in the quotation above Schaefer’s estimate for the number of giant comets this would require, all of them having to pass in front of the star in the last century.
The paper is Schaefer, “KIC 8462852 Faded at an Average Rate of 0.165+-0.013 Magnitudes Per Century From 1890 To 1989,” submitted to Astrophysical Journal Letters (abstract).
Harry Ray writes:
“COMBINED with RUMERS of a HINT of “reddening” in the system(which, IF CONFIRMED”
any change you can elaborate on this?
Has anyone looked at the Red dwarf in the past and present photos to determine if it is coming or going and work out the velocity of it relative to Tabby’s star. I tried opening the plates but my computer takes ages to open even a little plate!
Anyone else out there that can do it?
Eniac January 19, 2016 at 12:09
Sorry, what was your explanation for the long term dimming?
@Eniac So the the missing IR signature is suddenly just out of the equation? I thought the lack of IR light was the reason the cometary scenario was abandoned?
As far as I know, the missing IR only eliminates ring scenarios, where an extended dusty disk sometimes occults the star. Comets that happen to cross between the star and our line of sight because their orbits are so aligned would not emit noticable IR, I would think.
That ‘reddening’ sounds interesting, Harry.
Martin Elvis of the Harvard / CFA sends word that “Harvard plate stacks suffered from flooding today. Some equipment damage. Details later.”
I hope that there was no loss of irrecoverable data.
Bill @ January 19, 2016 at 10:44
“Several people have expressed a desire for the Harvard raw data. It is in fact available through the DASCH Lightcurve access webpage:
http://dasch.rc.fas.harvard.edu/lightcurve.php ”
Thank you for sharing that information – now I will delve into the raw data and look for more detailed trends/artefacts.
Any chance you have the coordinates for the 2 comparison stars and do they appear on the same plates as KIC 8462852?
@Mark Zambelli January 19, 2016 at 18:25
‘That ‘reddening’ sounds interesting, Harry.’
Stars at greater distances from earth become increasingly reddened due to scattering of light by interstellar gas and dust, this star is around 1400 lyrs away so reddening is expected.
Thank you Jason Wright for clarification.
I misunderstood: What I assumed was a “gnats ass” change in Tabby’s output is actually 20% dimming since 1890! Appears to be good data. And the puzzling lack of infrared, which I at first considered evidence AGAINST ETI technology, may instead turn out to be indirect evidence SUPPORTING something “unnatural”. Wheres the flux? The missing flux could be (1) Directed, away from us or (2) Emitted at a very low temperature, in an IR window we can’t see or (3) Stored and/or converted into something: ie, work is being extracted (but not at 100% efficiency) Can anyone think of a “natural” way to do 1 through 3? How about combinations? Combine 1 & 2: Somehow, naturally, it is directed away from us at a window we can’t see?
Paul:
You could start a new, separate from CD blog, just about Tabby’s Star. Just look at all the comments here and in the last few months! So many that the thread of the individual conversations gets difficult to follow. Ever consider re-vamping to, say…. how FaceBook’s posts work? FB Posts can have what I call “separate replies” and/or an “overall comment.”
Schaefer: “Likely the best try is to get a spectrum of the star during a dip. From the spectrum, we might see absorption lines from any gas associated with the ‘occulter,’ we might see a reddening that would point to the occulter being mainly dust, or we might see a color neutral dip that would point to a solid body. Thus, a spectrum would tell us the nature of the occulter, and this would greatly narrow down models.”
http://motherboard.vice.com/read/that-alien-megastructures-star-continues-to-defy-explanation
@Rafik
“Any chance you have the coordinates for the 2 comparison stars and do they appear on the same plates as KIC 8462852?”
TYC 3162-1001-1 : 20 06 55.887 +44 26 43.374
TYC 3162-879-1 : cannot locate
Even if this is a swarm of comets orbiting that star – and not a young star at that – why are we not seeing more such events? And the astronomers already dismissed a planetary collision for this system, does that still apply?
Kepler scanned a very narrow range of the sky. Did we really just get lucky finding a rare celestial phenomenon, or does that mean there are a lot of them and we just haven’t had the means to detect such things yet? I have read more than once that the chances of us coming across an unusual/rare celestial event in the Universe’s 13.8 billion year history are more than slim, so where do we stand with Tabby’s Star if this is still true?
Do people get what I am trying to say here?
http://simbad.u-strasbg.fr/simbad/sim-id?Ident=TYC+3162-879-1&NbIdent=1&Radius=2&Radius.unit=arcmin&submit=submit+id
LJK: “Even if this is a swarm of comets orbiting that star – and not a young star at that – why are we not seeing more such events? And the astronomers already dismissed a planetary collision for this system, does that still apply?”
That’s why we need to look at red dwarf next to Tabby’s. Did it disturb Tabby’s Oort cloud? Is is bound, or passing by or just in the foreground (background)? Gotta start eliminating things!
I think the conclusion in the Schaefer paper that
“The KIC8462852 light curve from 1890 to 1989 shows a highly significant secular trend in fading over 100 years, ”
could be a bit premature and may not stand up to peer review.
Looking at the “raw data” there is clearly a period of stability in the magnitude of KIC8462852 from 1915 through to 1952.
There appears to be a very slight dimming trend during this period of 39 years from around 12.34 to 12.35, well within error limits and similar to changes observed in the 2 comparison stars.
Putting data into 5-year “bins”, weighting each “bin” equalling and including “outlier” observations, can have a distorting effect when looking for secular trends.
However, even with the 5-year bin method the result is a similar period of stability from 1915 to 1952 equivalent to dimming/brightening shown by the 2 comparison stars.
A similar result happens when looking at raw data (and excluding “outliers”) for the period 1962-89. A slight dimming is noticed over the 28 years from 12.45 to 12.46, again similar to the movements in magnitude noticed in the comparison stars.
The period prior to 1915 shows a significant dimming trend (at a much higher rate than in the Schaefer paper) but the data seems to be of poorer quality. There is another period of significant dimming in magnitude between 1952 and 1967 – during the “Menzel Gap”. It is worth noting that some data is available for 1961-62 but this data seems to be poor quality.
I do not consider that, in the long term data, we are observing a “secular” trend in dimming of KIC8462852. Rather, there appears to be short dynamic periods of significant variation in magnitude and other periods of relative stability, 67 years in total, when KIC8462852 is behaving like other stars.
I do agree with the notion in the Schaefer paper that whatever is causing the variation in magnitude in the short term observed by Kepler is likely to be responsible for the long term variations. These long term variations appear to occur in shorter periods of time followed by longer periods of stability in the magnitude of KIC8462852.
Corey Hart: Unfortunately, NO! It was just a tweet among MANY at JasonWwight@AstroWright/Twitter. ALL of the tweets BEFORE and AFTER this one were about KIC8462852, so I ASSUMED this one was,too. The KICKER here is that it WASN’T made by Jason himself! I forgot who made it. I wonder how many CCD images of KIC8462852 were made between 1990 and 2010, CONNECTING THE DOTS BETWEEN DASCH and Kepler? If this IS a LINEAR progression, as Schaefer SUGGESTS, a 20% dip between 1890 and 1989 would THEN mean a 4% dip between 1990 and 2010, the ACCURACY OF WHICH would be MUCH GREATER than that of the PLATES! Anybody have an idea of how many, and how they may be accessed?
@Bill
@Rafik
TYC 3162-879-1 20 06 08.977 +44 24 30.19
http://simbad.u-strasbg.fr/simbad/sim-id?Ident=TYC+3162-879-1
@ljk
Kepler looked at many thousands of F stars like Tabby’s Star. It’s not clear to me how many other F stars have been observed before at a level that would have made us notice similar events in them (20% dimming is huge and obvious, but there are only ~2 such events for Tabby’s star, and it’s not obvious such events would be obviously extraordinary at lower precision).
But yes, given that there are billions of F stars in the Milky Way, one must assume that there are millions of stars like it in the Galaxy; more if Tabby’s star is only obvious because of a special geometry.
A planetary collision should result in lots of IR emission, which we do not see.
Corey Hart: UPDATE! I tracked down the tweet. It was by John Debes, and said “there is reddining in System E(B-V)0.11. I think this means E(B-V)=0.11 magnatudes, but I am not SURE!
Paul, re@EricSECT comment on changing your blog format.
If you were to consider a change, you might look at the comment format on Judith Curry’s blog: judithcurry.com
It allows replies to comments and there may be several different discussions going on a particular article. Personally, I really enjoy Centauri Dreams, But replies might make it even better.
@Michael
“Stars at greater distances from earth become increasingly reddened due to scattering of light by interstellar gas and dust, this star is around 1400 lyrs away so reddening is expected.”
Quite correct. I’m aware of the ISM causing reddening as you mention, but I had assumed Harry was referring to an ‘unexpected’ reddening of some kind that tallys to the odd dimming.
Thanks, Harry, for digging out the tweet.
Daniel Suggs writes:
I’m not against the idea but I’m very wary of new plugins from a security standpoint. But have been considering changes like this for some time. Thanks.
OK, nested comments now enabled thanks to Judith Curry and Daniel Suggs’ pointers. Much appreciated, both of you!
I have had a closer look at the data from 1890-1915 and if we remove the “outliers” there is almost no change in magnitude for KIC8462852.
Over the total period 1890-1952 there is no “secular” change in magnitude. KIC8462852 shows a very slight dimming over this period of 63 years from 12.352 to 12.356.
Over the period 1967-89 there is also no significant movement in magnitude with just a slight dimming from 12.450 to 12.455.
There remains a noticeable variation between these two stable periods during the “Menzel Gap”. (The data for 1962 appears to be of poorer quality, but if we remove the bright “outlier” measurements we get an average of around 12.46 – but I have low confidence in this data point.)
Would like to hear anyone’s thoughts on why this may be the case.
My conclusion on examining the Harvard data is that KIC8462852 has remained stable from 1890 to 1952 then during the “Menzal Gap” experienced a dimming of around 0.1 in magnitude.
Since 1967 (and possibly back to 1962) KIC8462852 has again remained relatively stable in magnitude up to 1989.
I know this might not be the answer we were looking for but the data does not appear to support a secular trend in dimming over the past century.
My impression upon seeing the individual brightness observations plotted against time was similar. It looked to me like a flat curve, then a gap, then a lower flat curve. My first thought was a shift in calibration — perhaps a new method of processing the exposed plates that is not quite as good, that causes the star images to be slightly smaller. But since it appears that the surrounding stars did not change their recorded brightnesses to change, that could not be it. My current guess is a relatively sudden drop of about 0.15 in blue magnitude between 1953 and 1962, which is unfortunately not covered by the Harvard plates. We really need to see the Sonnenberg plates, which apparently continued without interruption through the 1950s and 1960s.
Rather than outlier rejection, which I implicitly distrust I may work through these data using a median filter and several binnings. The differences across time in the raw data are too significant for me to believe they are all outlier induced artifact. My only question is whether the clear drops are continuous or discontinuous.
That is going to depend very strongly on what criterion you use for outliers, is it not?
What criterion are you using?
I removed values 12 and below and 12.75 and above from the data – that is roughly the top and bottom 3%.
I also noticed I missed some datapoints in the 1967-89 series. The increase in dimming during this period may actually be larger than what I had indicated earlier moving from around 12.43 to 12.46.
I need to go back and revisit that period to make sure I have covered everything correctly.
You should post your results, in a graph attachment.
I have a word document with The Schaefer Graph, and two graphs I have prepared one for KIC8462852 and one for TYC 3162-879-1.
Happy to share with anyone who is interested.
Not sure what is the best way of putting such a document on-line for people to look at – bit of a novice in that area.
I still think we are looking at a reasonably stable star (over the long term) that has undergone a dimming event during the “Menzel Gap”.
There may be a small long term dimming trend. However, additional data would be required before we could be confident about any such secular trend in dimming.
If an “event” occurred in the 1950’s or early 1960’s would there be any IR measurement of such an event?
For the record here is Jason Wright’s reply to my question on this:
https://twitter.com/astro_wright/status/690185301487132672
I will look at this through a median filter maybe tonight. Overwhelmed with analytics I am actually getting paid to do right now…
Looking at Figure 1 in the paper and projecting out to present it looks like the value for “B” should be greater than 12.5. But that does not seem to be in agreement with the database value.
What “database value” are you referring to?
Yes a simple extension of the “secular” trend in dimming reported by Schaefer would result in an expected current magnitude measurement of around 12.49. As outlined in my post above I do not think the data supports a ‘secular” dimming trend. The data supports two long periods of stability in magnitude separated by a significant dimming that occurred during the “Menzel Gap”.
I would be reasonably confident that either the magnitude of KIC846285 has remained around 12.45-12.46 or has returned to around 12.35.
However, these magnitude measurements are based on photographic plates so we need to be able to either take current photographs with similar plates or have a database using current techniques that “overlaps” with the old photographic plate record.
http://simbad.u-strasbg.fr/simbad/sim-basic?Ident=KIC+8462852&submit=SIMBAD+search
B = 12.82 based on Hipparcos data circa 1989 to 1993
seems much greater than the trends indicated in figure 1
Tang et al. (2013)* examined all of the Kepler planet candidate stars for data in the DASCH plates. 109 of these have > 100 good measurements, of these “No variation is found at 3? level for these host stars.”
To me, this indicates that the Harvard plates do provide a stable century long photometric reference. They are likely to be not the only one – we are interested in an 12th magnitude star at +44 deg declination. There will be a lot of photometric data on this object, especially for the post WWII period.
* http://authors.library.caltech.edu/41070/7/671759.pdf
Mr. Eubanks: I just heard Paul Carr mention you on his podcast in reference to KIC8462852 submillimeter absorption data. Can you link me to more info on this?
I should mention that this seemed to be a reference to an anomaly beyond infrared, that could be consistent with waste heat. I haven’t seen a reference to this anywhere else, so I’m very curious.
Rafik writes:
Anyone who would like to see Rafik’s documents can write me back-channel and I’ll be glad to connect you with Rafik for the download. The ‘Contact’ tab on the main page has the email to use to reach me.
I have shared the graphs with Paul – so you can contact him if interested.
In quick summary;
there appears a small dimming of .021 in magnitude between 1894 and 1952,
followed by a large dimming event of .075 between 1952 and 1962,
and then a slight dimming of .009 between 1962 and 1989.
Would love to hear anyone’s thoughts on what could cause this.
A brown dwarf or supergiant Earth slingshotted into Tabby’s Star.
The ejecta would show the exact same Doppler signature as Tabby’s for a while, and well until it cools off. The mess would look rather irregular, and it could cause significant dimming, but nothing special should be detected on the IR side until it totally cools off.
I don’t know the region where Tabby’s Star lives well enough, it’s not near any clusters, which could make this explanation less likely.
But every now and then planets and stars leave their normal orbits, so the chance that this could happen, however unlikely, is non-zero.
The good thing is that this conjecture can be proven false with more detailed observation of the spectral signature over a longer period.
A brown dwarf impacting the surface of the star would release an enormous amount of energy and cause an identifiable spectrum signature of metals, lithium including an IR one etc. Even an earth type planet would dirty the atmosphere with metals which would be identifiable, the spectrum of this star is shown to be unremarkable i.e. normal for its type. A good bet would be to get a good look at the spectrum but been so far away it may prove difficult.
I am in favour of an evaporating moon/s and or planet creating firstly these deep dips (episodic eruptions) possibly with a ring system and then the longer duration dimming as the cloud expands into space. I have a feeling it is water been broken down into hydrogen and hydroxyl ions that has the dimming effect, a good spectrum analysis could reveal this.
I like the idea that an evaporating planet or super ring structure could be causing these observations.
Something like J1407b may help explain some of the dimming events.
http://www.rochester.edu/newscenter/gigantic-ring-system-around-j1407b/
The concern is the lack of IR emissions which would have been expected if rings and/or jets of matter were causing the ‘dips’/dimming.
Another option could be an unbound (or loosely bound) cool brown dwarf with an enormous (and complex) ring structure. I suspect the problem with this option is the fact that significant dimming has occurred since the 1950s not just in the past 4 or 5 years.
Water is a powerful absorber of IR and higher energy light.
https://upload.wikimedia.org/wikipedia/commons/1/18/Absorption_spectrum_of_liquid_water.png
If the planet is in orbit it is in an eccentric and highly inclined orbit, almost polar, as the star is tilted ~68 degrees a ring system could have an almost perpendicular position to the line of sight and thus maximise the blocking power. We may be dealing with a low massed planet that allows a large amount of material escape because of its low gravity or from the moons, the moons may be altering the distributing of material skewing the light curve.
http://vega.lpl.arizona.edu/~gilda/images/evap_e.gif
It is worth pointing out that this “Planet X” could potentially be a Stevenson planet* with a potential biosphere. If it has kept its hydrogen and helium atmosphere, it could have a warm surface, even surface oceans, from trapping of the heat released by radioactivity in a dense atmospheric greenhouse.
* http://www.nature.com/nature/journal/v400/n6739/abs/400032a0.html
Are any close the KIC 8462852 showing any signs of beginning to dim in a similar manner?
That’s what the check stars were for. Nearby stars of about the same color. It’s all in Schaefer’s paper:
http://arxiv.org/abs/1601.03256?utm_content=buffer78842&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer
Here is an article of an evaporating planet,
http://vega.lpl.arizona.edu/~gilda/extrass.html
Notice the depth of the dimming in the light curve is around what we get for Tabbys star. If a water moon was evaporating which has a lower gravity the cloud could be as big and further out from the star, it would also tend to form a ring around the planet making the dimming event much larger. The strange deformations in the light curve could amplified be due to a gravity darkening effect.
Apparantly, Bradley Schaefer is NOW going to look at a COMPLETELY DIFFERENT STACK of plates! I hope these NEW plates have images from EVEN EARLIER THAN 1890!
If you are talking about the Sonnenberg Observatory collection of survey plates (which Shaefer says is the only comparable collection out there), it only goes back to the 1930s. But it does completely cover the 1950s and 1960s , unlike the Harvard Observatory collection he had used in his recent paper, which is a big gap in the existing data. Also it extends closer to the present day than the Harvard collection, which ends in 1990.
Someone in this long discussion thread rightly mentioned the important issue of the relatively short lifespan of an F3 type star such as KIC 8462852.
I would roughly guesstimate, based on mass and luminosity, that the main sequence lifespan of this star is only about 2.5 – 3 gy.
More importantly, the window for complex/multi-cellular life will become very short, if not totally absent, also see this post and comments:
https://centauri-dreams.org/?p=25359&cpage=1#comments
And if the intelligence is not indigenous, but migrated there from another planetary system, I could think of much more suitable and long-lived systems.
Therefore, I tend to have strong doubts about the intelligence explanation for KIC 8462852.
F type stars emit substantially more UV than G types and therefore likely to produce more oxygen when water is broken down. If oxygen was formed from the breakdown of water and life evolved to use early on the rate of evolution may have been accelerated somewhat. In effect speeding up advanced life’s development before the star leaves the main sequence. There tends also to be more energy available in the light spectrum of F-type stars than G-types if we use our photo life forms as a baseline. So it could be a six or two threes situation.
http://plantphys.info/plant_physiology/images/psnpigmentspec.gif
I find this simulator a helpful way to see the habitable zone around a star move outwards for various massed stars, have a play around, the low massed stars have HUGE lifespans.
http://astro.unl.edu/naap/habitablezones/animations/stellarHabitableZone.html
Going on the Dyson Swarm theme, perhaps the ETI building the structure are not natives to the system but colonizers remaking the place to suit them?
The 15% obscuration re Eder red to in the abstract is in the EUV, so how is that applicable?
Ronald: My sentiments EXACTLY regarding the F3 star. BUT, don’t forget about the PUTATIVE M dwarf companion! If THAT STAR is a much OLDER star that was captured(naturally or UNNATURALLY) by the F3 star, and its inhabitants are now in the process of constructing a Matrioshka Brain around the F3 star, STARTING with the OUTER matrioshk, ALL OF THE EXISTING DATA WOULD fit, including the LACK of EXCESS IR! A good science fiction read pertaining to this scenario is Brian Aldiss’s “Heliconia” trilogy.
Here is something interesting and relevant here. Astronomers have found something like s super Planet X around a nearby star – this object maybe 10 Jupiter masses, is maybe 6800 AU from its host star and would have a period of ~ 1 million years (should it be in a bound orbit, which is pretty likely but not certain at present).
Now, WISE has ruled out a Jupiter or superJupiter closer than about 1 light year from the Sun, but this shows that “Planet X’s” are possible, and also that they can apparently form in situ, at a considerable distance from their star.
The Paper : http://arxiv.org/abs/1601.06162
An article in the press: http://www.smh.com.au/technology/sci-tech/astronomy/astronomers-find-parent-star-of-lonely-planet-20160125-gmddg8.html
It may have formed closer in and got thrown out in a gravity tussle with another planet, 40 odd million years is a long time so it could have occurred. It does have a strange analogy with a possible planet 9 in our solar system though.
Are you thinking that the Tabby’s Star Anomaly may be planetary system far away from the star, with a wide debris field of cold objects?
A co-moving/bound object with a very extensive cool ring/disk/debris field sounds like a better explanation than a “swarm” of giant comets lasting decades.
I guess we will have to wait until more “plates” are analysed so we have a comprehensive history from the 1890s onwards.
One question I have is if Schaefer is convinced we have a “secluar” trend in dimming then shouldn’t we have seen a dimming during the Kepler years of between 0.5-1.0%?
We should have a very accurate measurement of the magnitude of KIC 8462852 from Kepler and any “secular” dimming would have been obvious.
Anyone able to enlighten us on this aspect?
It would have to be done with a MICROSCOPE! You would need to choose the most QUIESCENT part of the graph RIGHT AFTER OBSERVATIONS STARTED as your START POINT, and the most quiescent part of the graph JUST BEFORE THE 80 DAY 2013 EVENT starts as your end point, draw a streight line between these two points, and see if the line slopes DOWNWARD! MORE IMPORTANT, however, would be to access ANY AND ALL CCD IMAGES of KIC8462852 between 1890 and 2010, and plot a graph of them, LINKING the DASCH and Kepler data sets. A 4-5% dimming over these 20 years, in my opinion, would CONFIRM Schaefer’s results.
That would be difficult to do, because the spectral response of CCDs are different from that of the blue-sensitive photo emulsions used in the Harvard Observatory survey. A slight difference in spectral response can make a big difference in measured brightness.
Exactly. The important thing will be to use surveys where care has been taken to keep spectral bands and equipment issues comparable. For example, the Palomar Schmidt telescope survey in the 1950’s was matched by Palomar II more recently – comparing Palomar and Palomar II should be pretty straightforward for a 12 magnitude star.
It would have to be huge, much bigger than the 7 Earth masses Thompson, et.al. considered within 200 AU as a possible fit to the mm wave data. however, this completely fails to explain the timing between the big dips as reported by Boyajian, et. al.
I looked up the opacity of Saturn’s rings (http://articles.adsabs.harvard.edu/cgi-bin/nph-iarticle_query?bibcode=1984prin.conf..737C&db_key=AST&page_ind=4&data_type=GIF&type=SCREEN_VIEW&classic=YES), it seems that optical tau averages about 0.4, which corresponds to 90% transmission looking straight through. So, assuming the same for the putative ring system eclipsing Tabby’s star, it would have to be larger than the the star, and still would only result in 10% obscuration (you could play around with geometries and get a little different answer, but as a large as the star is just crazy).
Oops – that’s the log of optical Tau is 0.4. Works out to about 2.5, or about 0.80% transmission. In the B ring it can get higher.
The mass of the Saturn ring system based on Voyager observations was estimated to be about 3 x 10^19 kg which is not a great deal of material and it is around 300 000 km in diameter. So it is not unreasonable to expect a larger more massive ring system as Rafik commented earlier .
http://www.rochester.edu/newscenter/gigantic-ring-system-around-j1407b/
most of the high-Tau regions are within 100,000 km of Saturn. The stuff way out contributes little.
If we tried a thought experiment in we exposed Saturn to higher light levels we would most likely see Titan and all the moons for that matter under goes eruptive outgassing. Any water would quickly turn to vapour and be broken up into ions, any methane exposed to U.V would under go chemical changes to form complex molecules or been broken up into carbon based compounds. All of these molecules and atoms absorb part of the light of the spectrum and contribute to the dimming effect, dust may be present but any IR light it emits may be absorbed by the water vapour. A though examination of the spectrum could reveal these details.
A planet HD 209458b is under going an evaporation phase and blocks out a lot of light up to 15% with just hydrogen, hydroxyl and carbon ions. A moon evaporating would create an even bigger cloud, complicated by gravity interactions with the planet, due to its low escape velocity. The magnetosphere of the planet may even affect the shape of the cloud as it interacts with ions in its sphere of influence.
http://vega.lpl.arizona.edu/~gilda/extrass.html
And this made me laugh,
https://twitter.com/astrodave2/status/655042955720421376/photo/1
HD 209458b is a transiting gas giant planet around a G class star in a 3.5 day orbit. It gets roughly 400 suns (compared to Earth at 1 AU) power. Clearly a completely different signature from Tabby’s star.
If the object is on an eccentric orbit with the orbit ‘aphelion’ pointing towards us the object would rush in ‘get cooked’ and move back out again, flash fried if you will. A lower gravity object will allow a lot of escaping material and hence a larger cloud for less light.
The chart you show is for liquid water. Water vapor mainly absorbs in the infrared, and is effectively transparent in the visual. With Tabby’s Star pumping out UV, much of that would dissociated into OH and H, which are again transparent in visual.
Paul – Just out of curiosity, what’s your best guess at an explanation about Tabby’s star?
Honestly, Andrew, I don’t have one. This object is so mysterious that I can’t get a handle on it. If it’s a new astrophysical process we’re seeing, it’s a major find just for that reason. I do not discount SETI-related explanations, either. My view is that we just don’t have enough data — we need a new round of observations, and that probably won’t settle it. I suspect Tabby’s Star is going to be an unresolved problem for some time to come.
PS: More about this in tomorrow’s post.
I’m happy to admit that I have no idea, any more than I know what dark matter is or how the origin of life happened or what came before the Big Bang. I hope it is something really exotic, but am not holding my breath.
http://arxiv.org/abs/1601.07314
KIC 8462852 did likely not fade during the last 100 years
Abstract: A recent analysis found a “completely unprecedented” dimming of 0.165±0.013 magnitudes per century in the F3 main sequence star KIC8462852. This star is interesting, as it shows episodes of day-long dips with up to 20% dimming of unknown origin. We re-analyze the same Harvard archival Johnson B photometry and find comparable dimmings, and structural breaks, for 18 of 28 checked F-dwards (64%) in the Kepler field of view. We conclude that the Harvard plates photometry suffers from imperfect long-term (1890–1989) calibration. The most likely explanation for the century-long dimming of KIC8462852 is thus a data artefact, and it is probably not of astrophysical origin.
Wow!
Thanks Bill!!! …but the check stars? They were the calibrators, no? How can these two opposite conclusions be reconciled? Science at its best and we’ve got a ringside seat.
Hi Coacervate.
Schaefer used 5-year “bins” and then looked at the overall trend for the whole time period. As I discussed early, I think this is a simplistic and potentially skewed view of the data.
My quick analysis (using 1-year bins) showed that there was a significant dimming event during the Menzel gap for KIC8462852 and a slight dimming trend at other times (but that was within the error bounds of the data).
The new study appears to support that view of the data.
What I am yet to be convinced about is that the “break” during the Menzel gap is solely due to an artefact of the data.
Hippke and Angerhausen state that they show a “structural break” at 12? confidence during that time. They mention a 10? confidence “structural break” for KIC7180968. Finding one other significant structural break in 28 stars during the Menzel gap I would think is insufficient to completely discount the possibility of a real dimming event for KIC8462852.
I think we need further data from the likes of Palomar POSS and POSS II to resolve this issue.
Very interesting paper – and apart from the “structural break” during the Menzel gap there no longer appears to be a long term dimming for KIC8462852.
I would be very interested to know if the “structural breaks” for the other stars in the study show a similar dimming to KIC8462852, a brightening as shown for KIC 7180968 or a random mixture of dimming/brightening.
This would tell us if there is likely to have been any event affecting KIC8462852 during the “Menzel Gap” or the break is caused by an artefact of the data.
I hope that someone is examining the Palomar plates from 1948-58 and Palomar II from the 1980’s to 1990’s at a similar comprehensive level as has been undertaken in this study.
Hippke and Angerhausen state: The two comparison stars used by Schaefer (2016) (TYC 3162-1001-1 and TYC 3162-879-1) show a con- stant luminosity (within their errors). For our 28 com- parison stars, we find a constant luminosity for 10 of 28 stars (36%). Assuming that all stars have been drawn randomly from the same sample, the chance of drawing 2 of 2 constant stars is 13%. It might be attributed to bad luck that these apparent data discontinuities were not seen in the first place.”
I don’t understand how some of the stars on the same plate can be constant and (many?) others show apparent dimming? Can this be attributed to artifacts within the plates themselves?
My first guess would be that some main-sequence F stars are in fact variable on a decade-long scale, in the manner seen among many of the F stars seen in yesterday’s paper. So this would be a previously unknown type of variable star.
Bill: THAT makes the most sense, so far. Have the data been examined for non-constant stars in proximity to Tabby’s and compared to, say, a section of sky 180 degrees away? Could point to interstellar clouds near Tabby’s, etc.
Discovering variable F stars would be, perhaps, even more exciting than discoverint life. We can point to another example of life in the universe, but a dimming F star WOULD be news.
I think we are talking about many different plates covered by the 28 stars. The new study covers F stars from the whole Kepler field of view – not sure how many plates that would equate to.
Even the comparison stars in the Schaefer study do not appear to always come from the same plates as KIC8462852. Sometimes they may be on the same plates at other times they may be on different plates – I am sure someone can clarify that for us.
Schaefer actually looked at 5 check stars, but only reported two in detail because they were the closest in color.
Yes, and in retrospect “it would be nice” to be able to see what the other 3 check stars are doing, even if they aren’t good matches in color.
Note in his Figure 1 that all 3 stars shown had low magnitudes in the 1908 bin. With only 3 points, it is impossible to be sure if this is significant, or just happenstance.
Harvard ran the “Sky Patrol” (an effort to regularly photograph the entire sky) for decades, which is where a lot of these plates came from. I believe that the “Menzel Gap” was due to Menzel shutting off the Sky Patrol, and a successor restarting it.
This Sky and Telescope article from 1965 is entitled “Harvard Observatory’s New Patrol Cameras” (volume 29, page 200). This implies that the Sky Patrol came back online with new cameras, and quite possibly new color sensitivities, which could certainly affect the photometric stability of the Sky Patrol data across the gap. (It is also possible that how the Patrol tiled the sky was changed at the same time; two stars might possibly be on the same plate before, and different plates after, or vice versa, which would make this pair a bad choice for a target and a check star.)
http://adsabs.harvard.edu/abs/1965S%26T….29..200I
This paper seems relevant : A disintegrating minor planet transiting a white dwarf ( http://arxiv.org/abs/1510.06387 ):
“Here, we report observations of a white dwarf being transited by at least one and likely multiple disintegrating planetesimals with periods ranging from 4.5 hours to 4.9 hours. The strongest transit signals occur every 4.5 hours and exhibit varying depths up to 40% and asymmetric profiles, indicative of a small object with a cometary tail of dusty effluent material. The star hosts a dusty debris disk and the star’s spectrum shows prominent lines from heavy elements like magnesium, aluminium, silicon,
calcium, iron, and nickel.”
I would think that if KIC8462852’s dimming was due to the disintegration of a planetary system, it would exhibit some of these signatures, which I don’t think have been observed.
From Schaefer (2016) :
“Magnitudes for stars on the photographic plates are always taken by comparing some measure of the image diameter with the diameters of comparison stars on the same plate.”
The probable reason that there are no large IR emissions is that the photovoltaic surface of the Dyson swarms are nearly 100% efficient in converting the radiant energy of the incident light into useable electric power.
Except we think the law of thermodynamics are probably unbreakable, and if you CAN break them, you have no use for even these completely fictional 100% efficient solar panels. The only way I can see around the waste heat problem is to direct the energy elsewhere, where we have yet to look for the waste heat. Maybe I should look in AllWISE for a source that isn’t known to correspond to a star.
Perhaps, but what is that power being _used_ for? Because presumably by one path or another it’s eventually degraded into waste heat.
If there is excess IR it is either (1) being directed away from our line of sight or (2) it is so low in temperature (<100K) we can't see it. Are there natural processes anyone can think of that will cause (1) and/or (2)?
for (2), an improbably huge and massive dust shell far enough from the star that it’s only marginally warmed by it. I presume that’s been ruled out.
I wouldn’t say strictly ruled out, but constrained by the mm wave observations to date. It seems very improbable.
They didn’t see much in the mm wave data, which is sensitive to very cold thermal emission (http://arxiv.org/abs/1512.03693v1). The WISE 22 micron W4 band (http://wise2.ipac.caltech.edu/docs/release/allsky/) is fairly sensitive around 100 K, and interestingly, B15 reports a slight excess there, but most of the W4 observations seems to be at the limit, so I’m not sure you can really use that except as an upper limit. I’m not a savvy WISE data user, so I’d hesitate to conclude anything.
The other option is that the energy harvesting device is very hot, right down on the surface of the star, so it is almost masked from observation, making just a slight bump in the curve when it eclipses. Of course, I have no idea how anyone could build such a thing….
Paul Carr: I have just discovered (and am now enjoying!) “The WOW! Signal” and “The Unseen” podcasts. Great listening!
General comment, Tabby’s Star and the lack of detectable excess IR: (1) Perhaps excess IR is direction-ally radiated, like a beam or a cone (could be natural?). Perhaps most of it is even “aimed” up or down in the ecliptic … so that whatever is causing the episodic dimming…. you can only see the excess IR at that dimming event…. it can ONLY be seen sporadically. We would have to be looking at the star while an actual event is taking place, which is in the ball park of every two years-ish for about a week’s duration. It may even be there BUT get smoothed out by averaging unless we are looking at it “real time”.
(2) Are we confident that our IR sensitivity is capable of detecting such a anomaly from 1500 light years away? We should be able to at least constrain what we can detect right now. Is it a reasonable amount?
I’m not sure – after all waste heat is at maximum entropy, so one would expect it to be more or less isotropic (if you can direct it, it’s not waste heat), but it does somewhat depend on the geometry, and we don’t understand that at all at present, and the IR observations are a bit sporadic. We also would not expect the occulter to only be absorbing and re-emitting when it crosses our line of sight – there should be nothing special about that particular direction.
The IR sensitivity is not as good at the long wavelengths, so all we can do with say, the WISE W4 band is establish an upper bound. The upper bound looks promising, but it’s hardly definitive. There are at least 4 papers now discussing the IR and mm excess, and none of them are claiming to see anything significant. The JWST, once it is operational, should be able to provide tighter constraints.
Paul Carr: Do you know how the Schaefer paper has been received in the weeks since the pre-publication? I remember that you had someone on The Wow Signal podcast who was skeptical of the work, and that the data binning was done improperly. I’m curious to know how the field has reacted to the manuscript in the meantime.
I think the ball is in Schaefer’s court. If there is a fade signal there, it n is poking just above the noise.
http://disownedsky.blogspot.com/2016/02/tabbys-star-for-perplexed.html
This is an excellent summary, Paul.
Congrats Paul – this is an excellent summary of the current state of play.
I particularly like this statement.
“So, Where are we Now?
At present, everyone is stumped, and we need more data.”
I have seen no suitable hypothesis/model that explains ALL the features in the Kepler light curve for KIC (and possible longer term dimming).
My best guess is that we have observed several separate intriguing events that require multiple causes – I seriously doubt we are looking at just one causal factor.
Thanks, I recently got some helpful comments from Jason Wright that I am thinking about. It’s done, but soon I plan to record an audio version. I want this to be accessible to the intelligent layperson not well versed in astronomy (unfortunately, such a person could well exist).