A KIC 8462852 (dipping again in March 2018)

[mfb]

If it is something that breaks apart the structure could get more complex each orbit. We'll see the what happens this time over the next few weeks, and then get more data mid 2019 (in the worst case we don't get any dips, but that would help ruling out models as well).

 

[craigi]

In the graph for the first dimming event, the baseline fluctuation seems to have disappeared completely for the 12 day period. Even the measurement error of about 0.0025 (inferred from the recently posted graph for the current dimming event) isn't apparent in the data. That makes me suspicious of either the data or how it is being represented.

Looking at the curve for the second dimming event, we can see that the graph seems to use a curve interpolation. I wonder how many data points actually contribute to that first curve. A small number over the 12 day period would explain a lot.

Alternatively, if there a large number of data points and the measurement error from the data for the current dimming event is relevant for the first dimming event, then we must ask how those measurements became correlated.

 

[mfb]

In the graph for the first dimming event, the baseline fluctuation seems to have disappeared completely for the 12 day period.

Note the different scales on both axes.
If you scale the noise seen in the large plot to the short period in time but the large range in y, you don't see it any more.
Both curves are probably smoothed (typical raw 15 minute measurements have significantly more noise), but on a scale of an hour or so, not on the scale of days.

 

[ExecNight]

Couldn't it be a pair of black holes between KIC 8462852 and Earth that are orbiting each other. Then this would explain how it regularly (and also irregularly) blocks the light from the star, since one black hole or the other keeps getting in the way while they are orbiting each other.

If these two are near collusion, gravitational waves from these two might be detected?

 

[Bandersnatch]

Then this would explain how it regularly (and also irregularly) blocks the light from the star, since one black hole or the other keeps getting in the way while they are orbiting each other.

No, this is the same scenario as with any other non-black hole body - all closed orbits are periodic, regardless of whether it's a black hole, a planet, or another star. In your sentence the bit in the parentheses does not then follow from the premise.

 

[mfb]

A black hole tends to focus starlight, so we would see more light. The shape of microlensing events is well-known, and looks completely different from what we see here.
Anything not close to the star would need an absurdly precise alignment in terms of velocity in order to stay in the line of sight for years. Whatever it is, it is close to the star.

 

[ExecNight]

A black hole tends to focus starlight, so we would see more light. The shape of microlensing events is well-known, and looks completely different from what we see here.
Anything not close to the star would need an absurdly precise alignment in terms of velocity in order to stay in the line of sight for years. Whatever it is, it is close to the star.

Very interesting indeed.

So facts are;

- Very close to the star.
- Can cause dimming up to %20
- Not a dead star itself, since universe is not old enough.
- Can't be a remnant planet
- Can't be a thick gas cloud.

Then, what is the current theory? Except the *ALIENS* one?

 

[arabianights]

i think it's time to bring in dark matter/dark energy to explain the strange behavior of distant stars.

 

[mfb]

A very messy gas/dust cloud as result of some recent collision process seems to be favored for now. The main components orbit the star every 750 days, and the process is still very dynamic so we see different events each orbit.
It is unclear if that explains the overall dimming, and it is hard to make that consistent with the lack of an infrared excess.

We are back at the baseline luminosity.

 

[Noel]

I'd like to invoke the term "Dark Shadows" as an explanation

 

[stefan r]

i think it's time to bring in dark matter/dark energy to explain the strange behavior of distant stars.

"dark matter" does not effect visible light. Noel nailed it:

I'd like to invoke the term "Dark Shadows" as an explanation

We still need to figure our why the aliens are building dark shadow megastructures. Why are they not radiating infrared or using radio?

Hasn't it been determined that the most likely scenario is that the star ate a planet? That causes an increase in brightness, which will then, of course, dims back to its original brightness over time. The planet gets ripped apart before being swallowed creating a massive dust cloud.

Why would the dust cloud have a 750 day orbital period? What happened to the infra red radiated by the dust? For example, dust around vega is easy to measure.

 

[newjerseyrunner]

Why would the dust cloud have a 750 day orbital period? What happened to the infra red radiated by the dust? For example, dust around vega is easy to measure.

Hmm... 750 days is a little long, I'd expect the dust cloud to be in a tight death spiral. However, I could come up with a scenario: two planet orbit close to each other around that 750 day radius. During a close approach, the planets swing around each other. One gets ripped apart leaving a dust cloud at the right distance and the more massive one gets thrown into the star, causing temporary brightening, which we are now observing the tail end of.

Question though: what would the dust cloud of a such a circumstance look like? Vega's dust ring is 2-3x the distance to Pluto, where this dust cloud is nice and cozy between what would be the orbits of Mars and Jupiter. Vega's dust cloud is also ancient, where if this star ate a planet, it happened in the last hundred years.

 

[stefan r]

Question though: what would the dust cloud of a such a circumstance look like? Vega's dust ring is 2-3x the distance to Pluto, where this dust cloud is nice and cozy between what would be the orbits of Mars and Jupiter. Vega's dust cloud is also ancient, where if this star ate a planet, it happened in the last hundred years.

I was thinking just of the re-emitted infra red. Dust further than pluto radiates in infrared so dust rings near a Mars radius should be blazing. It can not be some of the possible alien megastructures. For some reason they are reflecting light away or perpendicular. Or adsorbing and radiating away/perpendicular.

Hmm... 750 days is a little long, I'd expect the dust cloud to be in a tight death spiral. However, I could come up with a scenario: two planet orbit close to each other around that 750 day radius. During a close approach, the planets swing around each other. One gets ripped apart leaving a dust cloud at the right distance and the more massive one gets thrown into the star, causing temporary brightening, which we are now observing the tail end of.

A comet starts as an icy ball and grows a dust tail when it gets into the inner solar system. The the tail is already too large before finishing half a pass. When comets are completely dusted they turn into continuous long streams. Leonid meteor shower for example. The graph of tabby's star is fairly tight:

[zeit is German for time, normalized flux] The first event starts around 787 and ends 796 with most of it in one day.

I had not noticed the increased flux around day 1550 to 1560.

 

[Mr P]

Wild speculation perhaps but given the habitable zone for an F3 star is around 3au with a period of ~1500 days is it possible that we're looking at 2 structures on roughly opposite sides of the star? That should cancel out the wobble from tidal shifts though why you'd need to do that is another question.

 

[mfb]

The radial velocity is poorly constrained, there could be quite heavy objects orbiting it. But they have to be unusually cold.

 

[newjerseyrunner]

Is there any information about the mass of the orbiting object? Does it have any Doppler affect on the star at all?

Is there still a chance that the dips are coming from the star itself going through cycles from some natural self organized criticality?

 

[Mr P]

Mass of the object(s) appears to be negligible. Ref the original paper section 2.6

 

[craigi]

Again in this representation of the data, the fluctuation disappears almost entirely around day 788 and emerges again around day 795. Then disappears again completely between days 802 and 809.

The magnitude of the fluctuation outside of those ranges is about 0.001 or less, whereas the error bars in the recently posted graph for the current dimming event is about 0.0025.

I can't interpret this as a chance occurrence and since this is the section of data which presents the greatest difficulty for analysis, I think we should understand what's going on with this representation of the data in more detail.

 

[newjerseyrunner]

So everything seems to point to something comet-like. Absolutely huge physically, but almost no mass. A swarm of comets though would not make such a smooth light curve. There is also a known small star nearby that's passed close enough to disrupt objects about the distance of the Oort cloud.

What about an object much larger than a comet? What would happen if you took Pluto and threw it into the orbit of Jupiter? I imagine it'd act like a comet on steroids: the surface nitrogen would sublimate. That'd create a huge, very diffuse atmosphere that'd slowly leak into space. It'd also be cold, and wouldn't create an infrared signature.

 

[mfb]

Again in this representation of the data, the fluctuation disappears almost entirely around day 788 and emerges again around day 795. Then disappears again completely between days 802 and 809.

They don't have to disappear, they are just harder to see. Fluctuations on a constant line (where the curve is going up and down) are easier to see than on a heavily sloped line where you just get slightly smaller and larger first derivatives.

 

[Mr P]

So cometnado has morphed into supermegacomet, though I'm not convinced by the evaporation rate.
Then there's magneto-phase-thingy-avalanche if you can make head or tail of it.
Or 2 megastructures in the habitable zone with frozen backsides...

 

[craigi]

They don't have to disappear, they are just harder to see. Fluctuations on a constant line (where the curve is going up and down) are easier to see than on a heavily sloped line where you just get slightly smaller and larger first derivatives.

Sure. I wouldn't expect to be able to see fluctuations in the region where the curve dips below about 0.97 due to the high derivative, but that still leaves a lot of data where I would expect to see fluctuations, in the first range I refer to.

Also this doesn't account for the absence of fluctuations in the second range I refer to.

 

[Hdeasy]

The super-comet idea doesn't explain the long-term variation over the 4 years of Kepler data - and no other star in the survey showed this. At least the Dyson idea accommodates it as the long term dimming could be due to a steady, homogeneous filling out of a grid in the sphere. The short term asymmetric variations could be due to irregularly shaped patches.
Also I find it idiotic that all mention of the Dyson structures focuses on solar panels - any student of Orbitsville or Ringworld knows it's all about real estate.

 

[Mr P]

If you're going to consider the Dyson structure hypothesis (and dodging it really isn't going to do you any favours) you'd have to talk in terms of needs rather than wants, the sheer investment in time, material and who knows what other resources would preclude just about anything else.
If they are Dyson structures they aren't radiating heat which implies they're grasping at every last scrap of energy.
The obvious next question would be 'what for?'

 

[newjerseyrunner]

The super-comet idea doesn't explain the long-term variation over the 4 years of Kepler data - and no other star in the survey showed this.

Why wouldn't it? An icy dwarf planet suddenly tossed closer to its star would not only make a huge coma around itself but also lose a lot of material, which would form a ring of cold material around the star. Similar to how Enceladus's constant outgassing helps feeds Saturns rings. A ring forming would make the star brightness vary greatly. Would we be able to detect such a structure?

 

[craigi]

Data can be found and plotted here:
http://archive.stsci.edu/kepler/data_search/search.php?action=Search&ktc_kepler_id=8462852

 

[mfb]

An updated light curve, the dip is now clearly over. We'll see if we get more in the coming weeks.

 

[craigi]

An updated light curve, the dip is now clearly over. We'll see if we get more in the coming weeks.View attachment 204349

Do we think there's significance in the the tail-end data beyond 10 days or is this just within normal fluctuation?

 

[mfb]

Looks just like measurement issues to me, different experiments show differences of similar size.

 

[craigi]

My initial assumption was that there was an occluding object or cluster of objects with an orbital period of 750 days, but could it be that the orbital plane is rotating or oscillating with a period of 750 days and what we're seeing is completely different objects each time the plane is aligned?