It’s been a huge amount of work, but we finally have some conclusions.
First and foremost, the dips have now been observed by an instrument other than Kepler. So we can firmly rule out instrumental effects! (This was already clear, but it’s now true beyond all doubt for the dips).
Secondly, the dips are clearly chromatic:
Eva Bodman has done a lot of work to characterize how much deeper the dips are at blue wavelengths than red ones. If there were opaque objects blocking our view of the light, the star should get equally dim at all wavelengths. Instead, Eva finds that the blue (B) dips are much deeper—about twice as deep—as they are when we look at infrared wavelengths (i’ band, just beyond human vision).
This is consistent with ordinary astrophysical dust, and a major conclusion of our paper: the dips are not caused by opaque macroscopic objects (like megastructures or planets or stars) but by clouds of very small particles of dust (less than 1 micron in typical size). We can also say that these clouds are mostly transparent (“optically thin” in astrophysics parlance).
Secondly, we have spectra from Keck/HIRES both before and during the dips (in-dip spectra kindly contributed by John O’Meara, Jay Farihi, and Seth Redfield; see the black lines in the above figure for when they were taken. Pre-dip data taken by Andrew Howard and Howard Isaacson.) The difference between these spectra should bear the spectral fingerprints of whatever is causing the dips! So, is there atomic gas? Let’s check the neutral sodium lines (analysis courtesy of Jason Curtis):
The black points and red line in the figure above are from before and during the dips, respectively, and the black line at the bottom is the difference. As you can see, there is no obvious change in the spectrum at all. This strongly suggests that the dust causing the dips is not accompanied by much neutral sodium.
What about hot gas? If it’s really hot there shouldn’t be much in the way of dust, but if it’s warm there should be ionized calcium. How do the calcium lines look?
Again, no change, so it looks like there is no additional ionized gas accompanying the dust. So the dust—if that’s what it is—seems to be by itself with no accompanying gas.
In fact Jason Curtis has gone further, and shown that there does not appear to be any change in the stellar lines, either, during a dip, meaning the star is not moving, so does not have a nearby companion orbiting it.
So where are our 10 possibilities?
As I wrote, instrumental effects (#1) are now firmly ruled out.
The hypotheses I found most plausible involving an interstellar gas and dust cloud (#3 and #4), are not looking great. There should have been atomic gas in that case, and we see none.
My favorite (but “less-plausible”) hypothesis #5, a black hole disk, has not been similarly developed, so I think is still in play because we’re not sure what we would have expected to see for that one yet. A cold disk of dust could easily have had all of its gas frozen out onto grain surfaces, I suspect.
The unlikely hypotheses of an orbiting black hole disk (#6), spherical swarm of megastructures (#9) and pulsations (#12) continue to be unlikely.
But now even the more generic “alien megastructures” hypothesis (of any geometry) takes a severe blow from the chromatic nature of the dips: no opaque objects seem to be causing this. I suspect this will be the big headline here, so let me reemphasize: if the dips had been the same color at all wavelengths, we would have been scratching our heads and this hypothesis would be looking better than before (though still of unclear likelihood). The fact that the data came in the other way means that we now have no reason to think alien megastructures have anything to do with the dips of Tabby’s Star (Recall that Meng et al. had already come to a similar conclusion with respect to the long-term dimming, but it was the dips that got us thinking along these lines in the first place).
I still like the Solar System cloud idea (#2) but until it is developed to the point where we know what colors of dimming we would expect for a Solar System cloud, it remains of “unclear” plausibility.
The fact that the stellar lines did not change velocity during a dip helps us rule out pulsations (#12, if the star changed size then its atmosphere would be moving and would have a changing radial velocity) as well as close companions. Tabby had already ruled out the nearby companion hypothesis (which is why it wasn’t even on my list) but we now have independent confirmation.
Hypotheses invoking circumstellar material seem to be doing well. Steinn and I were originally pretty down on this class of solutions because of the lack of infrared excess and their inability to explain the long-term dimming, but Metzger et al. and Wyatt et al. (2017) ‘s models have shown how this could be explained, bringing this class of hypothesis up the plausibility scale to near the top (in my mind). To remind you, Wyatt et al. explain both the long- and short-term dimming with circumstellar material, while Metzger et al. have the long-term dimming being intrinsic and the dips due to exocomet-like debris).
We were also down on the family of solutions involving intrinsic variations, and we still don’t think the polar spots model (#11) and stellar cycle model (#10) have high likelihood. But in addition to the Metzger et al. hypothesis, Peter Foukal has developed a model where the entire star gets cooler, and this model is also consistent with the data (I think the dips seem to be too deep in the blue, but formally it’s consistent at the “2-sigma” level). Indeed, Peter himself finds the dips to be less chromatic than we do and very consistent with his model. So I’m ready to promote this class of solution up, in particular because it predicts no absorption features accompanying dimmings, which is indeed exactly what we see.
As for the circumstellar material solutions (“exocomets”), I’m not personally sure why that model does not predict neutral and ionized gas to accompany the dips, but I don’t think anyone has worked it out in detail yet, so it could be easy to explain.
So to recap:
Wyatt et al. and Metzger at al. have developed models involving circumstellar material like exocomets that seem to be consistent with the data we have. Wyatt et al. and Foukal have developed models where the star itself is getting dimmer that also seem supported. Both classes of model are now at the top of my list, though I still see major problems with both.
Hypotheses invoking intervening material like an interstellar cloud, seem to have taken a blow, though I still want to understand better if they are really ruled out by the lack of gas in the spectra, and whether circumstellar material like exocomets is similarly ruled out. I’m still fond of this solution, but it has gone down a notch in light of the new data.
I think my black hole disk hypothesis is still a dark horse in this race.
And the instrumental effects and alien megastructures hypotheses have been put to bed.
So that’s where we are. The next highest priorities (in my mind) are to scrutinize the in-dip spectra for any signature of the occulting material (I’m especially curious if the diffuse interstellar bands change depth), and modelers need to make detailed predictions of the atomic and ionized gas that should accompany dust in the exocomet, interstellar cloud, and black hole disk models to see if they can be made consistent with our in-dip spectra.