[Guest Post] Newly published paper on a 1574-day periodicity
As an author to a newly published paper about KIC 8462852, Tabby has invited me to tell you about it here. For those of you who don’t know me, I go by the Reddit handle of ‘gdsacco.’ I work in the cyber security domain, running a team for a global technology company. This is an important part of this story because it is not what I do for a living, it is what I do not do: astronomy. So how did I get swept into the most mysterious star in the galaxy?
Before this star started making news, I wasn’t even an amateur astronomer. So getting interested in this star was never because I wanted to know if there was some dust and gas in the interstellar medium, or if there was something wrong with the star itself that was causing these wild swings of flux. Like so many others that never before had an interest in astronomy, I got sucked into this story for one reason: the prospect that maybe, just maybe, we humans could finally show that we are not alone.
As it turns out, Reddit provided me (and many other would-be amateur astronomers) a forum of education in the field. If you peruse the KIC 8462852 subreddit, you’ll find a melting pot of experiences and backgrounds, with everyone after the same thing: what is up with this star? And while there are definitely personality clashes found there at times, most of the moderators are astronomy pros and generally do a good job at keeping speculation under control and the overall subreddit grounded. You just can’t throw out non-sense without getting pounced on. And as I look back on that aspect, that kind of environment serves as a kind of foundational astronomy classroom for those active there.
Elsie was a lost Kepler event
Then in May 2017, LCO started reporting a fresh new dip. And as we all now know, several more dips would happen afterwards. It was in July 2017, when I recognized that Elsie was actually a missing dip in the Kepler dataset!
Like so many others, I was busy trying to line up the Elsie light curve to Kepler dips (and failing to do so). Then when Celeste arrived one month later, it was an interesting timing comparison to the final days of Kepler observations in 2013. Something that occurred to me was while the amplitude intensity was clearly not aligned, Celeste fit Kepler D1519 in shape and duration. Then the exciting moment happened when I overlaid Celeste to D1519 and looked to see where that would place Elsie: the Elsie dip slipped right into a period of the Kepler dataset during a period (D1487) when Kepler was not observing! With this realization, I predicted the return of the D1540 group in July. Sure enough, we saw a long shallow trend of flux depression that lasted half the month and that fit perfectly in duration to the Kepler light curve. Now I had two dips lined up nicely and a third falling in a missing data period in the Kepler dataset. With that, I was soon able to refine a periodicity calculation to ~1574 days. Using that, you would expect Kepler D1568’s peak to return on August 9, 2017. Right on schedule, Skara Brae peaked on that same day, as confirmed by LCO. At this point, it was suggested by several Reddit readers to pull together a paper. I sought co-authors from the community and we were able to get the paper posted to arXiv in October 2017 (we are currently working on posting an update there).
The Castelaz et al., paper
Along the way, there were skeptics to the hypothetical periodicity. That’s because between Kepler and LCO, there was not nearly enough data to show statistical significance. Then, a new paper by Castelaz et al. was published at JAAVSO that did two important things: (1) it supported the Schaeffer earlier finding of long term secular dimming, and (2) it identified five potential historical dips using achieved plates at the Maria Mitchell Observatory. The five dips were exactly what we were looking for to further the significance to the proposed 1574-day periodicity. Using the new dips’ dates, we found the good news we were hoping for: two of the five historical dips would have fallen within the same window of time (using 1574 and Kepler D1487 - D1568 and May 2017 - September 2017). For these two dips that did fit our window, our question was, do they align to any of the Kepler and LCO dips using an ~1574-day periodicity? Excitedly, I found that both hit the jackpot by landing precisely on the day as expected.
Maria Mitchell Observatory Match 1:
Skara Brae minus 1574 = Kepler D1568
Kepler D1568 minus (1574.4 X 9) = October 22,1978
Maria Mitchell Observatory Match 2:
July 2017 dip minus 1574 = Kepler D1542
Kepler D1542 minus (1574.4 X 18) = August 21,1935
To add significance, we also noted in the paper that Hippke et al. used Sternberg historical plates to identify a dip on October 24, 1978! Now we had two separate observatories with high quality plates seeing the same dip. Armed with these results, our paper was updated and presented to the Journal of the American Association of Variable Star Observers (JAAVSO) in May 2018 where it was accepted and published. See here for the updated / published version of the paper.
But what about the other three Castelaz et al. dips? We don’t completely know yet. Potentially, they may suggest the other Kepler dips outside of the Kepler Q4 events (D792, etc.) are on a different orbit. Or, maybe not. For example, using 1574 days, we would expect the return of D260 on May 3, 2018. Unfortunately, the days before, during, and after that date had poor weather conditions for the LCO network. That said, Bruce Gary did make mention on his webpage of a possible small (~1%) dip on May 3, 2018. So, who knows? Stay tuned!
What it all means
I know so many want to know in plain English what all this means. If confirmed, a 1574-day periodicity further constrains several potential long-standing causes of the ongoing dips, such as intrinsic variation or dust in the interstellar medium. Simply put, it means there is something in orbit around the star at about 3 AU (within the habitability zone assuming a circular orbit). Furthermore, this paper may lend support to the Boyajian et al. (2018) paper, in which fine dust was identified as a possible cause of the dips. It also adds intrigue to the Boyajian et al. (2016) paper with respect to the observation that many dip separations are multiples of 24.2 days. Interestingly, a 1574-day period is equivalent to 65 even periods of 24.2 days. Furthermore, when you combine orbiting material with century-long dimming and no detected IR excess, it adds more mystery to this star….not less. The fact is, we still don’t know what the source is of such dust. Perhaps a ring of broken comets that are slowly moving into our line of sight thereby dimming the star over decades or even centuries. Or maybe, just maybe, something unnatural. In any case, whatever the cause, this paper should make you want to say "WTF" even more so!
While we need to keep observations going to further develop statistical significance of a 1574-day periodicity, there’s something that has happened here that is far bigger than this star, or even potentially astronomy itself. The use of public-power through programs like Planet Hunters, Kick Starter, or social media like Reddit, is a largely untapped and plentiful resource that sits waiting to be unleashed. After all, this star was first discovered by Planet Hunter's 300,000+ group of volunteers in the public domain. Imagine for a moment if we hadn't had Tabby’s group of Planet Hunter's….Boyajian's Star would be completely unknown today. Or if Tabby had not provided me the LCO data to conduct the analysis, no one would have blamed her... I'm not an astronomer, right? So, regardless of how this amazing star's story unfolds going forward, perhaps the biggest story here isn't about the star at all. Maybe instead, it's an example of how motivated ordinary people can be used as an important new tool in science. Tabby is a real pioneer in this new space, and so far, it seems to be working just fine.
--- Gary Sacco (Florida, USA)
PS: [Message from host] The LCO observations are happening because of the wonderful backers of the 2016 Kickstarter project. The Kickstarter campaign has ended, but we are still accepting donations to purchase additional observing time on the LCO 0.4m network. Thanks in advance for your support!