Revised data changed expectations for a star pair that was supposed to merge in 2022.

contact binary star
A simulated view of the contact binary system KIC 9832227. The larger star is about half again as wide as the Sun, whereas the smaller one is about 80% as wide as our star.
L. Molnar / Calvin College

Two binary stars – previously suspected to be merging – aren't going to get together in 2022. A fresh look at the data used in a 2017 study shows that one of the cited references had a typo, throwing off the calculations for binary system called KIC 9832227.

The 2017 study led by Larry Molnar (Calvin College) found that the observed orbital period between the stars shortened over time over several observation sets spanning 17 years. One of those data points was an outlier, recorded about eight years before the next dataset. The outlier — a set of observations taken between April 1999 to March 2000 — came from the Northern Sky Variability Survey, published in 2004 by Przemek Woźniak (Los Alamos National Observatory) and colleagues.

Recently, a new team led by graduate student Quentin Socia (San Diego State University) found that the 2004 catalog had different Julian dates listed in the preprint version than the final, published version. Further, the typo in the published version — the date Molnar carefully copied in his 2016 paper — put the star below the local horizon, making it an impossible observation.

Astronomers use the Julian calendar to calculate dates, but each observatory has the choice of using an adjusted Julian date to move the clock forward by 12 hours, making it easier to deal with local time zones. It was this little quirk that threw off the calculations; one version of the paper used the Julian date, while the other used the adjusted Julian date, said Molnar.

It's unclear how the date changed between preprint and publication. Molnar said he made every effort to get in touch with Woźniak to find out what happened, but couldn't find him; his coauthors said that Woźniak had likely retired. He also asked the Astronomical Journal if they had any records of e-mails or other correspondence from Woźniak related to the 2004 publication, but the journal has moved offices and publishers since then, and the records are lost.

For his own part, Molnar has many years of data he took on his own of the KIC 9832227 system, and his observations are still continuing. He says the adjusted data show the stars might be moving closer together, but he's not sure; at any rate, a merger is by no means imminent. “Even if they're not going to merge, it's still useful,” he says. “Contact binary stars are not well-understood in and of themselves.”. For example, one of his upcoming papers on KIC 9832227 will talk about hotspots observed in the binary system.

Socia's team uncovered the discrepancy using previously unpublished data from 2003 from the NASA Ames Vulcan Project. This project, using an instrument installed on Lick Observatory, was supposed to search for exoplanets in the Kepler spacecraft field of view. While no exoplanets have been found yet, Vulcan's field of view included KIC 9832227.

The team saw that the stellar eclipses were happening a half-hour later than predicted by the merger hypothesis. It was after digging into this further that the 12-hour time difference in the data from 2004 was discovered. The team even took the unusual step of asking Molnar to comment on their unpublished paper during peer review, just to make sure they understood the situation correctly. Typically, peer reviewers during publication have no affiliation with the work.

Socia initially planned to study KIC 9832227 as a summer project, but the work was so groundbreaking that it ended up being the major focus for his Master’s thesis, which he will complete next year. His future Ph.D. interest is in exoplanets. As for KIC 9832227? "That will be the end of it for me, personally," he said. "Dr. Molnar is far more knowledgeable and experienced, and I trust he knows what he's doing."

Socia's work was recently published in Astrophysical Journal Letters.

Comments


Image of Jim-Baughman

Jim-Baughman

September 14, 2018 at 8:02 pm

It would have been helpful to have more information on what this goof did to the timing of the prediction. Except for one toss-off remark about whether or not these stars will ever merge, this article spends all its time on the machinery of the error; how it happened, reactions from scientists. A little more scientific focus, please!

You must be logged in to post a comment.

Image of robin_astro

robin_astro

September 14, 2018 at 8:46 pm

Seems pretty clear to me.

" He says the adjusted data show the stars might be moving closer together, but he's not sure; at any rate, a merger is by no means imminent."

Or in scientific language - the measured rate of reduction in orbital period is less than originally thought, to the point that it is no longer statistically significant.

You must be logged in to post a comment.

Image of Jim-Baughman

Jim-Baughman

September 14, 2018 at 11:51 pm

You've isolated the "toss-off remark" that I mentioned in my post, which doesn't clarify anything if you read it closely. Your "scientific language" is unfortunately presumptuous. What the scientist said is that he's "not sure" if the stars are moving closer together, which you have expanded, without evidence, to say that orbital decay is "no longer statistically significant".

This is why the internet is so depressing. All opinions quack with equal loudness!

You must be logged in to post a comment.

Image of robin_astro

robin_astro

September 15, 2018 at 8:58 am

No presumption, I just followed up the supplied reference which with a couple of further clicks I find is also available without payment on ArXiv where the authors explain:-

"Our results cast doubt on the prediction that the stars will merge in 2022, as the data do not follow
the exponential decay at early epochs. The O-C does show real changes in period, but
our most favored explanation is that these are merely period wanderings common in
many contact binaries.

The internet is a fantastic tool for the enquiring mind, you just need to know how to use it

You must be logged in to post a comment.

Image of Jim-Baughman

Jim-Baughman

September 15, 2018 at 2:53 pm

"Our most favored explanation" is not the same thing as "reduction in orbital period is ...no longer statistically significant". Anybody can barge through any article and stretch and pull its conclusions to fit a preconceived idea. The point of the article is not to prove anything, but to point out that earlier conclusions were based on faulty data, and declare the issue is once again up in the air.

In the voting booth this fall, pull the R lever extra hard and your machine will record more than one vote.

You must be logged in to post a comment.

Image of Mana1144

Mana1144

November 14, 2018 at 11:52 pm

What i would like to understand is how 1 miss-transcribed data point completely negates the merger hypothesis? Everything indicates this one data point is an outlier so it shouldn't make or brake the merger theory. Furthermore the most recent data since the study began in 2013 should give clear indication of orbital decay should it not? Especially considering that as we get closer and closer to 2022 the stars should be orbiting at much faster rates. Astronomy is not my field so maybe someone with more knowledge can expound on this.

You must be logged in to post a comment.

Image of robin_astro

robin_astro

September 14, 2018 at 8:50 pm

There is even a reference to the paper to follow up. If only other on line popular science writers would do the same

You must be logged in to post a comment.

Image of Mana1144

Mana1144

November 15, 2018 at 12:13 am

From a different site this can be found:

"The resulting variation in brightness allows stargazers to track how fast the stars are merging. Their orbits are accelerating. And that enables the prediction when the stars will finally become close enough to tear each other apart."
"Ongoing observations between 2007 and 2016 demonstrated that a dramatic acceleration is indeed underway."

These stars will definitely merge, the only question is when exactly? I have a hard time believing that 1 data point from 1999 can really impact the timing by years.

You must be logged in to post a comment.

You must be logged in to post a comment.