That Blast of Radio Waves Produced By Colliding Dead Stars? Not So Fast.

Sometimes science happens quickly. Over the weekend, follow-up observations of a faraway galaxy challenged the conclusions of a study published just last week, which was reported with much fanfare by news outlets that didn’t dig deeply enough to uncover already existing skepticism about the results.

On Feb. 24, astronomers reported in Natureastronomers reported in Nature that a hugely energetic blast of radio waves originated in a galaxy about 6 billion light-years away. These blasts of radio waves, called fast radio bursts (FRBs), have puzzled scientists since the first one was discovered in 2007.

For nearly a decade, one of the major goals in understanding FRBs has been finding a burst’s host galaxy—with that information in hand, scientists would know how far away a burst came from (relatively nearby? Far, far away?), and what kinds of stars live in the burst’s neighborhood. Knowing those things would help identify an astrophysical source for the bursts, which have been thought to arise from colliding dead stars, evaporating primordial black holes, magnetars, and a host of other exotic objects.

So when the Square Kilometer Array’s Evan Keane and colleagues identified a candidate galaxy for a burst observed in April 2015, it was a hugely important observation. The team’s data suggested that FRB 150418 came from 6 billion light-years away—well outside the Milky Way—and went off in a neighborhood filled with old, dying stars.

The team interpreted those data to mean the burst may have been triggered by colliding neutron stars, or the dense, crushed corpses of very large stars. When those stars collided, they not only produced a short burst of gamma-rays, but a blast of radio waves as well. What’s more, by measuring the amount of matter between Earth and the burst’s host galaxy, the observation helped solve what’s known as the “missing mass” problem—a longstanding conundrum in which the amount of observable matter in the universe doesn’t quite match predictions.

It’s a nice, clean story. Trouble is, the association between the fast radio burst and its candidate galaxy may have just fallen apart.

Keane and his colleagues identified the host galaxy based on their observation of “a fading radio afterglow” that overlapped with the galaxy. The team interpreted the fading glow to be a remnant of the collision that generated the FRB, meaning the burst must have originated from within the overlapping galaxy.

Even though the work done by Keane and colleagues is “a technical tour de force,” the link between the galaxy and FRB is uncertain, says Caltech’s Gregg Hallinan, who studies the radio sky and looks for transient events and variables.

“It looks like they may have been unlucky in that the rate of transient radio sources at these frequencies is much higher than you might expect,” he says.

Many things in the sky produce variable amounts of radio waves, sometimes getting brighter, sometimes fading, and it’s not clear how common those sources are or how frequently they vary. Without a good understanding of the variable sky, it’s hard to say whether the “afterglow” really is associated with the FRB, or is something else.

“The most important first step is to check whether the radio afterglow has faded away,” Hallinan said to me, last week. “If it is associated with a short gamma-ray burst [the same event that produced the FRB], it will definitely have faded.”

Disappointingly, very few news stories reported any skepticism about the results or the need for more confirmation.

On Feb. 27 and Feb. 28, Harvard University’s Edo Berger and Peter Williams re-observed the FRB’s candidate galaxy, looking for any sign of the radio afterglow. The team already suspected the afterglow wasn’t the work of the FRB: Even though the initial radio glow had faded, it continued to shine too brightly, and for too long, to make sense. Instead, Berger and Williams (and also Hallinan) suspected the “afterglow” was really an active galactic nucleus, the hugely bright and compact region in the center of the galaxy. These nuclei are common and vary in brightness depending on how active the galaxy’s central black hole is. They also produce detectable glows for long periods of time.

Colliding neutron stars, on the other hand, wouldn’t generate a long-lived afterglow.

In short, if the afterglow appeared in Berger and Williams’ follow-up observations, it couldn’t be associated with the FRB. The team moved quickly, once the Nature paper appeared. “We wrote a proposal the next morning, it was approved in the afternoon, we got the data this weekend, and the rest is history,” Berger says.

Not only did the team find the “afterglow,” Berger and his colleagues also observed that it had brightened. That means associating it with the FRB is problematic; the more likely explanation is that it is produced by an active galactic nucleus that happened to overlap with the FRB’s position on the sky.

If that’s true, then the link between the burst and its galaxy has fizzled, along with any implications for the source of fast radio bursts.

“This recent observation has all but confirmed my fears that the fading radio source seen by Keane et al. is just a variable active galactic nucleus,” Hallinan says. “The claimed association is looking very unlikely at this stage, barring the possibility that the fast radio burst was actually produced by this same active galactic nucleus. There is no strong evidence to support this latter possibility at the moment, so it’s just idle speculation.”

Keane and his team say they’re aware of Berger and Williams’ work, and are working on follow-up observations and a response.

“FRBs certainly are an exciting field with a lot happening,” Keane says. “We are aware of that piece of work, and indeed are performing our own ongoing studies. When we’ve completed and fully considered those we will certainly report our findings. That will be in the peer reviewed scientific literature which, as you know, takes time.”

Behind the scenes, science at its best is a messy enterprise, characterized by challenges and disagreements, conflicting interpretations, and always the demand for cleaner, better data. But that’s how the scientific process works—or should work. It’s just that we don’t often get to see it play out under such a bright spotlight.

This post has been updated.