New Observations Challenge Popular Radio Burst Model

Fourteen years ago, the first fast radio burst (FRB) was discovered. By now, many hundreds of these energetic, millisecond-duration bursts from deep space have been detected (most of them by the CHIME radio observatory in British Columbia, Canada), but astronomers still struggle to explain their enigmatic properties. A new publication in this week’s Nature “adds a new piece to the puzzle,” says Victoria Kaspi (McGill University, Canada). “In this field of research, surprising twists are almost as common as new results.”

Most astronomers agree that FRBs are probably explosions on the surfaces of highly magnetized neutron stars (so-called magnetars). But it’s unclear why most FRBs appear to be one-off events, while others flare repeatedly. In some cases, these repeating bursts show signs of periodicity, and scientists had come up with an attractive model to explain this behavior, involving stellar winds in binary systems.

However, new observations by European radio telescopes may rule out this model.

Astronomers knew that FRB 20180916B, located in a galaxy some 460 million light-years away, produces multiple bursts about every 16 days, during a ‘window’ that lasts for a few days. “The idea was that the magnetar is part of a binary system with a 16.29-day period,” says Inés Pastor-Marazuela (University of Amsterdam and ASTRON, The Netherlands), the first author of the new paper. If the companion star had a thick stellar wind that absorbs radio waves, the bursts would only be visible when the magnetar was on ‘our’ side of the orbit, she explains.

However, simultaneous observations of FRB 20180916B by the Low-Frequency Array (LOFAR) and the 14-dish Westerbork Synthesis Radio Telescope (WSRT) in the Netherlands challenge the predictions of this model. Since stellar winds should better absorb lower-frequency radio waves than higher-frequency ones, astronomers expected that the bursts observed by LOFAR (down to 110 megahertz) would only be visible in a narrower time window than the bursts observed by WSRT (at around 1.4 gigahertz). “We found the exact opposite,” says coauthor Joeri van Leeuwen (University of Amsterdam and ASTRON, The Netherlands). Moreover, the peak in the number of high-frequency bursts preceded the low-frequency peak by a few days, which also isn’t expected in the binary wind model.

“I agree that the observations are challenging for the model,” says Kaspi, who’s part of a team that has independently studied the LOFAR data (which are publicly available) but didn’t have access to the simultaneous Westerbork observations. However, she’s not yet convinced that the binary idea is completely ruled out. “We need more sources and better statistics.”

What could be an alternative explanation? Perhaps, says Pastor-Marazuela, the 16.29-day period is actually the rotation period of the burst source, instead of its orbital period. If the explosions originate in a small, localized region of the magnetar’s surface, this region will be carried in and out of sight by the compact object’s rotation.

Kaspi counters that a rotation period of 16.29 days would be incredibly long: Magnetars (and neutron stars in general) usually complete tens, hundreds, or even a few thousand revolutions per minute. “But nature can be very creative,” she adds. “Never say never.”

FRB 20180916B could be a very unusual case, says van Leeuwen. In particular, he is surprised by the fact that no single burst was detected by both LOFAR and WSRT, even though the two facilities were observing simultaneously. “It’s something I had never expected,” he says. But even if this particular source is special, it could shed more light on the properties of FRBs in general. “Think of Oliver Sacks, the famous neurologist,” van Leeuwen says. “He learned a lot about the human brain by studying his most interesting patients.”