Scientists spot a 'kilonova' flash so bright they can barely explain it

Scientists may have caught the blinding flash of two dense neutron stars colliding to form a strange magnetic star.

The first sign of the massive event was a gamma-ray beacon that appeared in telescope data on May 22, prompting astronomers to assemble their best instruments. That response was important: Scientists believe gamma-ray bursts usually stem from neutron stars colliding so they are eager to see as many views of such fireworks as possible. But as observations came in, researchers realized there was something strange going on: The flash included far more infrared light than predicted, 10 times more. The scientists behind the new research think that discrepancy may mean the crash produced something unexpected.

"These observations do not fit traditional explanations for short gamma-ray bursts," Wen-fai Fong, an astronomer at Northwestern University in Illinois and lead author on the new research, said in a statement. "Given what we know about the radio and X-rays from this blast, it just doesn't match up."

Related: Gamma-ray universe: Photos by NASA's Fermi Space Telescope

An artist's depiction of a brief gamma-ray burst that was 10 times brighter than the next brightest such event.of a brief gamma-ray burst that was 10 times brighter than the next brightest such event.

An artist's depiction of a brief gamma-ray burst that was 10 times brighter than the next brightest such event.of a brief gamma-ray burst that was 10 times brighter than the next brightest such event. (Image credit: D. Player/STScI/NASA/ESA)

Astronomers used a host of facilities to study the event, including NASA's Swift Observatory in space, the Very Large Array in New Mexico and the Keck Observatory in Hawaii, but it was the Hubble Space Telescope that spotted the extremely bright infrared radiation from the burst that told scientists something particularly strange was going on.

"The Hubble observations were designed to search for infrared emission that results from the creation of heavy elements — like gold, platinum, and uranium — during a neutron-star collision," Edo Berger, an astronomer at the Center for Astronomy jointly run by Harvard University and the Smithsonian Institution and co-author on the new research, said in the statement. Neutron stars are the superdense remains of exploded stars and the bright afterglow from a collision of two such objects is called a kilonova.

"Surprisingly, we found much brighter infrared emission than we ever expected, suggesting that there was additional energy input from a magnetar that was the remnant of the merger," Berger said. "The fact that we see this infrared emission, and that it is so bright shows that short gamma-ray bursts indeed form from neutron star collisions, but surprisingly the aftermath of the collision may not be a black hole, but rather likely a magnetar."

A magnetar is a cosmic curiosity, an unusual class of supermagnetic neutron stars. But scientists have long wondered how magnetars become so magnetic, so observing a possible formation event is particularly valuable for scientists.

"We know that magnetars exist because we see them in our galaxy," Fong said in a second statement. "We think most of them are formed in the explosive deaths of massive stars, leaving these highly magnetized neutron stars behind. However, it is possible that a small fraction form in neutron star mergers. We have never seen evidence of that before, let alone in infrared light, making this discovery special."

And this time, researchers were able to catch an early enough view of the blast to catch the fading infrared peak in all its glory.

"Amazingly, Hubble was able to take an image only three days after the burst," Fong said. "You need another observation to prove that there is a fading counterpart associated with the merger, as opposed to a static source. When Hubble looked again at 16 days and 55 days, we knew we had not only nabbed the fading source, but that we had also discovered something very unusual."

The research is described in a paper announced for publication in The Astrophysical Journal today (Nov. 12) and available to read on the preprint server arXiv.org.

Email Meghan Bartels at mbartels@space.com or follow her on Twitter @meghanbartels. Follow us on Twitter @Spacedotcom and on Facebook.

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: community@space.com.

Meghan Bartels
Senior Writer

Meghan is a senior writer at Space.com and has more than five years' experience as a science journalist based in New York City. She joined Space.com in July 2018, with previous writing published in outlets including Newsweek and Audubon. Meghan earned an MA in science journalism from New York University and a BA in classics from Georgetown University, and in her free time she enjoys reading and visiting museums. Follow her on Twitter at @meghanbartels.

  • rod
    The arXiv paper cited says "The NIR counterpart, revealed by our HST observations at a rest-frame time of ≈2.3 days, has a luminosity of ≈(1.3−1.7)×10^42 erg s−1. This is substantially lower than on-axis short GRB afterglow detections, but is a factor of ≈8-17 more luminous than the kilonova of GW170817, and significantly more luminous than any kilonova candidate for which comparable observations exist.", https://arxiv.org/abs/2008.08593
    Considering some 10^42 erg s^-1 released, that is still some blast. I am glad the Earth is not orbiting in this location :)
    Reply
  • samiup
    rod said:
    The arXiv paper cited says "The NIR counterpart, revealed by our HST observations at a rest-frame time of ≈2.3 days, has a luminosity of ≈(1.3−1.7)×10^42 erg s−1. This is substantially lower than on-axis short GRB afterglow detections, but is a factor of ≈8-17 more luminous than the kilonova of GW170817, and significantly more luminous than any kilonova candidate for which comparable observations exist.", https://arxiv.org/abs/2008.08593
    Considering some 10^42 erg s^-1 released, that is still some blast. I am glad the Earth is not orbiting in this location :)

    That's nothing...
    Back in my days, 10^77 erg s−1 was pretty common.
    Reply
  • Torbjorn Larsson
    Nifty video, explaining the artist's conception, here: https://www.sciencealert.com/we-may-have-just-witnessed-the-birth-of-a-magnetar-from-colliding-neutron-stars .

    rod said:
    I am glad the Earth is not orbiting in this location :)

    :)

    A similar context here:

    An international team of researchers led by Dr. Diederik Kruijssen at the Center for Astronomy at the University of Heidelberg (ZAH) and Dr. Joel Pfeffer at Liverpool John Moores University has now managed to infer the Milky Way's merger history and reconstruct its family tree, using only its globular clusters.

    https://phys.org/news/2020-11-family-tree-milky-deciphered.html ]
    From the paper Conclusions:

    Focusing on the assembly history of the Milky Way, our results add to a growing body of evidence that the Milky Way experienced an unusual path to adolescence. Not only did it assemble unusually quickly for its mass, but it also experienced a striking paucity of major accretion events, with only a handful of minor mergers shaping the Galactic stellar halo. The fact that it grew most of its stellar mass through secular processes and in-situ star formation implies that it may not be the most representative example for understanding the evolution and assembly of the galaxy population, but is a correspondingly more pleasant environment to live in.

    By the way, I recommend reading the somewhat technical abstract and then looking at the more detailed figure 9 of the merger tree to get a quick look-see on the putative merger history. Paper here: https://arxiv.org/pdf/2003.01119.pdf .
    Reply
  • Torbjorn Larsson
    FWIW, this is just in on an interesting application of kilonova research to improve the method of multimesssenger observation of the universe expansion rate:

    We’re not quite sure how old we are — cosmologically, that is. The main methods scientists use to measure the age of the universe don’t agree with each other. Many physicists hope a newly applicable technique that incorporates gravitational-wave observations will solve this age discrepancy once and for all.

    But this new technique may not be as straightforward as researchers hoped. A new paper by Hsin-Yu Chen, a postdoc at the MIT Kavli Institute for Astrophysics and Space Research, describes a potential problem ...

    Using telescopes, it is easy to measure how fast these galaxies move away from us, but it's difficult to measure their distance. But when observing gravitational-waves signals, it is the other way around: We can measure distance directly from the gravitational-wave observation, but it’s hard to measure how fast these gravitational-wave sources are flying away from us. For that part, we need help from traditional telescopes; we need to capture the light produced by the gravitational-wave sources.

    To measure the age of the universe in this newer way, you need these two components: gravitational waves and light. It is actually a very straightforward method that astronomers consider rather clean.

    If we don't know what the geometry of the emission is, then we might preferentially observe one specific viewing angle and lead to a bias in our measurement.

    ... the gravitational waves give us some idea of the viewing angle of the sources we observe. By combining observation of the kilonova with this viewing angle constraint from the gravitational-wave side, it’s possible that after many, many observations we could figure out whether there is a bias or not.

    https://news.mit.edu/2020/3-questions-hsin-yu-chen-treading-lightly-when-dating-universe-1113 ]
    It's not all that bad. The paper abstract discuss systematic biases that are 2-3 % or so, much less than the current 9 % tension between expansion rate measurements.
    Reply
  • rod
    Torbjorn, interesting posts and info. I have a number of reports showing H0 ranging from recent measurements, 67.9 to 82 km/s/Mpc which changes the Hubble time calculation too.
    Reply