The Millisecond Signals From the Edge of the Universe
In July 2001 the Parkes radio telescope in New South Wales, Australia, swept a patch of sky near the Small Magellanic Cloud, recorded a few minutes of ordinary looking data, and filed it away. Six years passed before anyone realised what was in the file. In 2007 Duncan Lorimer, an astronomer at West Virginia University, handed those archived recordings to a student named David Narkevic and asked him to comb them for pulses. Buried in the numbers was a single spike: a burst of radio energy lasting less than five milliseconds, bright enough for that instant to outshine almost everything else in the sky, and then nothing. It had happened once, on 24 July 2001, and never came back.
The spike is now known as the Lorimer Burst, and it was the first fast radio burst, or FRB, ever identified. What made it extraordinary was not only its brightness but a fingerprint hidden in its shape. Radio waves of different frequencies do not cross the universe at exactly the same speed. When a sharp pulse travels through the thin plasma of interstellar and intergalactic space, its higher frequencies arrive slightly ahead of its lower ones, and the pulse is smeared into a rapid downward sweep. The size of that smearing is called the dispersion measure, and it works like an odometer: the more space a signal has crossed, the more free electrons it has met, and the wider its frequencies spread. The Lorimer Burst's dispersion measure was about ten times larger than our own galaxy could account for along that line of sight. Read at face value, the burst had come from far outside the Milky Way, perhaps billions of light years away, releasing in thousandths of a second as much energy as the Sun pours out in days.
For several years almost nobody believed it. A single unrepeatable flash in old data is the kind of signal that usually turns out to be a glitch, a satellite or a faulty receiver, and the doubt hardened when Parkes began turning up other short bursts that looked similar but behaved wrongly. Astronomers nicknamed them perytons, after a mythical winged creature. Perytons showed up in many of the telescope's beams at once, which no distant point source can do, and they clustered around midday and weekends. They were plainly local and man made, and their existence poisoned the well. If the telescope could manufacture convincing fake bursts, there was no obvious reason to trust the Lorimer Burst either.
The perytons were run to ground in 2015. A team led by Emily Petroff installed new monitoring equipment at Parkes and waited for the interference to return. When fresh perytons appeared, they came with a companion signal at 2.4 gigahertz, a frequency familiar to anyone who has stood in a kitchen. The source was the observatory's own microwave ovens. Staff impatient for lunch were pulling the doors open before the timer finished, and an oven interrupted in mid cycle releases a brief chirp from its magnetron as it shuts down. Caught by one of the most sensitive telescopes on Earth, that chirp had been masquerading as a signal from deep space. The episode reads like a joke, but it mattered. Once the perytons were separated out, astronomers could show that the remaining bursts behaved nothing like kitchen appliances. The field was real.
The bursts became a central problem of astrophysics when it turned out that some of them repeat. In 2012 the Arecibo telescope in Puerto Rico recorded a burst catalogued as FRB 121102. In 2015 observers caught it flashing again from precisely the same point on the sky, and then again, sometimes several times within an hour. That single fact narrowed the field. A one off flash could be a catastrophe, an event that happens once and is finished. A source that bursts over and over cannot be a one time destruction. Whatever FRB 121102 is, it survives its own explosions.
Because it repeated, telescopes could be trained on it long enough to fix its position, and in 2017 the trail led to an address. FRB 121102 sits in a small, dim, irregular dwarf galaxy roughly three billion light years away, inside a bright knot of active star formation. Directly on top of the burst source astronomers found a persistent glow of radio emission, a faint but continuous hum from the same patch of sky. It was the first FRB placed in a known galaxy, and the neighbourhood was informative. Dwarf galaxies crowded with young stars are exactly where the universe makes its most extreme magnetic objects.
Those objects are called magnetars. A magnetar is a neutron star, the collapsed core left behind when a massive star dies, packing more than the mass of the Sun into a sphere the size of a city, with a magnetic field trillions of times stronger than Earth's, violent enough in principle to wipe a credit card at half the distance to the Moon. Young magnetars are unstable. They flare and convulse as that magnetic energy is released, and they are well known sources of X-rays and gamma rays.
On 28 April 2020 one of them was caught in the act inside our own galaxy. SGR 1935+2154 is a catalogued Milky Way magnetar about thirty thousand light years away. On that day it threw off a burst of radio waves so powerful that two very different instruments recorded it at once: the CHIME telescope in British Columbia, Canada, and a modest array of small antennas in California called STARE2. Scaled for distance, the burst had the profile of a genuine fast radio burst, the first ever detected inside the Milky Way. It arrived together with a flash of X-rays from the same magnetar, picked up by orbiting observatories, so there was no doubt which object had fired.
Two further findings reshaped the picture. One repeating source, FRB 180916, keeps a calendar: its bursts cluster in a window that comes round roughly every sixteen days, active for a few days, silent for the rest, then active again. The other is sheer volume. CHIME watches a huge strip of sky every night, and it turned the trickle of known bursts into a flood, cataloguing first hundreds and then thousands. Extrapolated across the whole sky, the numbers imply that fast radio bursts are going off thousands of times a day.
The bursts have also become useful before being explained. The dispersion measure that once looked like a nuisance is a running tally of all the ordinary matter a signal has passed through, and for years cosmologists could not find enough of that matter in the nearby universe. Theory predicted atoms that telescopes could not see, hidden as thin cold gas between the galaxies. By combining the distance of localised bursts with the smearing of their signals, astronomers weighed that gas and closed the account on the missing ordinary matter of the cosmos.
Conclusions and Open Questions
The strongest conclusion available is that magnetars can produce fast radio bursts. That case rests on the localisation of FRB 121102 to a star forming knot in a dwarf galaxy, the kind of place where young magnetars are born, and above all on the April 2020 detection from SGR 1935+2154. Most working astronomers now treat some version of the magnetar model, pressed after the 2017 localisation and vindicated in 2020, as the default explanation.
The weakness of that model is scale. The galactic burst from SGR 1935+2154 was still thousands of times fainter than the bursts going off billions of light years away. If ordinary magnetars make the weak ones, it is not clear what makes the titanic ones. Nor is it settled where the radio waves are generated. One camp argues the emission comes from close in, inside the magnetar's magnetosphere; another holds that it is produced far outside, where an outflow slams into surrounding material. Both models can be tuned to fit, and neither has been confirmed.
A second open question is whether FRBs are one phenomenon at all. Some sources repeat; most have flashed once and gone silent. One theory holds that these are the same objects seen from different angles or at different stages of life, another that they are two populations with two separate causes. Repeating sources do tend to produce broader and differently shaped pulses, which argues for a real distinction. FRB may yet turn out to be a word like explosion, a label for a violent event with several unrelated causes.
The sixteen day cycle of FRB 180916 has no accepted explanation. Some argue the source orbits a companion star whose wind periodically clears a path for the signal to escape. Others propose that the magnetar is slowly precessing, wobbling like a dying top and sweeping its beam past Earth on a schedule. A magnetar spinning every few seconds has no obvious reason to keep a sixteen day calendar, and both proposals require geometry that has never been observed directly.
The question readers always raise deserves a straight answer. Could fast radio bursts be artificial? A few researchers have published versions of the idea, most prominently a 2017 paper by the astrophysicists Manasvi Lingam and Avi Loeb, which proposed that the bursts are leakage from enormous beams used to push light sail spacecraft between the stars. It is not a forbidden hypothesis and it makes testable predictions. Its weakness is decisive: the bursts come from all over the sky and from many different galaxies, they match natural neutron stars under extreme stress better with every year of data, and invoking intelligence explains nothing that magnetars do not already explain.
What remains genuinely unexplained is the mechanism itself. Nobody has a complete account of how any object converts stored magnetic energy into a coherent radio pulse of that power in a thousandth of a second. The open questions are these: how many kinds of fast radio burst exist, what sets the periodic sources apart, and whether the physics behind the brightest bursts is in the textbooks or has yet to be written. The first burst waited six years in an archive before anyone looked. It is worth asking how many more are still unread.