![](https://media.wired.com/photos/60354ee440867c4bdef7c684/191:100/w_1280,c_limit/Science_SnowCloud-4K.jpg)
“The reason that rapid variability is interesting is because it’s usually a sign of something extreme going on,” says Murphy. Detecting extreme events can mean spotting hidden supernovae or catching nearby stars releasing flares so large that they wipe out any potential for life on planets in their orbit. This fast variability is hard to observe, however, since it requires a radio source to be far away (small in our field of view) and for whatever is obstructing it to be large and close to home.
In 2019, Murphy worked on an unrelated investigation of the radio wave aftermath of the merging of two neutron stars. The team used ASKAP to scan a tract of the cosmos nine and 33 days after the merger. But after that analysis ended, the data remained a treasure trove for further analysis of variations in the night sky. “We got like 30,000 galaxies—30,000 radio sources—in that field. So I had to deal with lots of data,” Wang says.
Wang wanted to find the most capricious radio signals in the sky. She wrote a script to weed out the data from stagnant radio blips they didn’t care about, but was still left with thousands of radio sources that appeared to be varying. The vast majority had uninteresting explanations, or were artifacts of the detection process. Still, Wang scrutinized each one. “So I just click, click, click, click for several days,” says Wang, “and eventually,I found it.”
Of the 30,000 distant galaxies, only six were actually scintillating rapidly. “Of those six, five were in a dead straight line,” says Murphy. “When you discover something like that, you think there’s something strange going on here.”
To Wang and Murphy, something strange also meant that there might be something wrong. Their team had to confirm that their result wasn’t just some weird one-off. They reimaged the sky from a different angle so the interesting feature appeared elsewhere, ruling out unreliable pixels. But in the end, they couldn’t blame it on telescope misbehavior. “So then you’re left with the idea that this must be something astronomical,” says Murphy. “It must be real.”
Encouraged, Wang and Murphy collected more snapshots of the scintillating signals over 11 months—seven nights of observation in all. That timespan let them tease out the size and shape of what they believe to be an interfering gas cloud, as the backlights shifted in relation to Earth, the first example of such an approach. Their results show that the filament of gas is thin and about a third of one light-year long—20,000 times longer than Earth’s distance from the Sun.
How did this weird cloud form? Murphy’s team can’t know for sure, but they think a star’s immense gravity shredded a gas cloud into these proportions. Black holes are known to create these gas streams, but none are nearby. “So rather than a black hole,” Murphy says, “we have some kind of plasma cloud that’s been disrupted by a star and stretched it out so that we have this long tidal stream.”
One aspect of the cloud stumped Murphy’s team. She says only warm charged gas, plasma, could cause the twinkling. But based on her team’s models, they think the cloud could form its shape only by moving fast—about 30 kilometers per second—and that means that a larger portion of it would actually be very cold. So cold, in fact, that hydrogen droplets inside could freeze like snowflakes.
Françoise Combes, a Collège de France astrophysicist not involved with the work, is sold on the team’s find. In fact, Combes’ own work two decades ago hypothesized that not only do cold clouds exist but also that they make up a large portion of the Milky Way’s missing baryons. She thinks this cloud is likely just the small tip of a much larger fractal cloud structure throughout the galactic disk. “Scintillations are the signature of the existence of this hierarchy of molecular cloud scales,” she wrote in an email to WIRED. “There is plenty of space to have a large fraction of dark baryons under the form of cold molecular clouds.”