New observations of a very unusual and mysterious star located approximately 15,000 light-years away from Earth have revealed a bizarre pattern of stellar activity that astronomers say they've never witnessed before.

The star in question is called Swift J1818.0–1607, and it was only discovered last year. Swift J1818.0–1607 is what's known as a magnetar – a rare breed of neutron star, which forms when supergiant stars fail to go supernova, instead collapsing into incredibly dense cores.

Unlike most neutron stars, however, magnetars are known for producing an extremely powerful magnetic field. Only about 30 of these strange objects have ever been detected in the Milky Way, but even among their weird ilk, Swift J1818.0–1607 is an unusual outlier.

That's because only a handful of known magnetars have ever been observed emitting radio waves of a kind similar to pulsars – another type of neutron star, much more common than magnetars, but remarkable nonetheless for the way in which they beam pulses of radiation from their magnetic poles.

Amidst this exclusive clade of 'radio loud' magnetars, though, none have ever been observed pulsing in quite the same way as Swift J1818.0–1607, leading some in the astronomy community to suggest that it might represent some kind of "missing link" between magnetars and pulsars.

Now, a series of new observations led by astronomers from the ARC Centre of Excellence for Gravitational Wave Discovery (OzGrav) in Australia do nothing to suggest that Swift J1818.0–1607's reputation for weirdness is undeserved.

In eight observations with the CSIRO's Parkes radio telescope over a period of five months in 2020, the researchers observed the magnetar's radio pulses distinctly change in character – resembling a pulsar's pulses in May, then changing to a different form of bright/weak flickering in June, before adopting a mysterious mix of both pulsar-like and magnetar-like radio pulses in July.

010 magnetar 2Artist's impression of Swift J1818.0–1607. (OzGrav/Carl Knox)

In the researchers' new study, they describe this seeming identity crisis of sorts in rather sober scientific terms, saying Swift J1818.0–1607 "[exhibited] significant temporal profile evolution over this period".

But don't let that language fool you into thinking this wasn't an extraordinary thing being witnessed.

"This bizarre behaviour has never been seen before in any other radio-loud magnetar," explains the study's lead author, Marcus Lower from Swinburne University and CSIRO.

"It appears to have only been a short-lived phenomenon as by our next observation it had settled permanently into this new magnetar-like state."

While the mixed messages being relayed by Swift J1818.0–1607 can't fully be explained, the researchers suggest the fluctuations could represent a form of stellar evolution that we don't yet entirely understand.

In part, that's because while this magnetar may be a one of a kind (for now), the truth is magnetars in general, let alone radio-loud magnetars, still represent a very young field of study.

"This raises a number of questions," Lower explained to The Sydney Morning Herald.

"Maybe this magnetar evolved from a more regular pulsar over time… or perhaps we are missing other magnetars in the Milky Way because they're so far away from us that the low-frequency radio waves we're seeing are too scattered for us to detect."

In other words, this seemingly bizarre behavior could be more usual than we know, it's just we're limited in what we currently understand about magnetars. Still, we're finding out more all the time.

The new observations of Swift J1818.0–1607 suggest the magnetic axis of the magnetar isn't aligned with its rotation axis, but instead dips into its southern hemisphere. If so, that's a first for a magnetar – which usually have their magnetic fields aligned with their spin axis.

But it could also explain some of the changes in radio emission profile observed, potentially reflecting radio pulses being emitted at different heights about the surface of the neutron star.

At least one data point, however – an observation called MJD 59062 on August 1 last year – doesn't match that version of events. But the team has a hypothesis for MJD 59062 too.

"Our best geometric model for this date suggests that the radio beam briefly flipped over to a completely different magnetic pole located in the northern hemisphere of the magnetar," Lower says.

The researchers say continued monitoring of Swift J1818.0–1607 will help us to figure out what's really going on here.

"We're looking closely at our data to try to capture one of those flips while it's occurring, because if we can do that we could possibly map out the magnetic field between the magnetic poles," Lower told The Sydney Morning Herald.

"Knowing the actual geometry of magnetars is also quite important in neutron star theory and being able to predict how they will evolve over time."

The findings are reported in Monthly Notices of the Royal Astronomical Society.