In 2020, something near the centre of the Milky Way switched on and off six times in radio waves, then disappeared from every follow-up search in X-rays and infrared — leaving astronomers with only the name ASKAP J173608.2−321635 and the possibility of an entirely new kind of object.

ASKAP J173608.2-321635 is not a name built for memory. It is a coordinate label, the kind astronomers use when an object has not yet earned a more familiar identity.

That is part of the point. The source appeared six times in 2020 in radio observations near the centre of the Milky Way, then refused to show up in the usual places astronomers would look for an explanation. X-ray follow-up found no counterpart. Near-infrared observations found no counterpart. The radio signal itself was variable, highly polarised and difficult to fit into an existing category.

The result is not proof of a new kind of object. It is a well-documented failure of the usual explanations to cover all the evidence.

What ASKAP saw

The main discovery paper was published in The Astrophysical Journal by Ziteng Wang and colleagues in 2021. The team reported the discovery of ASKAP J173608.2-321635 as a highly polarised transient point source using the Australian Square Kilometre Array Pathfinder, or ASKAP.

The source was found in the ASKAP Variables and Slow Transients survey, known as VAST. It was detected six times between January and September 2020 at 888 megahertz, in the Galactic plane and about four degrees from the Galactic Centre. When visible, it showed about 25 percent circular polarisation, a striking feature because circularly polarised radio emission can point toward coherent or strongly magnetised processes.

It was also steep-spectrum and highly variable. In plainer terms, it was brighter at lower radio frequencies, changed strongly over time, and did not behave like a steady background radio source.

That is exactly the kind of object a wide-field radio survey is built to catch. The VAST pilot survey, described by Tara Murphy and colleagues, was designed to search for sources that vary over timescales from seconds to years. ASKAP’s wide field makes it especially useful for finding objects that would be easy to miss if a telescope had to know where and when to look.

The non-detections matter

After the ASKAP detections, the team monitored the field with MeerKAT from November 2020 to February 2021. The source was not detected at first, then appeared on February 7, 2021, reaching a peak flux density of 5.6 millijanskys. It faded quickly, on a timescale of about one day.

MeerKAT added a second important detail. The source remained circularly polarised and also showed up to 80 percent linear polarisation. Its rotation measure, a quantity related to how magnetised plasma twists the polarisation of radio waves, changed significantly over three days.

Then came the absence. Wang and colleagues reported no X-ray counterpart in Swift or Chandra observations about a week after the first MeerKAT detection. They also reported no counterpart in new or archival near-infrared observations down to a J-band magnitude of 20.8.

Those non-detections are not just empty boxes in a table. They remove some ordinary explanations. A flaring star, a nearby brown dwarf, an X-ray binary, a magnetar, a pulsar and an older class called Galactic Centre radio transients each explain part of the behaviour, but none fits comfortably across all wavelengths and all timescales.

Why ordinary labels do not quite work

A low-mass star or brown dwarf can produce radio flares, sometimes highly polarised ones. But the lack of an infrared counterpart is difficult if the object is nearby and stellar. A pulsar could explain polarised radio emission, but searches did not reveal the expected pulses, and scattering toward the inner Galaxy complicates the comparison.

A magnetar is also tempting. Magnetars are strongly magnetised neutron stars, and some can produce unusual radio behaviour. But the lack of X-rays is a problem, because known magnetars are usually much more visible in high-energy observations. The source’s slow, irregular appearance also does not map neatly onto the regular pulses expected from a rotating neutron star.

The closest family resemblance may be with Galactic Centre radio transients, a small group of radio sources discovered in earlier surveys toward the inner Milky Way. Even there, the fit is not exact. ASKAP J173608.2-321635 is located near the Galactic Centre direction, but not at Sagittarius A* itself, and its polarisation and variability give it a particular signature.

That is why the original paper ended cautiously. The authors did not claim discovery of a new class outright. They wrote that the source may represent part of a new class of objects being discovered through radio imaging surveys.

A later radio glimpse sharpened the picture

The object did not vanish from radio astronomy altogether. In a 2024 preprint, Kierra Weatherhead and colleagues reported spectrum and polarisation measurements from THOR-GC and VLITE, based on serendipitous observations of the same transient at three epochs in March 2020, April 2020 and February 2021.

That work found the source only on April 11, 2020, in those data, with a flux density of 20.6 millijanskys at 1.23 gigahertz. VLITE also detected it at 339 megahertz. The authors inferred a spectral break below one gigahertz and a rotation-measure range larger than previously reported.

Their possible interpretation was not an exotic new particle or a signal from outside astrophysics. They suggested the data could be consistent with a highly supersonic neutron star interacting with a changing environment. That is still a possibility, not an identification.

It is an important shift in emphasis. The longer the source remains constrained only by intermittent radio detections and non-detections elsewhere, the more the problem becomes environmental as well as intrinsic. Astronomers need to know not only what object is emitting, but what magnetised material the radio waves are passing through.

Why radio surveys are finding these objects now

ASKAP J173608.2-321635 belongs to a broader change in astronomy. Wide-field radio surveys are starting to find objects that do not announce themselves in optical catalogues, X-ray archives or infrared images. Some may turn out to be unusual versions of known systems. Others may require a new category.

The safest reading is that ASKAP J173608.2-321635 remains unclassified. Its six 2020 ASKAP detections, February 2021 MeerKAT flare, strong polarisation, changing rotation measure and missing X-ray and infrared counterparts are real observational constraints. The label “new kind of object” is still a possibility, not a settled result.

The next step is also plain. The source needs to switch on again while radio, X-ray, infrared and perhaps optical facilities are watching at the same time. Until then, it remains what its name implies: a position on the sky, a set of measurements, and a gap in the catalogue.