A bold claim can now be made in radio astronomy: The most sensitive, widest-area, low-frequency radio map of the Milky Way has been produced, presenting an image of our galaxy well beyond anything previously achieved at these wavelengths. Combining the GaLactic and Extragalactic All-sky MWA survey with its high-resolution successor, GLEAM-X, this achievement was enabled by a sophisticated imaging technique called image domain gridding.
The Murchison Widefield Array, located in the radio-quiet expanse of Western Australia, was the observational backbone. It comprises 4,096 spider-like dipole antennas splayed across several square kilometres, enabling it to grab shots of wide swaths of the sky at one go. In 2013-2015, it mapped the southern hemisphere between 72 and 231 MHz in GLEAM. These observations produced the first “radio colour” view of the Milky Way, in which emissions from different astrophysical processes are differentiated through colour coding: low-frequency emissions from supernova remnants glow orange, while higher-frequency emissions from star-forming regions appear blue.
The 2018 upgrade to the MWA extended its maximum baselines to 5 km, enabling GLEAM-X to deliver far greater resolution and sensitivity. Still, each survey had its limitations: GLEAM captured structures on a large scale but lacked fine detail, while GLEAM-X resolved small-scale features but missed the broader context. Merging them required overcoming a fundamental computational challenge: correcting for ionospheric distortions that shift apparent source positions as radio waves pass through irregularities in Earth’s upper atmosphere.
In this case, gridding in the image domain was essential. Instead of performing expensive convolution kernel evaluations as done in conventional Fourier-domain gridding, image domain gridding split the visibility data into a set of subsets, each with their own low-resolution images, which are then transformed and integrated onto a master grid. This allows for efficient handling of rapidly varying direction-dependent effects, such as ionospheric phase fluctuations, while mapping very well onto massively parallel GPU architectures. It resulted in a seamless mosaic of 95% of the Milky Way visible from the southern hemisphere, stitched together during more than one million hours of processing time at the Pawsey Supercomputing Research Centre.
The scientific potential of this map is enormous. Mantovanini’s work concentrates on supernova remnants-the expanding shells of gas and magnetic fields that result from the explosion of a massive star. Hundreds of them are known, but the new image might host thousands, many beyond the sensitivity of earlier surveys. Colour separation also greatly helps the identification of H II regions, planetary nebulae, and pulsars. By measuring pulsar brightness across GLEAM-X’s frequency range, astronomers can probe emission mechanisms and refine distance estimates.
This work also emphasizes how important advanced calibration strategies are. It was demonstrated in MWA ionospheric studies that low-frequency arrays operate in a regime where planar ionospheric corrections are inadequate; curvature in the phase screens over the array must be accounted for, especially on kilometre-scale baselines. A promising solution in this regard for both imaging and real-time beamforming is the LEAP algorithm, which performs parallel direction-dependent calibration without requiring a full sky model. While this map sets a new benchmark, it is also a preview. The forthcoming SKA-Low telescope with instantaneous frequency coverage of 50–350 MHz, collecting area of 0.4 km² in phase one, and baselines up to 65 km will be thousands of times more sensitive than the MWA.
Its high resolution will require even more precise ionospheric calibration, no doubt featuring GPU-accelerated, station-based direction-dependent solutions similar to LEAP. For now, the combined GLEAM and GLEAM-X mosaic stands as a milestone in low-frequency radio astronomy – an intricate portrait of the Milky Way’s hidden structures, from relics of ancient stellar explosions to the cradles of newborn stars.
