Independent Reanalysis Reproduces M87 Black Hole Ring, Strengthening the 2019 Image

A four-member team at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Western Australia has taken the raw data behind the first image of a black hole, run it through an entirely new pipeline, and come out with the same ring. Their paper, published in Nature on Nov. 12, describes a reconstruction of the M87 black hole that is nearly identical in size and shape to the picture released in 2019.

That image showed a bright yellow-orange ring around a dark central region, the black hole's shadow. It was built from observations made in April 2017 by radio telescopes at eight sites around the world, including the South Pole, Chile and Antarctica, linked to work as a single Earth-sized antenna. More than 20 institutions took part, among them the Korea Astronomy and Space Science Institute. The target, M87, sits at the center of the Virgo cluster, about 55 million light-years away, and has a mass of roughly 6.5 billion Suns, figures as reported by the Event Horizon Telescope (EHT) collaboration.

The network's thin coverage was both its strength and its weakness. Nithyanandan Thyagarajan, the CSIRO researcher who supervised the work, explained that with so few telescopes scattered across the globe, large gaps open in the measurements, and imaging algorithms must fill them in. Observations at very short radio wavelengths are also prone to atmospheric distortion. That left more room for interpretation than most astronomical data allows. In 2022, a team at Kobe University in Japan reported that applying the EHT's own methods with slightly different assumptions produced no ring at all, only a clump of radio emission.

The CSIRO group attacked the problem in two ways. The EHT processed telescope data in pairs before combining them; the Australians bundled the data in threes, a scheme the researchers say preserves information about the object's structure while canceling errors introduced by the atmosphere or the instruments. They also trained their machine-learning model on images that were not of black holes, so no expectation of what a black hole should look like could steer the output.

Samuel Lai, a postdoctoral fellow at CSIRO and the study's lead author, put the difficulty this way: it is like standing in your backyard and trying to photograph a doughnut on the Moon. He stressed that the model had no idea what a black hole looked like when applied to the M87 data, calling the result the first truly independent validation of the EHT image.

The ring is light from matter whipping around the event horizon at close to the speed of light before being swallowed. The black hole's gravity bends spacetime so severely that light from matter behind it is deflected forward, outlining a shadow that would otherwise stay hidden. One side of the ring is brighter because the disk rotates, and the portion moving toward Earth is boosted in brightness, the same Doppler effect that makes an approaching siren louder. The result supports the EHT's central claim that general relativity, published by Einstein in 1915, accurately describes spacetime around a black hole.

Laurent Loinard, director of the Event Horizon Telescope, welcomed the outcome, calling it extremely gratifying that an elegant, entirely independent method recovers M87's ring structure across several years, in excellent agreement with the collaboration's published results. The multi-year agreement matters because it suggests the ring is a stable feature of the source, not a quirk of one processing run.

The EHT is now expanding its array and targeting more objects, including Sagittarius A*, the supermassive black hole at the center of the Milky Way. Independent verification of the processing pipeline gives researchers more confidence that features in those future images reflect real structure rather than assumptions built into reconstruction software.