Independent Reanalysis Reproduces M87 Black Hole Ring Shape

Independent researchers reproduced the iconic M87 black hole ring, closing a seven-year debate over whether the original image was an artifact of image processing.

The Tech TL;DR:

• Australian researchers successfully used a new algorithm and machine learning to independently reconstruct the 2017 M87 black hole dataset.
• The resulting ring shape matches the 2019 Event Horizon Telescope image in size and structure, strengthening evidence for general relativity.
• Grouping telescope data in triplets instead of pairs helped eliminate atmospheric and instrumental errors during processing.

A four-member team at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia reprocessed the original Event Horizon Telescope dataset using a newly built imaging algorithm and machine-learning techniques. Published on November 12 in a Nature paper, their recreation of the M87 black hole matches the dimensions and configuration of the 2019 release almost perfectly.

The result shores up the Event Horizon Telescope collaboration’s central claim that general relativity accurately describes the warped spacetime around a black hole. By clearing up this uncertainty, it also resolves a persistent question tied to one of the most widely recognized scientific pictures of the past ten years.

The M87 black hole sits at the center of the Virgo cluster 55 million light-years from Earth and possesses a mass 6.5 billion times that of the sun.

“It is like standing in your backyard and trying to take a photo of a doughnut on the Moon,” Samuel Lai, a postdoctoral fellow at CSIRO and the study’s lead author, said in a statement reported by BigGo Finance.

“Importantly, our model had no idea what a black hole looked like when we applied it to the M87 data. This is the first truly independent validation of the EHT image, an important step in the scientific process.”

Sparse telescope geometry creates measurement gaps

The 2019 picture showed a bright yellow-orange ring surrounding a dark central region, representing the black hole’s shadow against the black of space. Operating as a single planet-sized virtual dish, the setup relied on April 2017 data gathered by radio observatories situated across eight global locations, including Antarctica, Chile, and the South Pole. More than 20 institutions took part, among them the Korea Astronomy and Space Science Institute.

The network’s sparse geometry presented both strengths and weaknesses. With only a handful of stations, large gaps open up in the measurements, requiring software to fill them in. Observations at very short radio wavelengths are also vulnerable to distortion by the atmosphere.

“Because there are only a small number of telescopes scattered around the world, there are large gaps in the measurements that must be filled in by imaging algorithms,” Nithyanandan Thyagarajan, the CSIRO researcher who supervised the work, said. “And since observations are at very short radio wavelengths, the atmosphere can distort the signal.”

That left more room for interpretation than in most astronomical data. In 2022, a team at Kobe University in Japan reported that applying the Event Horizon Telescope’s own analysis methods with slightly different assumptions produced no ring at all, only a clump of radio emission, BigGo Finance noted.

Bundling data in triplets preserves structural information

The Australian team attacked the problem on two fronts. Instead of evaluating telescope data in pairs as the Event Horizon Telescope previously did, the CSIRO group analyzed them in groups of three. The team explained that this method retains vital details about the target’s physical layout while simultaneously wiping out distortions caused by equipment or atmospheric conditions.

They also trained their machine-learning model on images that were not of black holes. This step ensured that expectations about what a black hole should look like could not bias the output.

Laurent Loinard, director of the Event Horizon Telescope, welcomed the confirmation.

Because multiple years of observations yield consistent findings, the ring appears to be a permanent characteristic of the target rather than an accidental artifact of one specific data analysis cycle. The finding carries weight for the next generation of black hole science as the Event Horizon Telescope expands its array and targets additional objects, including Sagittarius A*, the supermassive black hole at the center of the Milky Way.