Astronomers have used machine learning to sharpen the Event Horizon Telescope’s first image of a black hole — an exercise that demonstrates the value of artificial intelligence in fine-tuning cosmic observations.
The image is intended to guide scientists in testing their hypotheses about black hole behavior and the gravitational rules of the road under extreme conditions.
The EHT image of the supermassive black hole at the center of an elliptical galaxy called M87, some 55 million light-years from Earth, wowed the scientific world in 2019. The image was created by combining observations from a global array of radio telescopes – but gaps in the data meant the image was incomplete and somewhat blurry.
Remove all ads on Universe today
Join our Patreon for just $3!
Get the ad-free experience for life
In a study published this week in Astrophysical Journal Letters, an international team of astronomers described how they filled in the gaps by analyzing more than 30,000 simulated images of black holes.
“With our new PRIMO machine learning technique, we were able to reach the maximum resolution of the current array,” said the study’s lead author, Lia Medeiros of the Institute for Advanced Study, in a press release.
PRIMO slimmed down and sharpened the EHT’s view of the ring of hot matter swirling around the black hole as it fell into the gravitational singularity. That makes more than just a nicer picture, explained Medeiros.
“Since we can’t study black holes up close, the detail of an image plays a crucial role in our ability to understand its behavior,” she said. “The width of the ring in the image is now about a factor of two smaller, which will impose a severe limitation on our theoretical models and tests of gravity.”
The technique developed by Medeiros and her colleagues — known as principal component interferometry modeling, or PRIMO for short — analyzes large datasets of training images to find the most likely ways to fill in missing data. It’s similar to the way AI researchers used an analysis of Ludwig von Beethoven’s musical works to create a score for the composer’s unfinished 10th symphony.
Tens of thousands of simulated EHT images were fed into the PRIMO model, covering a wide range of structural patterns for the gas swirling into M87’s black hole. The simulations best suited to the available data were merged to create a highly accurate reconstruction of missing data. The resulting image was then reprocessed to match the actual maximum resolution of the EHT.
The researchers say the new image should lead to more accurate determinations of the mass of M87’s black hole and the extent of its event horizon and accretion ring. These determinations, in turn, could lead to more robust tests of alternative theories related to black holes and gravity.
The M87’s sharper image is just the beginning. PRIMO can also be used to sharpen the Event Horizon Telescope’s blurry view of Sagittarius A*, the supermassive black hole at the center of our own Milky Way. And that’s not all: the machine learning techniques used by PRIMO could be applied to much more than just black holes. “This could have important implications for interferometry, which plays a role in fields ranging from exoplanets to medicine,” Medeiros said.
In addition to Medeiros, the authors of “The Image of the M87 Black Hole Reconstructed With PRIMO” in Astrophysical Journal Letters include Dimitrios Psaltis, Tod Lauer and Feryal Özel. The development of the PRIMO algorithm was made possible by support from the National Science Foundation Astronomy and Astrophysics Postdoctoral Fellowship.