- Artificial intelligence searched almost 100 million slices from Hubbleidentifying more than 1,300 unusual objects in just a few days – more than 800 of them never before documented in the scientific literature.
- Most anomalies are rare or odd cosmic phenomena – including merging galaxies, gravitational lenses, ring galaxies, and galaxies with unusual shapes or features.
- A new AI tool called AnomalyMatch made discoveries that it allows astronomers to quickly comb through Hubble’s extensive archive and highlight objects that would be extremely time-consuming to search for manually.
AI combs the HST archive
A team of astronomers used a state-of-the-art technique supported by artificial intelligence to detect rare astronomical phenomena within archived data from the NASA-ESA Hubble Space Telescope. The team analyzed nearly 100 million image slices Archive of older Hubbles. These are small cropped portions of much larger images, each measuring just a few tens of pixels (7 to 8 arcseconds) on a side. In just two and a half days, they identified more than 1,300 strange-looking objects – more than 800 of which had never been documented in the scientific literature.
Most of the anomalies were galaxies undergoing mergers or interactions, manifesting in unusual ways morphology or trailing, elongated streams of stars and gas. Others were gravitational lenses, where the foreground galaxy’s gravity warps space-time and bends light from the background galaxy into arcs or rings.
Other discoveries included galaxies with massive star-forming clusters, jellyfish-looking galaxies with gaseous “tentacles,” and planet-forming disks in our own galaxy resembling hamburgers.
Remarkably, several dozen objects completely defied existing classification schemes!
The EarthSky Lunar Calendar 2026 is now available. Get yours today! It’s a great gift.
The enormous challenge of identifying galaxy types
Identifying such a diverse array of rare objects in the vast and growing repository of data from Hubble and other telescopes presents a huge challenge. Never in the history of astronomy has such a volume of observational data been available for analysis.
To solve this problem, researchers David O’Ryan and Pablo Gomez z ESA (European Space Agency) has developed an artificial intelligence tool that can examine millions of astronomical images in a fraction of the time it would take human experts. Their neural network, named AnomalyMatchit was trained to detect rare and unusual objects by recognizing patterns in the data – mimicking the way the human brain processes visual information. David O’Ryan, lead author of the study published in Astronomy & Astrophysics, said:
Archival observations from the Hubble Space Telescope now span 35 years and offer a rich data set in which astrophysical anomalies may be hidden.
Traditionally, anomalous images are discovered by manual inspection or random observation. While experienced astronomers excel at identifying unusual features, the sheer volume of data from Hubble makes comprehensive manual inspection impractical. Citizen science initiatives have helped expand the scope of data analysis. But even these efforts fall short when faced with archives as vast as Hubble’s or those of wide-field survey telescopes such as Euclid, an ESA mission with contributions from NASA.
This is significant progress
The work by O’Ryan and Gómez represents a significant advance. Using AnomalyMatch on the Hubble Legacy Archive, they performed the first systematic search for astrophysical anomalies across the entire data set. After the algorithm flagged likely candidates, the researchers manually checked the top-ranked sources and confirmed more than 1,300 as true anomalies. Gómez commented:
This is a powerful demonstration of how artificial intelligence can improve the scientific return of archival data sets. The discovery of so many previously undocumented anomalies in HST data underscores the instrument’s potential for future explorations.
Bottom Line: Astronomer-led artificial intelligence combed through the Hubble archive—about 100 million Hubble slices—and identified more than 1,300 unusual objects.
Leave a Reply