Planck Standard
Physics

Euclid Telescope Finds 15,000 Candidate Cosmic Magnifying Glasses, a Record for Dark Matter Hunters

Machine learning, citizen scientists and experts combed Euclid's first major data release for strong gravitational lenses, which were once so rare that fewer than 1,000 were confirmed.

Euclid Telescope Finds 15,000 Candidate Cosmic Magnifying Glasses, a Record for Dark Matter Hunters
Image via Phys.org / University of British Columbia

Researchers working with the European Space Agency's Euclid space telescope have identified about 15,000 new potential strong gravitational lenses, the largest collection of these cosmic tools for studying dark matter ever assembled.

Strong lenses are rare. They form when the gravity of a foreground galaxy bends the light of a more distant galaxy, and in the best cases the background light appears as a ring around the foreground one, called an Einstein ring after Albert Einstein's 1936 prediction. "Strong lenses have been difficult to find. The odds were about 1 in 10,000," said Hrishabh Srivastava, a doctoral student in physics and astronomy at the University of British Columbia. "Before Euclid, we had fewer than 1,000 confirmed strong lenses. Now we have 15,000 new candidates."

The catalog comes from the first major data release from Euclid, which totals 1.7 million gigabytes. A consortium of more than 2,000 scientists, including more than 30 Canadians, found more than 900 million galaxies and 60 million quasars across almost 5% of the night sky, using 13 months of observation. The release is described in 26 papers submitted to the journal Astronomy & Astrophysics and also includes spectra from more than 160 million sources.

Euclid launched in 2023 at a cost of more than $1 billion. Its goal is to map more than one-third of the sky and build a 3D map of the universe that can reveal the properties of dark energy and dark matter. Together they make up about 95% of the universe, yet neither has been directly detected. Dark matter is known only through its gravity, including the bending of light, and dark energy only through the accelerating expansion of the cosmos.

"It turns out the stuff humans are made of is relatively rare in the universe," said Douglas Scott, a UBC physics professor and Euclid collaboration member. "What dominates, dark matter and dark energy, is invisible so we're not used to dealing with it. But it makes the universe what it is." He added that Einstein would have found lenses easy to confirm with Euclid's data, "since they're all over the sky."

Srivastava was part of the team that searched for candidates. The process had four stages. Machine learning first sorted the huge dataset for likely lenses. Thousands of citizen scientists then reviewed what the AI picked. Experts, including Srivastava, checked the volunteers' work. The best candidates will be followed up with other telescopes.

The new catalog is larger than all earlier collections of strong lenses combined. Because one telescope gathered the data and the same methods analyzed it, researchers can compare lenses more reliably than before. Lenses can also be used to test dark energy by checking whether their measured properties fit the standard cosmological model, and they magnify the background galaxy for a clearer view.

Srivastava plans to use the sample to estimate the distribution of dark matter in the universe. He also hopes to build a machine-learning model strong enough to pick out strong lenses without human verification. Euclid operates above Earth's atmosphere, which blurs ground-based images, so it can spot faint, distant features that would otherwise be nearly impossible to detect.

Read next