Euclid found more cosmic lenses in a year than every earlier search combined

Europe’s Euclid space telescope has found more strong gravitational lenses in about a year of sky mapping than every earlier search combined. Its first major data release, published by the European Space Agency, holds roughly fifteen thousand high-confidence lens candidates. In each one, the gravity of a foreground galaxy bends the light of a more distant galaxy into arcs, rings or duplicate images.

That matters because a lens works like a scale. The amount of bending shows how much mass sits in the foreground galaxy, including the dark matter that no telescope can see directly. Until now, astronomers studied these systems one rare find at a time. With a catalogue this large, they can study them as a population, and the same release adds more than nine hundred million galaxies to the wider work of mapping the dark universe.

The release, called DR1-Foundation, covers about 1,900 square degrees of sky, an area roughly 9,500 times that of the full Moon. It contains more than a billion catalogued objects, over 60 million quasar candidates, spectra for more than 160 million sources and about 1.7 million gigabytes of data. It also covers only 14% of the area Euclid is scheduled to survey.

“Strong lensing provides unique insight into the dark Universe, allowing us to ‘weigh’ individual galaxies, including their invisible dark matter, and explore how the Universe is expanding,” said Natalie Lines of the University of Portsmouth, a member of the Euclid Consortium and first author of the lens catalogue paper.

How Euclid picked 15,000 lenses out of 900 million galaxies

Lenses are rare. Only about one galaxy in several thousand lines up with a background source closely enough to act as one, and many lensed images are too small to resolve through the blurring of Earth’s atmosphere. Euclid observes from space with a 1.2-metre mirror and two instruments: VIS, a visible-light camera that produces the sharp images, and NISP, which takes near-infrared images and splits light into spectra.

The search started with the roughly 900 million sources in the release and kept the 68 million that passed a first catalogue cut. Five machine-learning models scored those objects. The most promising went to around 7,000 citizen scientists on the Space Warps project, hosted on the Zooniverse platform, who made more than 4.5 million classifications of 300,000 objects. Finally, 67 lensing experts inspected more than 30,000 systems and graded them.

The grades matter. Of the 15,072 high-confidence candidates, 6,073 are grade A, which the team labels “sure lens,” and 8,999 are grade B, labelled “probable lens.” Another 9,010 grade C “possible” systems are published separately. Software then tried to reproduce each grade A and B system with a lens model, and for more than 9,500 lenses the team estimated distances and the physical properties of the foreground galaxy.

Most of those foreground galaxies are massive and no longer form new stars. They sit at a typical redshift of 0.68, a measure of how much the expansion of the universe has stretched their light, which is farther away than earlier lens samples from ground-based telescopes reached.

From rare finds to a census of thousands

The first strong lens, a quasar that appeared twice in the sky, was identified almost 50 years ago. Since then, the largest ground-based surveys, which together cover most of the sky visible from Earth, had found about 3,000 lenses of this galaxy-behind-galaxy type, according to the Euclid Consortium. Euclid’s 15,072 high-confidence candidates outnumber all previous catalogues combined.

Every one of them was found with the same telescope and the same method. When lenses come from many different surveys and selection methods, they are hard to compare. A single uniform sample lets researchers ask statistical questions, such as how the dark matter around massive galaxies changes from one galaxy to the next, that a few hundred mismatched systems cannot answer.

The scale of the sky coverage is new too. The area in this release is about ten times more sky than the Hubble Space Telescope has observed in 35 years, according to the consortium. Euclid’s earlier releases contained a few days of observations over about 60 square degrees. Valeria Pettorino, ESA’s Euclid project scientist, described the jump as “a change in scale.”

The catalogue also contains rare types that used to be one-off discoveries. It lists 84 candidate double-source-plane lenses, in which one foreground galaxy lenses two background galaxies at different distances, more than 1,000 candidate lenses formed by disc galaxies, and several hundred lensed quasars and active galactic nuclei. Before Euclid, the first two classes had few known examples.

The release holds more than lenses

Other papers published with the data show its range. One team resolved 15 faint stellar systems into individual stars for the first time, five of them new discoveries. One of the new ones, Ursa Minor II, lies about 65,000 light-years away in the Milky Way’s halo and contains only about 40 times the mass of the Sun in stars, which makes it one of the least luminous stellar systems known.

A second study measured the shapes of more than 40 million star-forming galaxies, a sample 440 to 52,800 times larger than comparable studies with the James Webb Space Telescope. Among lower-mass galaxies, it confirmed that elongated, cigar-shaped galaxies become rarer as the universe ages, while flattened discs dominate.

What DR1-Foundation does not settle

Euclid’s main goal is to measure dark energy, the unexplained cause of the universe’s accelerating expansion, and to map dark matter across cosmic history. This release does not do that yet. The measurements needed for cosmology, including weak-lensing maps and galaxy clustering, are still in production and will arrive with the full first data release.

The lenses themselves are still candidates. Only the grade A systems are rated as sure, and the discovery team writes that expert grading remains subjective, so it corrected for the biases of individual inspectors. The search also missed some real lenses. The team estimates it found about 73% of the definite systems in the area, with the rest likely fainter or more ambiguous. Confirming individual systems will need follow-up observations, including spectra.

The scientific papers are preprints. The eight science papers and 18 technical papers released with the data are intended for a special issue of the journal Astronomy & Astrophysics and have not yet completed peer review.

The sky coverage is also partial. At 14% of the planned survey, the sample may not represent the whole sky, and consortium scientists found a few regions whose data quality differs from the rest while they checked the processing.

“This is only 14% of what Euclid ultimately intends to deliver, but we can already see that the promises made more than ten years ago about the span of scientific endeavours Euclid would support are starting to come true,” said Marc Sauvage, an astrophysicist at the French Alternative Energies and Atomic Energy Commission (CEA) and lead of the Euclid Consortium.

Questions about Euclid’s first data release

What is in Euclid’s DR1-Foundation release?

About 13 months of observations covering 1,900 square degrees of sky: roughly 10,000 science images, catalogues of more than a billion objects including over 900 million galaxies, spectra for more than 160 million sources, about 15,000 strong lens candidates and over 60 million quasar candidates. In total it amounts to about 1.7 million gigabytes.

What is a strong gravitational lens?

It forms when a massive foreground galaxy bends the light of a galaxy far behind it, so the background galaxy appears as arcs, a ring or several copies. The amount of bending reveals the total mass of the foreground galaxy, including its dark matter, and the magnification lets astronomers study background galaxies that would otherwise be too faint.

How many galaxies has Euclid mapped so far?

More than 900 million galaxies are catalogued in this release, which covers about 14% of the planned survey area. ESA expects the full mission to catalogue about 1.5 billion galaxies.

Can anyone use Euclid’s data?

Yes. ESA has made the release public and hosts it on its Euclid Data Space platform, where users can explore, visualise and analyse the images, catalogues and spectra without downloading them.

When will Euclid measure dark energy?

Not with this release. ESA and the Euclid Consortium plan the first cosmology results, based on weak lensing and galaxy clustering, for the full first data release in mid-2027.

Cosmology results are due in mid-2027

Euclid launched in July 2023 and began routine survey work on 14 February 2024. The observations in this release were collected between spring 2024 and spring 2025, and ESA published them on 12 November 2026. The full DR1, with the first cosmology results, is due in mid-2027, and DR2, with about three times as much sky, is planned for 2029.

The first strong lens, the Twin Quasar, was identified in 1979. By the end of its six-year survey, Euclid is expected to find more than 100,000 high-confidence lens candidates, each one a natural scale for weighing a galaxy’s dark matter.

Reference: Euclid Collaboration: Lines et al., “Euclid Data Release 1 (DR1). Euclid Lens Identification Pipeline and Scientific Exploration (ECLIPSE) A: Strong lens catalogue overview,” submitted to Astronomy & Astrophysics, 2026. arXiv:2611.15883. DOI: 10.48550/arXiv.2611.15883

Tags: astronomy, Dark Matter, dark energy, esa, euclid, gravitational lensing