Cosmology: Citizen Science Project with LMU Participati …

Warps in spacetime do not only show up in science fiction like Star Trek or the film Interstellar. In real life, we can see the warping effect that gravity has on spacetime in the form of gravitational lensing. The enormous gravity of a massive object – such as a galaxy or a cluster of galaxies – distorts the shape of spacetime and can bend the light rays coming from a distant galaxy behind. By warping spacetime, the foreground galaxy acts like a magnifying glass.

Light from the background object that would be obscured doesn’t travel in a straight line anymore. Instead, it curves around the intervening mass, often producing multiple images, stretched arcs, or even a complete ring known as an ‘Einstein ring’, like the one recently discovered by Euclid. Strong gravitational lenses offer a striking demonstration of Einstein’s theory of general relativity, showing that matter in the Universe can act as a natural telescope, bringing distant objects into sight.

ESA’s Euclid telescope is revolutionising the studies of strong gravitational lensing by providing very sensitive imaging over large swaths of the sky in unprecedented detail. This is exactly what is needed to identify rare gravitational lenses.

In March 2025, almost 500 galaxy-galaxy strong lenses were found in the first Quick Data Release (Q1), nestled in just the first 0.04% of Euclid data, most of them previously unknown. This pioneering catalogue was created thanks to the combined effort from citizen scientists, machine learning (ML) – a subset of what is commonly labelled as artificial intelligence (AI) – and researchers.

Early glimpse of new Euclid images

As Euclid continues its survey, sending around 100 GB of data back to Earth every day, ESA and the Euclid Consortium once again need help from citizen scientists to identify strong gravitational lenses in a large data set.

To this end, the Space Warps team has launched a citizen science project based on new Euclid images that will be part of the future Euclid Data Release 1 (DR1). While this data is not public yet, by participating in this new citizen science project, you can get an early glimpse of these new images of galaxies captured by the telescope.

For this project, you will be inspecting new high-quality imaging data from Euclid, in which many previously unknown strong lenses are hiding. About three hundred thousand images pre-selected by AI algorithms will be shown, which are fine-tuned with the results from the initial citizen-science Euclid strong lens search. These are the highest-ranked candidates from a whopping 72 million galaxies in DR1, classified by ML algorithms. Scientists expect that this exquisite, high-quality data will reveal more than 10.000 new lenses.

The current DR1 data set now benefits from that refinement, which helped avoid overlooking a considerable number of potential strong lensing detections.
Leon Ecker, PhD student at LMU

Präzise Vorauswahl gegen die Datenflut

Scientists from LMU and the Max Planck Institute for Extraterrestrial Physics (MPE) in Garching near Munich, Germany, were strongly involved in selecting data that were fed into the Space Warps sample. In order to capture galaxies, Euclid must peer through the Milky Way, our home galaxy. Therefore, multiple other objects, such as stars, nebulae, and other phenomena, must be identified and removed from the catalogue before focusing on the lensing candidates.

“During the first attempt using the Q1 data, this step reduced 29 million objects to only one million that went into closer scrutiny via ML, followed by human inspection,” LMU PhD student Leon Ecker explains. However, Ecker, together with Maximilian Fabricius, Stella Seitz, and Roberto Saglia (all from LMU and MPE), found that the initial selection was too broad. They refined the procedure and identified 27,000 additional candidates, which, in the end, yielded another 72 strong-lensing objects the initial approach missed – 14 percent of the complete sample of almost 600. “The current DR1 data set now benefits from that refinement, which helped avoid overlooking a considerable number of potential strong lensing detections,” Ecker adds.

Welche Erkenntnisse liefern starke Gravitationslinsen?

The Euclid mission explores how the Universe has expanded and how its structure has changed through cosmic history using mainly two methods: weak lensing and baryonic acoustic oscillations. From this, scientists can learn more about the role of gravity and the nature of dark matter and dark energy. Strong gravitational lenses can also provide insights into these central questions. For example, strong lensing features can ‘weigh’ individual galaxies and galaxy clusters. This reveals the total matter (whether dark or light) and traces the distribution of dark matter. By studying strong lenses across cosmic time, scientists can trace the expansion of the Universe and its apparent acceleration. This will provide additional insight into the role of dark energy.

“We’ve already seen the success of combining AI with visual inspection by citizen volunteers and scientists on Space Warps, efficiently finding hundreds of high‑probability lens candidates in an initial small Euclid search in 2024”, explains Aprajita Verma, Space Warps’ co-founder and project lead at the University of Oxford, UK. “In this brand new DR1 data, 30 times larger than the initial search and together with our improved algorithms, we are expecting to find more than 10,000 high-quality lens candidates. This is more than four times the number of lenses that we have been able to find since the first gravitational lens was discovered nearly 50 years ago.”

This step-change is made possible by Euclid. The mission can map large areas of the sky with unprecedented sharpness, an ideal combination for finding rare objects such as strong gravitational lenses.

Additional Informations

Citizen Science: Space Warps