Largest 2D map of the universe helps scientists discover new gravitational lenses

Using data from the DESI Legacy Imaging Surveys, with help from artificial intelligence, an international team of scientists has discovered 70 new gravitational lenses—rare cosmic alignments that act as natural magnifying glasses. The new lenses further expand one of the largest collections of confirmed lenses to date and will allow scientists to study dark matter, galaxy evolution and the structure of the universe.

In August 2026, the DESI Legacy Imaging Surveys team released the largest 2D map of the universe ever created. The 5.6-trillion-pixel map contains nearly 4 billion celestial objects, including stars, galaxies, black holes and asteroids.

This map was created using data from three ground-based sky surveys:

The Dark Energy Camera Legacy Survey (DECaLS) was conducted using the 570-megapixel Department of Energy (DOE)–fabricated Dark Energy Camera (DECam), mounted on the U.S. National Science Foundation (NSF) Víctor M. Blanco 4-meter Telescope at NSF Cerro Tololo Inter-American Observatory (CTIO) in Chile, a program of NSF NOIRLab.

The Mayall z-band Legacy Survey (MzLS) was conducted using the NSF Nicholas U. Mayall 4-meter Telescope at NSF Kitt Peak National Observatory (KPNO) in Arizona, a program of NSF NOIRLab.

The Beijing-Arizona Sky Survey (BASS) at the University of Arizona's Steward Observatory was conducted with the UA Bok 2.3-meter Telescope at KPNO and supplemented by years of data from NASA's Wide-field Infrared Survey Explorer (WISE) satellite mission.

From candidates to confirmed lenses

The Legacy Surveys map has been 13 years in the making, and regular data releases from the team have enabled scientists to make countless discoveries across a range of subfields in astronomy and astrophysics. In one such research project, an international team led by Xiaosheng Huang (Santa Clara University and DOE's Lawrence Berkeley National Laboratory) used machine learning to inspect the vast Legacy Surveys data set and identify thousands of new gravitational lens candidates.

In an extensive follow-up study published in The Astrophysical Journal Supplement Series, led by Emerald Lin (an undergraduate student at UC Berkeley) and Ivonne Toro Bertolla (a research assistant at NSF NOIRLab and science operations assistant at Las Campanas Observatory), the team spectroscopically confirmed 70 of the candidates, establishing them as true instances of gravitational lensing.

Cosmic alignments magnify distant objects

A gravitational lens occurs when a massive object such as a galaxy, galaxy cluster or black hole lies directly between Earth and a more distant background object. The foreground object's gravity bends and magnifies the background light, producing arcs, rings and even multiple images of the same object.

These natural telescopes allow astronomers to study objects that would otherwise be too distant or faint to detect. They can magnify galaxies from the early universe, reveal the presence and distribution of invisible dark matter, help improve measurements of the universe's expansion rate and enable the detection of exoplanets—planets orbiting other stars.

AI narrows the search

"The DESI Legacy Imaging Surveys' deep and wide-field images have provided an unprecedented foundation for discovering new gravitational lenses," says Aleksandar Cikota, an associate scientist at NSF NOIRLab, co-author of the study and PI of the follow-up observing program. "What we're building now is a carefully confirmed sample that researchers can use for years to come."

While the Legacy Surveys provided the foundation for discovering these lenses, that foundation was vast. This created both an opportunity and a challenge: Thousands of gravitational lenses were likely waiting to be found, but visually searching for them was impractical. So the team had to devise a robust and efficient process for scouring the data to identify the lenses buried within them.

This is where they turned to artificial intelligence. Using a residual neural network—a form of machine learning designed to recognize subtle patterns in images—the team searched the survey data for the distinctive signatures of gravitational lensing. Earlier work using these techniques had created a catalog containing about 3,500 potential lenses.

The machine-learning approach dramatically accelerated the discovery process, allowing astronomers to focus valuable telescope time on follow-up observations of the most promising candidates.

"The combination of large surveys and AI-powered searches is transforming how we discover rare objects in the sky," says Cikota. "These tools allow us to systematically search for systems that, previously, only a few years ago, we might have discovered only serendipitously."

The initial lens searches also relied on advanced computing resources, including infrastructure provided by the National Energy Research Scientific Computing Center (NERSC), a DOE Office of Science user facility, underscoring how modern astronomy increasingly depends on powerful computation alongside telescopes.

Spectroscopy confirms 70 lenses

The AI-powered search uncovered 76 gravitational lens candidates, but this was only the first step. To determine whether an object is truly a gravitational lens, astronomers must use telescopes to measure the distances to both the lensing and background objects.

To confirm the lens candidates uncovered in the Legacy Surveys, the team used the Multi-Unit Spectroscopic Explorer (MUSE) instrument on the European Southern Observatory's Very Large Telescope. MUSE was particularly handy for this task since it employs a technique called integral field spectroscopy. Unlike a traditional spectrograph, which records light from only a narrow region of the sky, an integral field spectrograph collects a spectrum at every point across an image. The result is a three-dimensional data set containing both spatial and spectral information.

MUSE proved valuable for this study as it allowed the researchers to simultaneously study the lensing and background objects, as well as other nearby objects, in the same observation. The instrument's broad wavelength coverage also enabled the detection of faint spectral features that revealed the distances and physical properties of the objects involved. Using MUSE observations collected between 2022 and 2024, the team confirmed that 70 of the candidates were true gravitational lenses.

A growing resource for cosmic studies

While the newly confirmed lenses are scientifically interesting on their own, their greatest impact may come from what they enable in the future. The newly discovered gravitational lenses have been added to the DESI Strong Lens Foundry Project, one of the largest collections of confirmed lenses to date.

The large, well-characterized lens sample provides an essential foundation for studies of dark matter, galaxy structure and cosmology. Combined with observations from facilities like the Hubble Space Telescope, the James Webb Space Telescope and the recently launched Nancy Grace Roman Space Telescope, as well as with large surveys like NSF–DOE Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST), these lenses will help researchers probe the distribution of dark matter, study distant galaxies and test models of cosmic evolution.

The work also highlights the enduring scientific value of the DESI Legacy Imaging Surveys. What began as a massive imaging campaign to create a 3D map of the universe is now serving as the foundation for a growing catalog of gravitational lenses that will support discoveries for years to come.

"Each confirmed lens adds to a resource that will benefit the entire astronomical community," says Cikota. "Together, we are building a powerful foundation for future discoveries."