DESI completes planned 3D map of the universe and continues exploring

Editor’s note: The following press release was published by Lawrence Berkeley National Laboratory.
Fermilab contributed key elements to the Dark Energy Spectroscopic Instrument, which include the online databases used for data acquisition and the software that ensures that each of the 5,000 robotic positioners is precisely pointing to their celestial targets to within a tenth of the width of a human hair.

Other Fermilab contributions include the corrector barrel, hexapod and the testing and packaging of DESI’s charge-coupled devices, or CCDs. The CCDs convert the light passing through the lenses from distant galaxies into digital information that can then be analyzed by the collaboration. Fermilab scientists continue to participate in data-taking and analysis.

Last night, the 5,000 fiber-optic eyes of the Dark Energy Spectroscopic Instrument (DESI) swiveled onto a patch of sky near the Little Dipper. Roughly every 20 minutes, they locked on to distant pinpricks of light, gathering photons that had traveled toward Earth for billions of years. When the sun rose, collaborators marked completion of a major milestone: successfully surveying all of the area in DESI’s originally planned map of the universe.

The five-year survey, finished ahead of schedule and with vastly more data than expected, has produced the largest high-resolution 3D map of the universe ever made. Researchers use that map to explore dark energy, the fundamental ingredient that makes up about 70% of our universe and is driving its accelerating expansion.

By comparing how galaxies clustered in the past with their distribution today, researchers have traced dark energy’s influence over 11 billion years of cosmic history. Surprising results using DESI’s first three years of data hinted that dark energy, once thought to be a “cosmological constant,” might be evolving over time. With the full set of five years of data, researchers will have significantly more information to test whether that hint disappears or grows. If confirmed, it would mark a major shift in how we think about our universe and its potential fate, which hinges on the balance between matter and dark energy.

DESI’s quest to understand dark energy is a global endeavor. The international experiment brings together the expertise of more than 900 researchers (including 300 PhD students) from over 70 institutions. The project is managed by the Department of Energy’s Lawrence Berkeley National Laboratory (Berkeley Lab), and the instrument was constructed and is operated with funding from the DOE Office of Science. DESI is mounted on the U.S. National Science Foundation’s Nicholas U. Mayall 4-meter Telescope at Kitt Peak National Observatory (a program of NSF NOIRLab) in Arizona.

“DESI’s five-year survey has been spectacularly successful,” said Michael Levi, DESI director and a scientist at Berkeley Lab. “The instrument performed better than anticipated. The results have been incredibly exciting. And the size and scope of the map and how quickly we’ve been able to execute is phenomenal. We’re going to celebrate completion of the original survey and then get started on the work of churning through the data, because we’re all curious about what new surprises are waiting for us.”

DESI has now measured cosmological data for six times as many galaxies and quasars as all previous measurements combined. The collaboration will immediately begin processing the completed dataset, with the first dark energy results from DESI’s full five-year survey expected in 2027. In the meantime, DESI scientists continue to analyze the survey’s first three years of data, refining dark energy measurements and producing additional results on the structure and evolution of the universe, with several papers planned later this year.

“The Dark Energy Spectroscopic Instrument has truly exceeded all expectations, delivering an unprecedented 3D map of the universe that will revolutionize our understanding of dark energy,” said Kathy Turner, Program Manager for the Cosmic Frontier in the Office of High Energy Physics at the Department of Energy. “From its inception, we envisioned a project that would push the boundaries of cosmology, and to see it come to such a spectacularly successful completion for its initial survey, ahead of schedule and with such rich data, is incredibly rewarding. The dedication and ingenuity of the entire DESI collaboration have made this world-leading science a reality, and I am immensely proud of the groundbreaking results we are already seeing and the discoveries yet to come as we continue to explore the mysteries of our cosmos.”

An observing machine

DESI began collecting data in May 2021. Since then, the instrument has far surpassed the collaboration’s original goals. The plan was to capture light from 34 million galaxies and quasars (extremely distant yet bright objects with black holes at their cores) over the five-year sky survey. DESI instead observed more than 47 million galaxies and quasars and 20 million stars.

The project’s success is even more impressive in light of several challenges. DESI is a complicated machine with thousands of parts to maintain. In 2020, final tests of the instrument were interrupted by the COVID-19 pandemic. In 2022, the Contreras Fire swept over Kitt Peak but, through the efforts of firefighters and staff, did not damage the telescope. Recovery efforts were slowed by monsoons and mudslides.

“DESI is a complicated but wonderfully robust system, and it’s been a huge amount of fun to see it come together and work so well for such a long time,” said Connie Rockosi, co-instrument scientist for DESI and a professor at UC Santa Cruz and UC Observatories. “We’ve learned about the instrument over five years, and we know its personality and behavior pretty well. That’s important because having the instrument be so efficient is why we’re here at the end of DESI’s original survey with such great data and so much science coming out.”

To map objects, researchers use specially-designed software to optimize DESI observations and decide where to point the telescope. Robotic positioners precisely line up optical fibers that are accurate to within 10 microns, or less than the width of a hair. Ten spectrographs then measure and split the light into its separate colors to determine each object’s position, velocity, and chemical composition. Each night, roughly 80 gigabytes of data streams through ESnet, DOE’s high-speed science network, to supercomputers at Berkeley Lab’s National Energy Research Scientific Computing Center (NERSC). Initial processing lets researchers do quality assurance and make any adjustments needed for the next night of observations.

Collaborators across the project found ways to make DESI more efficient. Efforts spanned telescope operations, tweaks to the instrument hardware, updates to software, observing protocols, methods to reduce the data, and more.

“There’s been constant monitoring and intervention to make the whole thing tick,” said Adam Myers, co-manager for DESI’s survey operations and professor at the University of Wyoming. “And the DESI team is remarkable. This huge group of people have all been working on whether they could save one or two or three percent in their particular area, and when you add it all up, it results in these amazing gains in efficiency.”

DESI is designed to make several overlapping passes of the sky to observe its full footprint (and sometimes make repeated observations of faint objects). The survey was so efficient, the team completed an entire additional pass over the sky for the “Bright-Time Survey,” which is carried out when reflected light from the moon hinders observations of faint and distant objects. All told, DESI made five passes during the Bright-Time Survey and seven during the Dark-Time Survey, covering about two-thirds of the northern night sky.

The sky’s the limit

DESI will continue observations through 2028 and grow its map by about 20%, from 14,000 square degrees to 17,000 square degrees. (For comparison, the moon covers approximately 0.2 square degrees, and the full sky has over 41,000 square degrees). The extended map will cover parts of the sky that are more challenging to observe: areas that are closer to the plane of the Milky Way, where bright nearby stars can make it harder to see more distant objects, or further to the south, where the telescope must account for peering through more of Earth’s atmosphere.

The experiment will also revisit the existing area of the map to collect data from a new set of galaxies: more distant and faint “luminous red galaxies.” These will provide an even denser and more detailed map in the regions DESI has already covered, giving researchers a clearer picture of the universe’s history.

Researchers will also study nearby dwarf galaxies and stellar streams, bands of stars torn from smaller galaxies by the Milky Way’s gravity. The hope is to better understand dark matter, the invisible form of matter that accounts for most of the mass in the universe but has never been directly detected.

The extended map is already underway. When it became clear that DESI would operate beyond its original survey plan, researchers began interleaving the new observations with the ongoing DESI survey to optimize the use of telescope time and keep the instrument from sitting idle.

“We’ve built a remarkable piece of equipment that met all our expectations and then some,” Levi said. “Now we’re pushing beyond our original plan. We don’t know what we’ll find, but we think it’ll be pretty exciting.”

DESI is supported by the DOE Office of Science and by the National Energy Research Scientific Computing Center, a DOE Office of Science national user facility. Additional support for DESI is provided by the U.S. National Science Foundation; the Science and Technology Facilities Council of the United Kingdom; the Gordon and Betty Moore Foundation; the Heising-Simons Foundation; the French Alternative Energies and Atomic Energy Commission (CEA); the Secretariat of Science, Humanities, Technology and Innovation (SECIHTI) of Mexico; the Ministry of Science and Innovation of Spain; and the DESI member institutions.

The DESI collaboration is honored to be permitted to conduct scientific research on I’oligam Du’ag (Kitt Peak), a mountain with particular significance to the Tohono O’odham Nation.

Particle accelerators are some of the most powerful tools that humanity has ever built, supercharging discoveries in physics, chemistry, materials science and biology. The new technologies developed for accelerators have been critical in many modern advancements from producing lifesaving medical isotopes for cancer treatment to breakthroughs in fusion research to eliminating forever chemicals in water.

But particle accelerators are complicated beasts. The most advanced particle accelerators take years to research, design and build, and have tens or hundreds of thousands of devices all working in tandem to deliver a huge variety of particle beams.

“MOAT’s ultimate vision is that we integrate AI so fully into the design, construction and operations of accelerators that we fundamentally transform the pace of discovery and the resulting innovations.”

To help address this complexity, Fermilab is playing a central role in the Multi-Office particle Accelerator Team, known as MOAT, which is creating a unified advanced artificial intelligence system that is incorporated into the entire life cycle of particle accelerators to increase efficiency and innovation.

“MOAT’s ultimate vision is that we integrate AI so fully into the design, construction and operations of accelerators that we fundamentally transform the pace of discovery and the resulting innovations,” said Jonathan Jarvis, MOAT collaborator and director of Fermilab’s Accelerator Research Division.

The U.S. Department of Energy’s Genesis Mission is a historic effort to advance AI and accelerate scientific discovery. MOAT is part of the Transformational AI Models Consortium, or ModCon. Fundamental to the mission, ModCon will develop and deploy self-improving AI models that leverage the DOE’s data, facilities and expertise.

Researchers from DOE national laboratories — including Berkeley, Argonne, Fermilab, Jefferson, Oak Ridge, SLAC and Brookhaven — are working together to develop MOAT.

“There are so many applications of accelerators,” said Jean-Luc Vay, head of the Advanced Modeling Program at Lawrence Berkeley National Laboratory and the MOAT project lead. “They have a really big impact across many fields.”

Fermilab’s accelerator technology test facility, called FAST/IOTA, will serve as a key demonstrator for MOAT’s AI tools. FAST/IOTA also offers flexibility for testing across several different types of accelerators and particle beams.

MOAT’s AI systems are still in the early stages of development, but recently MOAT presented the first demonstration of their work to the DOE Office of Science. This showcase highlighted the team’s initial deployment of the Osprey AI tool, which uses AI agents to accelerate specific tasks by a factor of 100. AI agents are autonomous software systems that can reason, plan and take actions with minimal supervision, and they are a key element of MOAT’s approach and long-term vision.

“Usually each of our labs would develop our own standalone prototype,” said Thorsten Hellert, MOAT collaborator at Berkeley Lab and creator of Osprey. “The Genesis Mission has really compelled our community to work together to develop and deploy this new AI software collectively.”

One immediate path to optimization lies in the decades of knowledge created from running a particle accelerator complex like Fermilab’s. Accelerator operators are the heart of these systems, ensuring that the accelerator is running optimally to deliver particles to experiments. When a team responds to an error in the complex, it is critical that they can search for instances of successful problem-solving passed down from previous operators. MOAT’s AI systems will be trained on all of these documented fixes from Fermilab and other DOE accelerator complexes, providing an immediate solution with citations for where the information was found.

“The Genesis Mission has really compelled our community to work together to develop and deploy this new AI software collectively.”

MOAT will also develop digital twins of each accelerator complex. These will serve as testbeds for virtual diagnostics and speculative beam tuning before any changes are applied. Unlike existing simulations, the virtual twin will be interconnected with the real particle accelerator, allowing for a continuous feedback loop. This enables the AI to learn how the accelerator responds to adjustments and evolve the digital twin to more accurately reflect the performance of the real components inside the accelerator.  

MOAT’s AI will be able to be integrated into the conception and R&D phases of accelerators. Fully realized, MOAT’s vision stands to save billions of dollars, years of effort and dramatically increase the performance and value of particle accelerators.

“The goal is for MOAT to speed up how we can discover and expand our knowledge in fundamental physics, chemistry, biology, materials science, and more, faster than would be possible otherwise,” said Vay. “We hope the resulting research would enable us to multiply the research that can be done, whether it’s for new medication, fusion or one of the other particle accelerator applications.”

The international ICARUS collaboration announced its first physics results on neutrino oscillation searches in a paper recently posted to the preprint server arXiv. They did not observe muon-neutrino disappearance in the data collected with the neutrino beam at Fermilab. The analysis is unique for its rigorous treatment of uncertainties around detector performance. It also demonstrates the quality of the ICARUS data and advances the development of analysis techniques and software tools.

ICARUS is located at the U.S. Department of Energy’s Fermi National Accelerator Laboratory near Chicago and is the world’s first large liquid-argon neutrino detector. It began operating in 2010 at Italy’s Gran Sasso National Laboratory, governed by INFN, the Italian Institute for Nuclear Physics. In 2014, the detector was moved to CERN to be refurbished and improved and, three years later, it journeyed to its new home at Fermilab to become part of the Short Baseline Neutrino (SBN) Program.

The search for neutrino oscillations

The SBN Program consists of three experiments situated along the lab’s neutrino beam: The Short Baseline Near Detector (SBND) is the closest to the neutrino source at just 110 meters away, followed by MicroBooNE at 470 meters, and finally ICARUS at 600 meters. All three SBN experiments study how mysterious, ultra-lightweight particles called neutrinos change as they move through space and matter.

“With ICARUS fully validated and operating we are entering, in concert with SBND, a new era of neutrino physics in which definitive, world‑leading measurements are finally within reach.”

Neutrinos come in three types, called flavors: electron, muon and tau. In a phenomenon called neutrino oscillation, neutrinos change flavor as they travel. To fully understand this behavior, physicists designed baseline experiments — in which particle detectors are placed at various distances along a neutrino beam, like in the SBN Program — to study it.

One specific phenomenon ICARUS is looking for could be evidence of a postulated fourth flavor of neutrino, the sterile neutrino. In the so-called 3+1 model of neutrino behavior, the sterile neutrino could mix with the three known flavors and cause oscillations that appear as muon‑neutrino disappearance over short distances. So, if ICARUS were to observe muon-neutrino disappearance, it would be evidence supporting the 3+1 sterile-neutrino hypothesis; if not, the collaboration could set limits on the model parameters.

First results from ICARUS

In this first analysis, which used data taken from 2022 to 2023, the collaboration did not observe evidence of muon-neutrino disappearance in ICARUS. But the result represents a first important milestone for the SBN Program. The collaboration demonstrated the ICARUS data’s excellent quality and its suitability for physics analyses, as well as the maturity of the software tools for event selection, fitting and detector simulation.

“These first disappearance results mark a major milestone for ICARUS and the broader Short Baseline Neutrino Program at Fermilab,” said Carlo Rubbia, 1984 Physics Nobel laureate and ICARUS spokesperson. “They demonstrate the exceptional performance and stability of the detector and confirm that we now have the precision analysis tools in place to rigorously explore the sterile‑neutrino hypothesis.”

“With ICARUS fully validated and operating we are entering, in concert with SBND, a new era of neutrino physics in which definitive, world‑leading measurements are finally within reach,” said Rubbia.

Preparing for DUNE

ICARUS, as well as SBND and MicroBooNE, is a liquid‑argon time projection chamber detector. Neutrinos are notoriously hard to detect, but when they occasionally interact with argon atoms in a liquid-argon TPC, they cause ionization, creating electrons. A high voltage drifts the electrons to wire planes inside the tank, resulting in a distinctive, precise signal that yields important information about the neutrino interaction and allows for a 3D reconstruction of the particles’ trajectories.

This same technology will be used for the forthcoming Deep Underground Neutrino Experiment, DUNE, led by Fermilab. Currently under construction at the Long Baseline Neutrino Facility at Fermilab in Batavia, Illinois, and in Lead, South Dakota, DUNE will be the world’s most comprehensive neutrino experiment. It will consist of liquid-argon TPC detectors that use technology pioneered by ICARUS — but more than 20 times larger.

The Underground Construction Association has awarded the Long-Baseline Neutrino Facility/Deep Underground Neutrino Experiment at the Sanford Underground Research Facility located in South Dakota, the prestigious 2026 Project of the Year Award. Hosted by the U.S. Department of Energy’s Fermi National Accelerator Laboratory, DUNE is a cutting-edge neutrino experiment comprised of three massive caverns located a mile below the surface. The underground space will house massive detectors and an entire laboratory system dedicated to neutrino research.

“The dedicated engineering teams who designed, excavated and constructed the colossal caverns in South Dakota, completed the project successfully and with an impeccable safety record,” said Fermilab Director Norbert Holtkamp. “Congratulations to the design and construction teams who achieved this important milestone. Construction of a project like this has never been done before in the U.S. They have my deep appreciation as we move to the next phase of making the underground laboratory a reality.”

LBNF won the award for a project in the $100M – $500M category. The construction teams of the LBNF/DUNE project included engineers from Arup, Delve Underground, Fermilab, Kiewit-Alberici Joint Venture, SURF and Thyssen Mining, Inc. Together, they pushed the limits of geotechnical engineering with the formation of two massive caverns, each measuring 65-feet wide, 92-feet tall, and 495-feet long (20 meters x 28 meters x 150 meters). The three caverns of the new research facility span an underground area close to the size of eight soccer fields.

“The dedicated engineering teams who designed, excavated and constructed the colossal caverns in South Dakota, completed the project successfully and with an impeccable safety record.”

For Mike Headley, the executive director of the South Dakota Science and Technology Authority and laboratory director at SURF, this award was made possible thanks to more than two decades of visionary leadership, generous philanthropy and dedicated labor.

“I think this UCA award is tremendous in that it recognizes the monumental scale of this project,” Headley said. “Building something like this on the surface would be challenging enough; it’s so impressive that through the cooperation of multiple partners, we have this enormous accomplishment nearly a mile underground,” Headley said of the project’s safety record; SURF logged more than one million hours during construction without a lost-time incident.

“The UCA Project of the Year Awards are presented to a project team or group that demonstrates insight and understanding of underground construction or a significant project, which may include practices, developing concepts, theories or technologies to overcome unusual problems within a project, resulting in little to no outstanding issues.”

Excavation work at the far site began in early 2019 and was completed in February 2024. During that time, the underground spaces were prepared for the DUNE project with the restoration and expansion of historic rock-handling systems that removed over 800,000 tons of rock from approximately 5,000 feet (1,520 meters) below ground. The rock traveled up the renovated mile-deep Ross shaft at SURF, continuing along an above-ground three-quarter-mile-long conveyor to a large former mining area called the Open Cut. LBNF consists of three long cut-out caverns; two of the caverns will house two far detector modules each, placed end-to-end, while the third will house cryogenics equipment and other utilities that keep the powerful detectors running.

Excavation successfully concluded with the stellar safety record of 1,135,105 hours worked without any lost-time injury.  

DUNE scientists will study the behavior of mysterious particles known as neutrinos to solve some of the biggest questions about our universe. Why is our universe composed of matter? How does an exploding star create a black hole? Are neutrinos connected to dark matter or other undiscovered particles? The project is the largest neutrino collaboration in history and consists of more than 1,500 scientists and engineers from over 35 countries.