A project more than two decades in the making, the NSF–DOE Vera C. Rubin Observatory is now live and has revealed its first images of the southern sky.
Looking up at the night sky, people can usually see the Moon, some stars and constellations, and even a bright planet or two through urban light pollution. In a more remote area, they may even see the arm of our home Milky Way Galaxy sprawling above. For thousands of years, humanity has lived under and looked up at these same celestial sights, feeling the tug of curiosity into what makes up the universe beyond. Scientists have long grappled with and attempted to answer this question. Telescopes have offered improved views of celestial objects near and far, quelling some curiosity but only uncovering more questions as they reveal more of the universe.
As of 2025, the most promising glimpse into the cosmos has come alive, with the potential to answer a slew of humanity’s questions (and, surely, to bring many more questions to light). Situated atop Cerro Pachón in the Andes Mountains of the Coquimbo Region in Chile, NSF–DOE Vera C. Rubin Observatory, funded by the U.S. National Science Foundation (NSF) and the Department of Energy’s (DOE’s) Office of Science, has achieved first-light commissioning images—the culmination of more than two decades of scientific effort and multi-institutional collaboration.
The NSF–DOE Vera C. Rubin Observatory—or “Rubin,” for short—is named for an American astronomer whose observations, along with those of fellow astronomer Kent Ford, forged convincing arguments for the presence of dark matter, opening new fields of study in astrophysics. The observatory, originally intended to be a step toward understanding dark energy and dark matter, will provide the most comprehensive survey of the southern sky ever conducted. (See S&TR, September 2017, Unlocking the History of the Solar System.) Its unique capabilities, including a novel mirror design, the world’s largest digital camera, and unmatched telescope speed, equip Rubin to efficiently accomplish this task. Throughout the next 10 years, the observatory will continuously take images of different parts of the sky with 30-second exposures, stitching together a picture of a quarter of the night sky every three days. Over the course of a year, the rotation of Earth enables the observatory to create a detailed movie of the entire southern sky in a survey called the Legacy Survey of Space and Time (LSST). This survey will provide humanity with literal astronomical volumes of data—about 20 terabytes per night—which will be made publicly available and will take decades to unravel.
Rubin Observatory is operated by NSF’s NOIRLab (formally named the National Optical-Infrared Astronomy Research Laboratory) and DOE’s SLAC National Accelerator Laboratory, and its scientific contributions since inception have come from institutions around the world. Lawrence Livermore has been involved since the beginning with contributions to Rubin’s optics, camera, and science missions, and Laboratory scientists are now reaping the benefits and exploring just some of the questions Rubin can answer, untangling the growing treasure trove of information the survey provides.
Involved History
Livermore’s involvement in LSST and Rubin began in the 1990s with the late physicist Kem Cook. His influence was twofold, starting with his status as the head of the Astronomy and Astrophysics group embedded within the University of California-wide Institute for Geophysics and Planetary Physics, with which he worked on the Massive Compact Halo Object (MACHO) dark matter search project in the nineties. (See S&TR, April 1996, MACHO: Collaboration Is Key to Success to Success.) The search for MACHOs—proposed components of dark matter that are made of standard matter but are too dim to be seen—required the highest-resolution large format cameras at the time. These cameras would survey stars at the center of the Milky Way Galaxy and nearby dwarf galaxies in search of brightening or dimming stars, an indicator that a MACHO had moved in front of them.
Livermore was a leader in such camera technologies at the time, hence its involvement in the MACHO search. The capabilities of early survey cameras inspired Cook and colleagues Tony Tyson of Bell Laboratories, who pioneered the concept of a large, 8-meter-class survey telescope with a very wide field of view, and Roger Angel of the University of Arizona, who invented new techniques for casting mirrors for astronomical telescopes, to begin exploring what a survey telescope larger than ever before might entail. “Tony (Tyson) and Roger (Angel) worked together to decide how to design this very large survey astronomical telescope, and Kem (Cook) became involved because he understood how to plan the observations that a telescope like this could make,” says Scot Olivier, program leader for Nonproliferation Research and Development in Livermore’s Global Security Principal Directorate and prior manager for Rubin’s camera optics and wavefront sensing system.
Their collaboration led to Cook’s development of a program called the Exposure Time Simulator to explore how long a telescope should be pointed at any given part of the sky and how quickly it should be moved around to completely observe the sky. “Ultimately, in around 2000, that program led to the initial survey strategy for Rubin to take a pair of 15-second exposures and march across the sky, covering the entire visible sky every three nights,” says Olivier. Originally called the Dark Matter Telescope because of its planned use in the dark matter search, most of the survey idea was in place by the year 2000 except the optical design for the instrument.
After that, former Laboratory Director William Goldstein, then leader of the Laboratory’s Physics Division, supported early Laboratory Directed Research and Development (LDRD) projects to begin Livermore’s science effort in preparation for LSST. Goldstein also served as Livermore’s chair member for the LSST Corporation, which accepted private donation money to begin long-lead items in the construction of Rubin before government funding began. Also in the early days of LSST, Livermore’s Don Sweeney, formerly program manager for Extreme Ultraviolet Nanolithography, served as project manager for the entire observatory from 2003 until 2012.
Cook’s other influence in LSST came when he brought in Lynn Seppala, now retired senior optical designer for the National Ignition Facility (NIF), who is partly responsible for Rubin’s innovative telescope optic design. Such a unique and unprecedented survey instrument needed similarly unique optics, requiring an expert mind to meet the challenges posed by the project’s goals.
Optical Excellence
Seppala caught Cook’s eye as a candidate to work on the Dark Matter Telescope at an international optical design conference in 1999, where Seppala earned third place in a worldwide design competition. From there, Cook gave Seppala several months of funding to begin tackling the optics for what would become the Rubin Observatory’s telescope, which proved to be a tall task. The target field of view had not been finalized, the chromatic effect of the atmosphere posed warping challenges, and the telescope was intended to achieve a more thorough astronomical survey than had ever been taken before. But Seppala’s history at NIF, as well as his experience with the Hubble Space Telescope, set him up to meet the challenge.
First design iterations received constructive feedback from Roger Angel. Originally, the optical design had situated the primary, secondary, and tertiary mirrors spread out from one another, which would require a long telescope that would incidentally be slow to pivot around the entire sky. The mechanical ineffectiveness of such a telescope urged Seppala and others to create a squatter design, and the solution emerged in the form of primary and tertiary mirrors being built into the same piece of glass. This design was persuasive, offering several advantages: compactness, permanent proper alignment of the primary and tertiary mirrors, and insensitivity to temperature changes. Seppala also needed to adapt to changing parameters, including the initial plan to have no filters in the system growing to include six filters, and the telescope’s desired field of view changing from 3 to 3.5 degrees. The optical design required correction for chromatic aberrations, or the inability to focus all transmitted colors of light into a single point, due to the vacuum window.
The design needed to incorporate three lenses into the camera—a positive, then negative, then positive lens—to compensate for the chromatic aberration of the filters. Each lens has its own chromatic aberration, but by summing the positive and negative contributions of the three lenses and the filters, one can almost eliminate the filter’s effects. “Designing something isn’t enough. A design must consider that problems can be solved by being clever,” says Seppala. “I looked ahead at potential problems while making the optics as simple and as foolproof as possible.” The resulting optical design contains an 8.4-meter primary mirror, a 3.5-meter convex secondary mirror, a 5.0-meter tertiary mirror, and three lenses inside the LSST camera—the largest being 1.6 meter in diameter—altogether capable of delivering a 3.5-degree field of view.
Seppala had also attempted to correct for the chromatic effects posed by the atmosphere near the horizon, which was unsuccessful and resulted in worse results than the original design. Unbeknownst to the team at the time, today’s software now enables this effect to be corrected after images are taken. Today, atmospheric dispersion at high airmass is addressed primarily via calibration and software corrections rather than hardware, reflecting tradeoffs made to preserve optical simplicity and stability. “As everything grew more and more complicated, I focused on the most difficult elements to produce and how I could make them as easily as possible,” says Seppala. “We automatically tested every design iteration by building in five optical design tests into each solution. In the back of my mind, I always held the idea that you can never test a component often enough or well enough.”
Design specifications developed by Seppala and later refined by Livermore optical engineer Brian Bauman were ultimately manifested under the leadership of engineering managers Scott Winters and Justin Wolfe—along with Livermore scientist Vincent Riot as camera manager. The unprecedented transmissive optics, fabricated by multiple industrial partners around the world, met every aspect of the Laboratory’s meticulous design.
Shooting the Southern Sky
To capture the light coming through such an innovative optical system requires an equally innovative camera detector. Rubin’s full night sky survey can be completed in only three days, which is achieved by a camera that can take images large enough to fit 50 full moons.
Rubin’s LSST camera was developed at SLAC National Accelerator Laboratory, where Riot, nationwide project manager for the National Nuclear Security Administration’s Scorpius project, went on a special assignment to serve as the nationwide project manager for the camera until 2023. During this time, he oversaw its fabrication, integration, and testing, a process that took about a decade and leaned on the integrated optical design from the telescope.
The large field of view for the camera requires extremely high resolution. Rubin’s camera contains the world’s largest astronomical charge-coupled device (CCD) focal plane, with 3.2 gigapixels. For reference, a modern smartphone image sensor contains tens to hundreds of millions of pixels. Not only must there be an abundance of pixels, but the aim of Rubin to detect even the faintest of objects in its field of view also requires that the pixels be large enough to gather as much light as possible. “What’s really driving the camera—what makes it so unique and large, and why not very many cameras are made the same way—is that we want to go deep and look at very fine objects,” says Riot. “Rubin needs to take images of as many objects as possible in 10 years, so the Rubin camera is close to the size of a minivan. This is very different from an iPhone camera, which has tiny pixels and fits in a pinhead.”
A further complication came from Rubin’s need to be sensitive to a range of wavelengths including red light, the longer wavelength of which requires that the CCD sensor have pixels with substantial depth. Sensors that meet all these criteria were not commercially available at the time, requiring the team to develop a sensor especially for the Rubin project and subvert the usual method of designing an instrument around an existing sensor. The resulting sensor has pixels with an area of 10 micrometers square and a depth of 100 micrometers, a special coating to prevent losing light from reflection, and a parallelized data readout to prevent any time wasted reading data between shots.
Some of Rubin’s first images became publicly available in June 2025. Already, the observatory is providing visually stunning glimpses into space, with information relevant to nearly all disciplines of astronomy waiting to be uncovered. “The survey is impressive visually, but it is also a game of statistics,” says Riot. “Imaging as many objects as possible provides an immense dataset to work with.”
Into the Dark
The main science goal driving LSST during its early days was to address what is referred to as Albert Einstein’s biggest blunder: Einstein introduced the cosmological constant to his general relativity equations to allow a static universe, later realizing he was mistaken after Hubble’s discovery of expansion. Modern observations reintroduced a cosmological constantlike term to explain accelerated expansion. In the late 1990s, two independent teams measuring type 1A supernovae found evidence that the universe’s expansion is accelerating, consistent with dark energy. In 2005, DOE convened a Dark Energy Task Force, which created multiple stages of dark energy experiments with stage four being intended to measure the equation of state of dark energy to a greater precision.
LSST is one of the stage-four surveys with this goal; the other two are the NASA Roman Telescope still under construction and the Lawrence Berkeley National Laboratory-led Dark Energy Spectroscopy Instrument (DESI). DESI has reported early results suggestive of possible evolution in the dark energy equation of state, although further data and cross-survey analyses are ongoing. DESI uses the spectroscopic galaxy survey method, in which the 3D distribution of galaxies over space cosmic time is mapped to measure baryon acoustic oscillations, or sound waves from the early universe that have been preserved in the distribution of galaxies. On the other hand, LSST uses weak gravitational lensing, or the distortion of light in the universe by the mass density surrounding it. Rubin’s strongest dark energy constraints are expected from weak gravitational lensing, or cosmic shear, combined with photometric redshift estimates across an immense galaxy sample. “LSST will make the definitive measurement of the lensing effect over a large cosmic volume to give us statistical constraints on the dark energy equation of state,” says Michael Schneider, the group leader for Astronomy and Astrophysics Analytics (AAA) in the Laboratory’s Physical and Life Sciences Principal Directorate. “Its strongest measure of dark energy will come from gravitational lensing, and that’s been the focus of Livermore research for the last 15 years as our dark energy science contribution to LSST.”
Much of Livermore’s dark energy contributions began when Schneider came to the Laboratory in 2010 to work on the science applications for Rubin after a postdoctoral research stint working on the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS) survey and previously having completed his doctorate with founding director Tony Tyson. Lawrence Livermore’s work has focused on precision shear measurement algorithms, calibration of systematic uncertainties, and the integration of machine learning into processing pipelines. In 2012, Schneider led an LDRD project that established algorithmic capabilities that reduce the key statistical uncertainty in LSST dark energy measurements, and in 2014, DOE Office of Science provided funding to continue the work. Schneider also earned an Early Career Research Program award from DOE in 2017, which furthered this effort, creating an algorithm that measures gravitational lensing and opening the door to incorporate machine learning into upcoming pipelines. (See S&TR, March 2022, Mission Fulfillment with Exponential Impact)
Now that Rubin is live, scientific discoveries in nearly every astronomic field will be made possible by the new data, going into and beyond the dark energy work that originally motivated the survey. “In one year of observing, Rubin will collect more information than all other telescopes have collected combined. This huge amount of data is being made readily available to researchers,” says Seppala. Adds Schneider, “Early on in the project, everyone involved recognized that it had science applications for essentially the whole field of astronomy.” To untangle this immense amount of data, LSST has stood up eight Science Collaborations, multi-institutional groups of scientists pursuing a specific area of astronomy. Olivier and Schneider were founding members of the Dark Energy Science Collaboration, to which Livermore has made major contributions as noted. However, this is not the only area in which the Laboratory is involved.
Asteroid Detection
In covering the entire southern sky, and in taking long exposures, Rubin is an excellent instrument for detecting moving objects such as asteroids, a focus area of the Solar System Science Collaboration. “Within its first few days, Rubin detected around 4,000 objects, more than half of which are new asteroid discoveries,” says Riot. Nate Golovich, an astronomer in the AAA group at Livermore, adds, “Because the telescope is so sensitive, it can identify objects moving both very slowly and very quickly.”
According to Golovich, a congressional mandate in the 2005 NASA authorization bill required NASA to detect 90 percent of near-Earth objects larger than 140 meters by 2020. That goal was not met, so the field looked to Rubin to help NASA with this endeavor and incorporated changes to Rubin to support more asteroid detection out of the survey. A major change included adding exposures in the form of a twilight survey, in which an area within 20 degrees of the ecliptic plane of Earth—or the imaginary plane containing Earth’s orbit around the Sun—will be covered to detect asteroids and comets that are otherwise difficult to detect in the main LSST. “With this addition, Rubin is doing more observations in the ecliptic plane just after sunset, and this will increase the delivery of asteroid discovery,” says Golovich. To further bolster asteroid discovery on LSST, Golovich headed an LDRD project to develop a moving-object detection algorithm, enabling the identification of asteroid populations missed with traditional surveys.
Rubin’s detection of moving objects also supports Livermore’s global security work. “Through normal survey operations, Rubin will detect objects in the very distant outer solar system—Kuiper Belt objects—and simultaneously detect things moving much more quickly and much closer to Earth, which connects back to the planetary defense mission that Livermore has been involved in for years,” says Golovich.
The Gift That Keeps Giving
Beyond the science applications Livermore and others will continue exploring, the data Rubin collects over the next 10 years will serve as a publicly available resource for all ages, opening the door to involvement by everyone from interested young minds to seasoned researchers. “One of the things that’s unique with Rubin is the educational approach,” says Riot. “A big portion of the project has been to develop a public access platform that can reach even elementary schools.” Adds Golovich, “This survey will deliver better data to the public than what I had to fight for to do my doctorate program, and the richness of the data will provide an incredible number of astronomy projects for future students.”
And the potential discoveries go beyond astronomy: handling such a large volume of data will encourage innovation in data science and AI to take on the challenge. In fact, Rubin will collect more than 500 petabytes of images, catalogs, and derived data products during the 10-year survey—an amount that far exceeds current algorithm-processing capabilities and will require decades to unravel. “Having an unrestricted dataset of this size and complexity is great fodder for data science and machine-learning disciplines to move forward,” says Schneider. Adds Riot, “There will be so much data that the only way to take full advantage of it is by having as many people as possible looking at it, trying new things, and investigating anomalies. Rubin is very different than past astronomy projects in that regard. The breakthroughs the data contains will be waiting for us to find for many years.”
—Lilly Ackerman
For further information contact Scot Olivier (925) 423-6483 (olivier1 [at] llnl.gov (olivier1[at]llnl[dot]gov)).
