U.S. DOE selects Berkeley Lab to lead DOE/NSF experiment that combines observatories at the South Pole and in Chile’s high desert
The largest collaborative undertaking yet to explore the relic light emitted by the infant universe has taken a step forward with the U.S. Department of Energy’s selection of Lawrence Berkeley National Laboratory (Berkeley Lab) to lead the partnership of national labs, universities, and other institutions that will carry out the DOE roles and responsibilities for the effort. This next-generation experiment, known as CMB-S4, or Cosmic Microwave Background Stage 4, is being planned to become a joint DOE and National Science Foundation project.
CMB-S4 will unite several existing collaborations to survey the microwave sky in unprecedented detail with 500,000 ultrasensitive detectors for 7 years. These detectors will be placed on 21 telescopes in two of our planet’s prime places for viewing deep space: the South Pole and the high Chilean desert. The project is intended to unlock many secrets in cosmology, fundamental physics, astrophysics, and astronomy.
Combining a mix of large and small telescopes at both sites, CMB-S4 will be the first experiment to access the entire scope of ground-based CMB science. It will measure ever-so-slight variations in the temperature and polarization, or directionality, of microwave light across most of the sky, to probe for ripples in space-time associated with a rapid expansion at the start of the universe known as inflation.
CMB-S4 will also help to measure the mass of the neutrino; map the growth of matter clustering over time in the universe; shed new light on mysterious dark matter, which makes up most of the universe’s matter but hasn’t yet been directly observed, and dark energy, which is driving an accelerating expansion of the universe; and aid in the detection and study of powerful space phenomena like gamma-ray bursts and jet-emitting blazars.
On Sept. 1, DOE Office of Science Director Chris Fall authorized the selection of Berkeley Lab as the lead laboratory for the DOE roles and responsibilities on CMB-S4, with Argonne National Laboratory, Fermi National Accelerator Laboratory, and SLAC National Accelerator Laboratory serving as partner labs. The CMB-S4 collaboration now numbers 236 members at 93 institutions in 14 countries and 21 U.S. states.
The project passed its first DOE milestone, known as Critical Decision 0 or CD-0, on July 26, 2019. It has been endorsed by the 2014 report of the Particle Physics Project Prioritization Panel (known as P5), which helps to set the future direction of particle physics-related research. The project also was recommended in the National Academy of Sciences Strategic Vision for Antarctic Science in 2015, and by the Astronomy and Astrophysics Advisory Committee in 2017.
Berkeley Lab Director Michael Witherell said, “The community of CMB scientists has come together to form a strong collaboration with a unified vision of what is needed for the next generation of discovery,” adding, “We will work with the universities and other laboratories, supported by the DOE and the NSF, to turn this vision into a CMB observatory that has unprecedented power and resolution.”
The NSF has been key to the development of CMB-S4, which builds on NSF’s existing program of university-led, ground-based CMB experiments. Four of these experiments – the Atacama Cosmology Telescope and POLARBEAR/Simons Array in Chile, and the South Pole Telescope and BICEP/Keck at the South Pole – helped to start CMB-S4 in 2013, and the design of CMB-S4 relies heavily on technologies developed and deployed by these teams and others. NSF is also helping to plan its possible future role with a grant awarded to the University of Chicago.
The CMB-S4 collaboration was established in 2018, and its current co-spokespeople are Julian Borrill, head of the Computational Cosmology Center at Berkeley Lab and a researcher at UC Berkeley’s Space Sciences Laboratory, and John Carlstrom, a professor of physics, astronomy, and astrophysics at the University of Chicago and scientist at Argonne Lab.
CMB-S4 builds on decades of experience with ground-based, satellite, and balloon-based experiments, and Berkeley Lab has had a prominent role in CMB research for decades, noted Natalie Roe, Berkeley Lab’s associate laboratory director for the Physical Sciences Area.
Berkeley Lab’s George Smoot, for example, shared the Nobel Prize in Physics in 2006 for leading a research team that discovered ever-slight temperature variations in the CMB light.
Adrian Lee, a Berkeley Lab physicist and UC Berkeley professor, has served on the leadership teams for a number of precursor experiments to CMB-S4, including POLARBEAR/Simons Array and the Simons Observatory. Lee noted that the Simons Observatory and POLARBEAR have contributed design elements that are relevant to CMB-S4 – such as in the areas of optics and cryogenics.
Borrill pioneered the use of supercomputers for CMB data analysis, led data management for the CMB research community for the past two decades at the DOE’s National Energy Research Scientific Computing Center (NERSC), and has served as the U.S. computational systems architect for the European Space Agency/NASA Planck satellite mission, which probed the CMB in great detail.
“What’s new about CMB-S4 is not the technology itself,” Borrill said, “but the scale at which we plan to deploy it – the sheer number of detectors, scale of the readout systems, number of telescopes, and volume of data to be processed.”
Roe noted that Berkeley Lab has particular expertise in data management, and in the design and fabrication of detectors for CMB experiments.
“This is a very big project,” Roe said. “We plan to staff up and bring in all of the expertise and capabilities from our sister labs and from the university community.”
CMB-S4 will exceed the capabilities of earlier generations of experiments by more than 10 times. It will have the combined viewing power of three large and 18 small telescopes. The major technology challenge for CMB-S4 is in its scale. While previous generations of instruments have used tens of thousands of detectors, the entire CMB-S4 project will require half a million.
The latest detector design, adapted from current experiments, will feature over 500 silicon wafers that each contain 1,000 superconducting detectors, on average – some wafers will contain up to 2,000 detectors.
Aritoki Suzuki, a Berkeley Lab staff scientist who is a detector team co-lead for CMB-S4, has been working with industry to develop faster and cheaper manufacturing processes for the detectors, as an option that can be considered, and noted that multiple manufacturing sites at research institutions are needed, too.
“Delivering nearly 500,000 detectors will be one of the biggest challenges of the project,” Suzuki said. “We will combine forces from national labs, universities, and industry partners to tackle this immense task.”
Another major hardware focus for the project will be the construction of new telescopes. The data-management challenges will be substantial, too, as these huge arrays of detectors will produce 1,000 times more data than the Planck satellite.
CMB-S4 plans to draw upon computing resources at Berkeley Lab’s NERSC and the Argonne Leadership Computing Facility (ALCF), and to apply to NSF’s Open Science Grid and eXtreme Science and Engineering Discovery Environment (XSEDE).
The project is hoping to deploy its first telescope in 2027, to be fully operational at all telescopes within a couple of years, and to run through 2035.
Next steps include preparing a project office at Berkeley Lab, getting ready for the next DOE milestone, known as Critical Decision 1, working toward becoming an NSF project, and working across the community to bring in the best expertise and capabilities.