Astrophysicist Chris Fryer and his family were driving back from a monthlong workshop on neutron-star mergers when he got word of a new gravitational-wave discovery. The fifth of the Laser Interferometer Gravitational-Wave Observatory’s catches, the August 2017 event appeared to be the first detected from two colliding neutron stars. “I recall vividly having my wife read me the email while I was driving and me telling her I didn’t believe her,” he says. His wife, Aimee Hungerford, is also an astrophysicist. A few days later, data started to become available.
“I’m sitting around thinking, ‘What can I contribute?’ ” recalls Fryer, a scientist at Los Alamos National Laboratory. Fryer’s work focuses on gamma-ray bursts, core-collapse supernovae, neutron stars, and nucleosynthesis—all topics that have some physical relevance to his “mission” work, as Department of Energy scientists call their national security research.
University physicists had done calculations about the violent crash—calculations of how much of which elements had formed, and how. But their models were simpler than DOE’s. For decades Los Alamos has worked on a massive software project, called the Los Alamos Suite of Atomic Physics Codes, that simulates plasma and spectroscopic results in a wide range of conditions—the kinds of conditions one might find in nuclear detonations here on Earth and in stellar explosions millions of light-years away.
So Fryer and his colleagues, in collaboration with scientists around the world, used the atomic physics codes and other software to decipher some of the details of the neutron-star collision. Fryer’s team couldn’t give others access to the code suite itself, which is export-controlled, but they could release the data it generated. The codes provided detailed opacities for various elements, which were crucial for quantifying how much of each element was forged in the explosion.
The neutron-star research illustrates why people like Fryer get into national security science: for the interdisciplinary work, the attention to detail, and the freedom to take up tangential projects. But that kind of science, and the secrecy around it, can get complicated. After all, science is supposed to be open, replicable, and untainted by outside interests—all things the atomic physics code suite is not. For a certain kind of physicist interested in a certain kind of science, the trade-offs nonetheless feel worthwhile.
An evolving landscape
Physics is no stranger to sealed work, or to the impact that opening the lid can have. During World War II, the development of radar led to modern radio astronomy. Similarly, infrastructure meant for tracking missiles, aiding communications, and understanding space weather’s security implications later furthered solar and atmospheric research.
The Manhattan Project is, of course, the ultimate exemplar of big, secret science. Tucked away in clandestine labs, thousands of researchers developed the first atomic weapons, in an era that made scientific secrecy official. In the early 1940s, the Office of Scientific Research and Development—the organization that, for a time, oversaw the Manhattan Project—created a classification system similar to the military’s. The choke hold grew tighter in 1946, when the Atomic Energy Act rendered information about atomic energy “born secret,” meaning that it was “automatically classified at the moment of its creation,” according to a 1982 National Academy of Sciences report.
Later, formerly hush-hush aspects of nuclear physics emerged from these national labs into the public square, invigorating disciplines like plasma physics and allowing scientific communities to grow. Scientists could collaborate across borders on fusion; they could publish more research in public journals; universities could teach high-temperature plasma physics courses. The Monte Carlo method of simulation, developed to mimic the insides of bombs, diffused across specialties; today its uses include structuring telecommunications networks, estimating risk in business, and identifying molecules’ supercooled states.
Of course, the end of the Cold War did not lead to totally open high-energy or nuclear physics. Today research related to nuclear weapons is kept quiet in part because the US no longer tests nuclear weapons. If you’re not detonating bombs to see how they behave—but you still have said weapons stashed beneath Albuquerque—you can only simulate their state.
Think about an iPhone, Fryer says. “You’re trying to see whether this phone works,” he says, “and you can test these different pieces.” But you can never test the phone itself; you can only simulate how those tests play together. That requires many lines of code. “Some of them we can’t talk about,” says Fryer. Others, like the code he used on the neutron-star work, can be talked about but not shared with people who don’t have the correct clearance. And some are fully public.
“In the past it’s been really hard to get the codes out into the open,” Fryer says. “And I think it’s a mistake.” As with the Manhattan Project, he explains, what modern scientists do for national security physics can advance physics physics—in part because it tends to be more granular than academics’ simulation software. The stakes, after all, are higher. “In astrophysics, if I’m wrong by an order of magnitude, I haven’t killed anybody, right?” he says.
To Fryer’s satisfaction, transparency is becoming more common as the lab creates new software that’s transparent from the get-go. A nuclear code called PRISM is undergoing an evaluation to decide whether it will be made open source, and the lab is working on an open-source code to model the winds and neutrino transport in astrophysical disks based on existing software called BHlight. The organization is reaching outward, too—contributing, for instance, to Stanford’s parallel-programming architecture, called Legion.
A delicate balance
The shift toward increased transparency mirrors a change in national security culture more broadly. Around the federal government, people are talking openly about “overclassification.” When it comes to national security science, releasing more information now—as was done in the past—means that university researchers (and the broader community in general) will have access to data and tools that can complement their work. Scientists like Fryer focus on nailing down errors, understanding uncertainties, and generally writing tedious, private papers that no one would want to cite even if they could. According to Fryer, the upside of not being able to publish everything publicly is that no one expects you to. In direct opposition to academic culture, a lot of times, it’s publish and you do perish.
National security physics tends to incorporate more physics at once than does academia. Fryer’s research, for example, includes neutrino transport, nuclear burning, magnetohydrodynamics, and radiation transport. “I know radiation transport way better than I learned as an astrophysicist,” he says. Work like his appeals to scientists, Fryer claims, who are more attached to the how of problem solving than to the what of the problem itself.
Historian Audra Wolfe, who analyzed Cold War science in her recent book Freedom’s Laboratory, agrees. “Interdisciplinarity was really built into these national security projects in a way that it just wasn’t in academia,” she says. That’s part of what makes it attractive to smart people who like big problems. But throughout history, the excitement of the challenge has sometimes clouded scientists’ judgment about which projects were ethical to work on, and why. “It is human nature to find value in whatever we spend our lives doing,” Wolfe says. “If you spend your life working in a nuclear weapons lab, you’re going to find some ways to tell stories about that experience to yourself and to others that show that this was a valuable experience.” To determine whether those stories are true, scientists must analyze their personal data objectively.
Ethical questions don’t weigh as heavily on physicist Tess Light, chief scientist for Los Alamos’s Space-Based Nuclear Detonation Detection program, which aims to make sure countries are upholding the 1996 Comprehensive Nuclear-Test-Ban Treaty.
The closed nature of her work does bother her. “That is actually one of the things that keep people from wanting to work in this field,” Light says. “As a scientist, what are you if you’re not publishing?” But she enjoys the less cutthroat mentality, and resources in her field are more abundant than they were in her basic-research experience. “There’s plenty of money,” she says. “And there’s plenty of work.”
Still, this work is not what she—initially an astrophysicist—first imagined for herself. Her philosophical shift started when she first applied for time on big telescopes and struggled to explain in proposals why her research mattered. “Because if you don’t let me observe this galaxy, in a million years it will have redshifted beyond what we can detect?” she recalls thinking. “That is hardly compelling.” Feeling uncomfortable getting paid to study something simply because she thought it was cool, she turned her gaze to the applied sciences.
Light now works with orbiting telescopes that look down instead of up, using instruments that ride the GPS III constellation to scan for electromagnetic pulses from nuclear tests. Those same instruments also let her do nonpolitical scientific studies, because those pulses also come from lightning. “We’ve done things that look at the connection between lightning rates and severe storms,” she says. For example, before a hurricane grows stronger, there’s a recognizable jump in the rate of lightning strikes, a finding the Los Alamos team first published in 2005, using open data. Identifying that jump as a storm churns toward land could save lives.
Although other such information exists publicly, Light’s global, persistent data from the GPS sensors doesn’t; it remains locked behind a classification code. “We can’t publish it,” she says. “I’ve been talking with my team, and some brave soul needs to take on the bureaucracy.”
The held-close nature of the data reminds Light of industry—where scientific innovation is often a trade secret—as do the research’s fast pace and formal test procedures, which resemble capitalist pursuits and the activities of engineering firms. “We’re more deliverable-driven than academia,” she says, “but have more R&D flexibility than most industry.” Also industrial is the ability to move between specialties, which is hard if you want to make a name for yourself in academia. “When I was in grad school, I had a professor who had studied one star—one particularly interesting star, but one star—his entire career,” she says.
Light wanted both more depth and more breadth and to focus on a problem that directly affected lives on Earth. She wanted to explicitly work on applications. That stands in contrast to fundamental science research, whose practitioners ask and answer questions about the nature of the universe for the sake of understanding the universe. Posing initial questions that stem from programmatic and political goals perhaps gives scientists like Light and Fryer a higher-resolution view of how science both shapes and is shaped by the planet on which it takes place. “Often, scientists in those positions had much more clear-eyed assessment of the relationship between science and politics,” says Wolfe.
For researchers considering national security work, that could be a plus or a minus, depending on mind-set. It’s a departure from the traditional scientific enterprise that is inculcated into researchers in graduate school, says Light. “For example, there are some people who feel that fundamental science and high-profile results are the mark of being a ‘real’ scientist,” she says. “If you buy into that, the classified and applied aspects of this work will feel like a distraction from what you want to be doing.”