Nuclear Threats Are Looming, And Nobody Knows How Many Nukes Are Out There
As best anyone can tell, there are 12,512 nuclear weapons in the world. But the world of nuclear deterrence is one of secrets and threats.
Deterrence, the idea that no one will attack you if you have a nuclear weapon, relies on the threat being ambiguous. But scientists and other experts dedicated to reducing that ambiguity work to count the warheads using open-source sources of information such as satellite imagery so the rest of the world can get a better, if imperfect, idea of how many world-ending weapons are out there.
The number of known nukes has grown in recent years, but it’s still down from its alarming 1985 peak of more than 60,000. Keeping track of all the nuclear weapons is hard, and imprecise, work. The nine countries that have them aren’t forthcoming with the exact number of weapons in their stockpile. The U.S. and Russia once shared information about their arsenals with each other as part of anti-proliferation treaties, but negligence on the part of both governments has seen those treaties weaken in the past few years.
So how do we get the number 12,512? The answer is complicated, but it involves high resolution satellite imagery, government reports, and military parades.
Mat Korda, a Senior Research Fellow at the Federation of American Scientists (FAS), is part of a team of researchers that tracks the world’s nuclear weapons. Every year, they publish their informed estimates as the Nuclear Notebook at the Bulletin of the Atomic Scientists.
For Korda, it’s not just about the numbers, which investigators have accepted will not be completely accurate. “The story is not whether or not the number has gone up or down,” he told Motherboard. “The story is about the broader trends we’re seeing in each country’s arsenal. What we’re seeing is that transparency is going down, military stockpiles are going up. The weapons themselves are getting more sophisticated. Some of these legacy weapons from the Cold War era are getting decommissioned and we’re getting new types of weapons. All of this comes together to make nuclear use more likely than at any time since the end of the Cold War.”
The nuclear powers don’t make the task of estimating their stockpiles easy. Some countries share information that hints at their nuclear capacity, but don’t explicitly state how many weapons they have. The U.K. has discussed setting limits to the amount of nukes it has, but hasn’t disclosed how many it has. “On the other end of the spectrum, you have countries that have never acknowledged that they have nuclear weapons. Like Israel,” Korda said. “Within that broad spectrum, there is not much official data about nuclear weapons stockpiles that the public has access to. And so we are forced to rely on little snippets of information that we can divine from things like treaty data, or government statements or leaks or freedom of information requests or military parade videos.”
Korda said there are three main sources the researchers use to figure out how many nukes everyone has: government sources, media sources, and open-source intelligence. Government sources are often the most helpful and clear, as long as you’re looking in the right place. “Treaties, official statements, or freedom of information requests that we can make to a particular agency,” he said. “These are things like budgetary documents where there’s a lot of stakeholders involved. And so we can rely on that kind of data.”
But government and media sources have biases that Korda and his team have to keep in mind when they’re working. “When the United States makes estimates about how many warheads China is going to build in the future, these are estimates based on assumptions. And those assumptions are based on biases or might be meant to send some kind of political message,” he said.
That doesn’t mean those political messages aren’t helpful. “Russia tends to do this fantastic thing, where, at the end of every year, the commanders of their missile forces, they all sit down for an interview with Redstar, or some other Russian publication,” Korda said. “And it's great, because they say, ‘Here's what we did this year, here's everything that we did.’ They go into detail, they say, ‘Here's how many ICBMs we put in the ground here is how many new units of this particular system that we've added to the arsenal.’ And next year, here are our goals.’”
Korda and others read these interviews, verify what’s said against satellite imagery, and update the lists of known nuclear assets if it checks out. The Russia example highlights how integral open-source intelligence, and especially commercial satellite imagery, has become to the process. “We also use, we look at parade videos, we look at images that countries have released, or that have been taken by civilians or militaries of their nuclear forces,” Korda said. “We can potentially measure things, especially when combined parades. That's a great opportunity for us to measure missiles and get a sense of perhaps some of their rudimentary performance like their range.”
Military parades in particular are a great way to get information about a country’s weapons stockpile. “You might have a parade that features all sorts of very specific missiles,” Korda said. “And then a few years later, some of those missiles have disappeared and they’ve been replaced with new ones. Perhaps that sends a message about what things are being prioritized by that country.”
Everything comes back to the satellites, which allow researchers to see what’s changing on a near-daily basis. Satellite imagery helped Korda and others uncover the recent buildup of intercontinental-ballistic missile silos in China. It also shows how this work is collaborative, with the work of one person or group building on that of another. In 2019, Hans Kristensen, a colleague of Korda’s at FAS, discovered new kinds of missile silos near training areas in Jilantai, China.
“The really interesting thing about Jilantai is, not only were they deploying 16 silos in that area, but they were also covered with these really interesting inflatable domes like you might see covering a tennis court in winter,” Korda said. “It was very clear that those domes had a direct connection to China’s ICBM program, because we can look at satellite imagery and see they were covering silo holes. We could also see types of material that would be used to make silo walls. At that point, we had a signature we could go and use for larger deployments.”
In 2021, a researcher at the James Martin Center for Nonproliferation Studies, building on the work of Kristensen, found satellite imagery that showed a buildup of the silos in the desert near the city of Yumen, China. “The signature linking that field to the silo program was those inflatable air domes. When some of the domes got removed you could see there were silo holes. So now, not only are these domes here but we now can see what it looks like when there’s 120 of these in a field. We can see, roughly, what’s the surface area of the space we’re looking at. How far apart is each silo?” Korda said.
Korda took that signature—the inflatable domes, the number of holes in the ground, their distance from one another—and went looking for others. Within a few hours of searching through satellite imagery by hand, he’d found a second field. “You don’t want to look through all of China because that’s a very large country and it’ll take forever,” he said.
To narrow the search, Korda said he made several assumptions. “If China’s going to build more than one silo field, my guess was that it was not going to be that far from the first one because these things cost a lot of money to build,” he said. “There’s a lot of time and effort taken to ship things across the country. And probably the same guys that built the first one are building the other one, just in sequence.” Once Korda had his search parameters down, it only took two hours to find the second field.
Later, using the same parameters, another researcher found a third field of new nuclear weapon’s silos. “It was clear that China was starting to think a little bit more about how to position some of its newer ICBMs in silos. Silos themselves aren’t new in China. But certainly the scale that we see now is a new thing,” Korda said.
This kind of work is what it takes to figure out how many nukes everyone has. According to Korda, it’s a constantly evolving process. He said that he’d love for his team to do real time updates of the stockpile, but it’s a big job. There’s only three of them. At certain points in the year, they sit down and crunch the numbers on a specific country. These numbers feed into an end of the year estimate. “Some of the data that we rely on come at very specific points of the year. Certain countries have military parades that correlate with national holidays. And we have to wait for those parades to learn new things,” he said.
It doesn’t have to be this way. Korda argued that nuclear weapon’s states could be more transparent about their arsenals. The default position of most states is secrecy, but Korda argued that warding off nuclear adversaries far outweighs any potential harm to national security.
“Explaining how many nuclear weapons are in the stockpile…that sends a strong deterrence message,” he said. “You tell your adversaries what the composition of your nuclear arsenal is and how you might use it in different circumstances. To me, that is a strong deterrent message relative to just keeping everything a secret.”
That secrecy, Korda argued, makes the world a more dangerous place. “When countries don't reveal any information about their nuclear arsenals, it opens up this pretty significant back in the information space, where propaganda and exaggeration and conspiracy theories and misinformation about nuclear weapons can really come to the surface,” he said.
And that secrecy means it’s hard to ever know how well Korda and his colleagues have done counting. “Everything we do is an estimate and estimates are based on assumptions we make. For some countries, we’re able to make assumptions that are quite accurate,” he said. “But for some countries, we have a lot less confidence because we lack data. When we put out a number for a country like Israel, North Korea, or even India and Pakistan, there are a lot of question marks.”
Even with all the tools of the modern world, it’s hard to know just how large the stockpile is and how sophisticated it’s become. For Korda, it’s a serious danger. “When one goes off it affects the lives of every single person on earth.”