Renewed Nuclear Testing in… the U.S.?

Jeffrey Lewis 0:05

You are listening to the Arms Control Wonk Podcast, a podcast on arms control, disarmament, and nonproliferation. I'm Jeffrey Lewis, the founding publisher of the Arms Control Wonk blog and a professor at the Middlebury Institute of International Studies at Monterey.

Anne Pellegrino 0:23

And I'm Anne Pellegrino, an independent analyst in Monterey, California.

Jeffrey 0:29

Not just an independent analyst, and you're also a baker, which leads to our question. What are you baking this week?

Anne 0:36

I made cookies. I made salted tahini chocolate chip cookies, and they are so good.

Jeffrey 0:49

That's fantastic. They're so good. I think, however, for the first time, the Lewis household might be threatening your baking skills because my wife made bagels.

Anne 1:01

I saw that. And so I was like, oh, well, I guess I won't offer loaves of bread anymore.

Jeffrey 1:06

No, we're still accepting loaves of bread. Maybe we'll swap them for some bagels.

Anne 1:13

Oh, yes. We would love that. I would love that. How do you feel about olives in your bread?

Jeffrey 1:20

I like them.

Anne 1:21

Okay. Okay. An olive loaf it is.

Jeffrey 1:26

I don't have some clever, clever segue from bagels to nuclear weapons explosions. You do. Make it. Do it. I don't.

Anne 1:37

What were you doing on Friday evening when this news dropped?

Jeffrey 1:42

So I think the joke that I probably should make is, I don't know, something about, you know, like bagel being zero and zero yield.

Anne 1:51

Is no joke. It was just like a Friday. I was like, what?

Jeffrey 1:55

Anne, have you ever read the Master and Commander series of novels?

Anne 2:00

I have not.

Jeffrey 2:02

So Patrick O'Brien, the author, always does this thing where he can't quite make a joke, where he basically describes the joke that is on the tip of his protagonist's tongue. So I can't make the segue this week, but there is a joke about bagels, bagel meaning zero, and Zero-yield nuclear testing, but it eludes me because I am not clever enough.

Anne 2:27

I will giggle as if the joke was made. But yeah, so that news bomb on Friday evening that on May 15th that Trump administration officials were in intense discussions about whether or not to do a, quote, rapid test.

Jeffrey 2:54

For listeners, we have done podcasts on allegations that Russia is conducting low-yield nuclear tests. We just, sorry, dropped an episode on China conducting low-yield nuclear tests.

Anne 3:09

Aww. There's the joke.

Jeffrey 3:13

But... One of the things, so John Hudson and Paul Sonne have a scoop in the Washington Post in which a senior administration official told them that in discussing how to respond to allegations of Chinese and Russian low-yield nuclear testing, which They started by making them themselves. But having created a public relations mess for themselves, they had to discuss what to do.

And someone suggested that they conduct a rapid nuclear test, not for any technical reason, but solely for political leverage with the idea that would be, quote, useful from a negotiating standpoint. Close float.

Anne 4:06

It's the only way to get the best deal, Jeffrey.

Jeffrey 4:13

Yeah, I think it was Cheryl Rofer who said that Trump's entire idea of negotiation is engage in like blustery rhetoric and then wait for the other side to surrender. And I think that's exactly it. Accuse the other side of cheating, detonate a nuclear weapon, and then a miracle occurs.

Anne 4:32

But this is like the ultimate blustery rhetoric, a desire to resume nuclear testing. But so this article dropped and I obviously was like freaking out and like put me into a bad mood Friday evening. But you rightly pointed out that this isn't the first time that this administration has kind of floated that weather balloon of an idea, or at least factions within the administration have expressed their desire to resume testing.

Is this different?

Jeffrey 5:14

No, my sense is it's exactly the same, you know. One of the things that I found really interesting about the story in the Washington Post is that there was, well, I'll say it this way. We shouldn't call them nuclear tests because tests are things that you instrument for diagnostic purposes because you want to learn something. And what they're talking about in the Washington Post story is the same thing Bill Hennigan wrote about in Time in 2018, which is not a test at all.

It's a nuclear explosion. But basically with little or no diagnostics, not done for any scientific or research purpose, but simply to send a message. Yeah, and the Hennigan piece in 2018, they said it was a simple test done purely for political purposes. And in 2020, what they were discussing was, again, this sort of rapid test, rapid because it's not complicated.

You don't have to do all the diagnostics. It's just a test for test sake.

Anne 6:18

Yeah, that was going to be my question. Because one of the things that people that has been circulating on Twitter after this most recent story came out was the fact that the idea that you could quickly, quote, test a nuclear weapon is not true, that it would take a couple of years to set that up. Again, you make the point we're not talking about a test, we're talking about a stunt.

Jeffrey 6:44

Well, and I think that's exactly right, that if you wanted to conduct a test that would be of value, I don't know who was saying that on Twitter, but that person is right, it would take a long time, both because you would have to Design the experiment and make sure you have in place the appropriate diagnostic equipment in order to collect all the data, but also there would be environmental issues.

So the phrase in the Hennigan piece I think was, oh no, it's not in front of me, but I think it talked about how you waive a bunch of the environmental restrictions.

Anne 7:21

Yeah, it says, when the Trump administration ordered the department to be ready to conduct a short notice nuclear test in as little as six months, he goes on to say, that's not enough time to install the warhead and shafts as deep as 4,000 feet and to fix all the proper technical instrumentation and diagnostics equipment. But the purpose of such a detonation, which the administration labels, quote, a simple test with waivers and simplified process, Would not be to ensure that the nation's most powerful weapons were in order, or to check whether a new type of war had worked, would be, quote, conducted for political purposes.

Jeffrey 7:58

Yeah, so I don't know who said it would take a couple years. That person's right, assuming that this was a real test, but this isn't a real test. This is a stunt. You know, it's the equivalent of the Viagra that Donald Trump takes, you know, it's It's something designed to get his Yeti pubes standing on end.

Anne 8:19

Jesus Christ. Well, the very informed and intelligent thread that I came across was from Amy Wolf, who was talking about if in the context of stockpile stewardship and actually conducting a test and factions within the administration and within the NNSA, Like you got to talk to like lab directors and who actually would be interested in conducting this test and issues with budgeting and, you know, All that sorts of good stuff.

But that's kind of a different conversation because, again, we're not talking about testing. We're talking about a political stunt.

Jeffrey 8:58

Yeah. So as in all things, Amy Wolf is right. Yes. But in this particular case, I think it doesn't apply to Stephen Miller and Johnny Foster out in the desert trying to hotwire Obama. Like, yes, she's right. You'd have to be insane to do it any other way.

Anne 9:26

Meet the Trump administration.

Jeffrey 9:28

You know, this is a thing that's been going on for a long time. I, in 2000... Late 2016, before Trump was inaugurated, there was a story in the New York Times that suggested there were people who were pushing to resume nuclear testing. And I actually wrote this article kind of mocking the idea that opened up by saying that Rick Perry, the incoming Secretary of Energy, he'd gotten a D in a class called meet And, and that not even he was stupid enough to resume nuclear testing.

And I got a note from someone who I won't disclose, but, you know, I would describe as sort of in the know saying like, don't mock them. Like they're serious. There are people who really want to do this. So it doesn't surprise me that in the years since that Idea got floated that they have developed this concept of a simple test without all of the diagnostics, without any particular scientific purpose, just to make a statement.

I mean, in some sense, it's very, very Trumpian.

Anne 10:34

It's like the most Trumpian thing that he could do.

Jeffrey 10:36

Yeah, I have to say, I mean, one of the things that really strikes me and It's a good chance to talk about a thing that happened in the past. You know, the Obama administration came very close to conducting a nuclear test.

Anne 10:48

I did not know that.

Jeffrey 10:49

Which most people don't know because they didn't in the end conduct a nuclear test. But there's a story by David Hoffman in 2011, also in the Washington Post, Which actually I'm quoted in because I helped David a little bit. It's one of my favorite bits of open source. He got an unclassified description from Bruce Goodwin, very generally describing how there had been a serious problem in the stockpile within recent years.

He was very careful not to say when or which weapon. And that it was as close as the U.S. ever got to resuming nuclear testing, but that in the end, they were able to use the diagnostic capabilities associated with supercomputers, which are just really leaps and bounds more sophisticated than anything that was in existence during the era of nuclear testing, to certify that, in fact, the weapon would be fine.

And what I did with David, which was really fun, is David gave me the kind of general description that Bruce had provided, which was, you know, very careful and didn't have any sensitive information in it at all. But I was basically then able to go through DOE documents where people weren't as careful as he was.

Anne 12:24

It's like putting together a puzzle where, yeah.

Jeffrey 12:29

Yeah, that's right. And it turns out you could see in the DOE documents that it was a problem that was identified in 2009 with the B83. And it looks like it was an issue with how the B-83 was moved around. You know, the B-83 is extremely large. And, you know, if you're like pushing it around on a cart in order to put it up in a bomber, you know, it's going to have some vibrations and bumps and things.

And I guess that was not great for the B-83. Which wouldn't...

Anne 13:11

Yeah, you wouldn't know that unless you did simulations. That information can't be cleaned from an actual detonation.

Jeffrey 13:21

Right, so they eventually figured out that they, I mean, what you could have done is you could have taken a B83, roughed it up a bunch, and detonated it in the desert, although I don't know how fun that sounds. But I think in the end, they decided that it wouldn't really have that big of an impact on yield, and I think they just changed some of the handling procedures.

But, you know, I mean, it's a really different case where, you know, that was an entirely technically driven event.

Anne 13:49

Mm-hmm.

Jeffrey 13:49

You know, no one wanted to conduct a test. It's that they had a weapon. There was a significant issue with the weapon. They needed to resolve the issue with the weapon. And they weren't sure in terms of supercomputing that they were going to be able to and models. I mean, we should talk about the role that those things play, but they just weren't sure that technically they were up to Diagnosing the problem and understanding how serious it was without a test.

But in, but in the end they got there. And, and so they, you know, that's a case where they, they had as good of a reason as they've ever had, but they didn't want to do it. And now we're living in this bizarro world where as far as I can tell, there aren't any of those problems. They don't have any technical need from a testing perspective to do an explosion, but they're just really excited about doing one.

Anne 14:44

I think, so I think it's interesting because like we can talk about, you know, politics versus like an apparent necessity to test the weapon. But in terms of like the global context of like who has nuclear weapons and who is also itching to resume testing or who would resume testing if the United States were to, I don't think that those, the rational explanation versus the Trump administration explanation really makes a difference.

To others looking in, they're like, oh, you're gonna do, you're gonna resume testing, like right behind you.

Jeffrey 15:22

Right, I mean, I think one of the most interesting things that's come out of our research looking at Nevaeh, Zemlya in Russia, and Lop Nur in China is that those countries also look like they're ready to do rapid tests for political purposes if need be. You know, there's lots of excavation at tunnels. The test sites are maintained.

I guess not year-round in all cases, but they're in good repair. And so they are kind of, there are people consistently there, even if it's seasonal. So, yeah, I would think that if the U.S. went ahead and broke the moratorium that has existed since the mid-1990s, you know, different countries went into the moratorium and different countries established moratorium in different years.

But, you know, certainly the U.S., China and Russia haven't tested since China's last test in 1996. If the U.S. broke that moratorium, I would think that China and Russia would both follow in short order and probably would have a lot more to gain than we do. Because we should talk about what role testing plays in stockpile stewardship, because I think people don't really, really get it.

Anne 16:40

So what role does it play?

Jeffrey 16:42

People have this idea that we tested for statistical confidence. Because that's a thing you can do with other kinds of weapons, right? Like a gun, you can just fire it a bunch. But the problem is when you test a nuclear weapon, that weapon has been destroyed, right? Every nuclear bomb, every bomb you drop, it doesn't have to be nuclear, but any bomb you drop, Is it has to be untested, right?

So what people imagine is that what you're doing is you're doing a big production run and then you're randomly selecting bombs and you're testing those bombs in order to have some level of statistical confidence, which you can do with a conventional explosive. But they never did that with nuclear weapons. You know, there just weren't enough nuclear tests.

And frankly, they never, it's funny, we used to say they never, and then they went ahead and did it. But to a first approximation, they never took a weapon out of the stockpile and tested it. All the weapons they did were science experiments. I say they never did it, but then people started pointing out that they never did it, and so then they initiated a program, I think, of doing one test of a device out of the stockpile.

It's just enough so that you can't say they never did it, but certainly not enough to establish statistical confidence.

Anne 18:09

I feel like there's a joke in there somewhere about nuclear bombs as artisanal bread loaves.

Jeffrey 18:17

Yes, they kind of are, right? So, you know, what the purpose of nuclear testing was, was not so much to validate statistically the weapons, but it was to validate the people and the process that produced them, right? So it's a little bit like, I'm going to hire you as a baker. And like, I try some of your samples to figure out like, are you a good baker?

And like, have you baked enough bread that you know what the hell you're doing? And you know, like, okay, every now and again, maybe a loaf is going to be off, but at some point you've, you've baked enough bread that people are like, yeah, Ann knows what she's doing.

Anne 19:01

I can then open a big production facility.

Jeffrey 19:05

Yeah, so I mean, I think the important thing to understand is when the US stopped explosively testing warheads, it's not like it stopped testing them, right? Because they still did non-explosive tests. But what they had to do was they had to transfer the basis of their confidence in the reliability of the stockpile From a culture that was centered on testing to a culture that's centered on computer models and validation using the data from the past.

I understand how explosive testing became embedded in the culture of the laboratories, but it was never statistical confidence. And so as we kind of look at the situation today, when I look at the US, Russia, and China, we by far have the most tests. I think we probably have the most data available from the tests we conducted.

I suspect our computer models are the envy of the world. And so it used to be our supercomputers were the envy of the world, but well, the Chinese have caught up. By the way, single-handedly, the weirdest part of this business for me is looking at changing expectations for a supercomputer performance. But I would think that of all the countries in the world, we are going to get the least technical benefit From a resumption of nuclear tests compared to other states.

And I would say Russia and China. So, you know, we're doing this political stunt, which actually opens the door for them to do something that would be of material benefit to them.

Anne 21:01

But of course, Trump doesn't understand that. He just wants big thing go boom. Well, you know, he gets an erection. Jeez.

Jeffrey 21:11

So, you know, I mean, we got to keep them out. Might have to bleep that one out. I mean, it's all about keeping the man happy and satisfied. Yeah, I mean, it's just, I don't know. I guess I would say this about testing, about explosive testing and the rise of supercomputers. And it's one more thing about Bruce Goodwin, who I think everybody likes quite a bit.

When the US was looking at signing the Comprehensive Test Ban Treaty, He actually got asked what, like how much supercomputing performance would the US need to live under a test ban? So, you know, the big question now is can you, but basically can you substitute computer models for nuclear explosions? And it's a kind of fascinating story, like how quickly things change.

So when the US was looking at signing the CTBT in the mid-1990s, there was a discussion about, well, how much supercomputing, how fast of a supercomputer would you need to be able to handle the data and the complexity of models that you'd really want to be confident? And Bruce Goodwin comes up with this number, 100 teraflops. And he actually tells someone later, and I'll quote him, I remember handing my answer in thinking they would kick me out of the room because it was insane at the time.

It was 100 teraflops. That would be for a machine that could do one calculation only and take three months to complete it. And so they wanted to have that by 2004. And back then, that was crazy. Well, the US breaks through The hundred teraflop number between 2004 and 2005. And I, I can't even keep track of it anymore because by 2007, it was past a thousand teraflops.

So I have no idea what the number is, but you know, we're now orders of magnitude past this unachievable number. But what's really important is when they came up with that number, China's supercomputers were basically non-existent, right? There were a couple of Chinese supercomputers that the US monitored very carefully. They were really not impressive. If you look at the ranking of the fastest supercomputers in the world today, right?

The Chinese weapons labs are traditionally right up there and often beating US weapons labs. So China's... Super computing performance has come way, way, way, way up. But you know what they don't have? A thousand tests of data.

Anne 24:05

Yeah.

Jeffrey 24:07

Right? They have 45 tests worth of data. So, Giving them a green light to resume underground nuclear testing and build out their data strikes me as fairly insane.

Anne 24:21

A very bad, very bad idea. Well, okay, so I think we wrap this up by then talking a little bit about What actually may come of this article and the administration's desire to do such tests. Like what actually happens? Like do norms change? Like what are we watching? Like the erosion of something?

Jeffrey 24:48

Yeah, I don't know how you feel, but I actually think it's far less likely that the US resumes large explosive tests. And what this really is, is this is one of these things that shifts our expectations Which is like a common thing in negotiations, right? Where they threaten to do a big yield test and we're like, oh my God, don't do that.

And now if they come back and want to change the US interpretation of zero yield to do very low yield tests, There are going to be a bunch of people who are like, well, that's a compromise, right? It's like there's no nuclear explosions versus we're going to do a really big nuclear explosion, and they're going to compromise in the middle on experiments that would be prohibited under the CTBT, and so therefore would undermine whatever chances that treaty has left of coming into force.

And they're going to do it, and we're going to be like, Happy about it, right? Because they've threatened us with this really big thing and then we're gonna be relieved when it's only this like tiny, equally terrible thing. You know, it's like, do you know this joke about how this guy goes into this bar And he bets the bartender $100 that he can pee in a shot glass at the other end of the bar.

And the bartender's like, there's no way you can do it. So he puts the shot glass down and puts $100 on the table. And he starts peeing. Of course, he misses the shot glass. He's peeing all over the bar. He's peeing all over the bartender. There's urine going everywhere. But not a drop is getting into the glass.

And the bartender's laughing. And at the end, you know, the bartender's like, well, you know, you owe me a hundred bucks. And the guy is laughing. And the bartender's like, well, like I'm the one who won a hundred bucks. Like, why are you laughing? And the guy says, cause I just bet the guy sitting over there, a thousand dollars that I could be on you and you'd laugh about it.

Anne 26:51

Right. So it's all about perspective. Yeah.

Jeffrey 26:54

You know, like if Trump comes in this and says, you know, oh, we're thinking about conducting a nuclear explosion. And we are all worked up about it, then I think that frees them up to do a whole bunch of stuff at a much smaller, lower level that is going to be almost as bad. Particularly when I don't actually think they could get away with a nuclear explosion, given how Las Vegas has sprawled toward the test site.

Anne 27:23

Right. Yeah.

Jeffrey 27:25

I mean...

Anne 27:26

Those pictures, those, yeah, those old picture, well, they're just going to say those old pictures of explosions where people are just like sitting on the hills outside Las Vegas watching is pretty stunning to me still.

Jeffrey 27:39

Oh yeah, you can see atmospheric explosions from Las Vegas. And if you look at how Las Vegas has sprawled in the 30 years, you know, 30-ish, almost 30 years since nuclear testing ended, a lot of that growth has been in the direction of the test site. So, you know, there are going to be a lot of people who are going to feel, you know, depending on how big the test is, maybe you feel a little rumble.

Not great. So what is that? Well, that's what freedom feels like. Do you know this joke when the Japanese used to complain about U.S. aircraft flying over? Because Japanese are very annoyed with noise pollution. And the thing the U.S. would always say to them, which wasn't terribly reassuring, was, that's the sound of freedom. So, hey, you know what?

We have to end on a silver lining. I think I have one.

Anne 28:32

Please do, because I don't. I don't.

Jeffrey 28:35

It's such a good thing that the U.S. never ratified the CPBT.

Anne 28:42

How is that a silver lining?

Jeffrey 28:44

Because Trump would have withdrawn from it.

Anne 28:47

Ugh. Yeah, we'll have to talk about the growing graveyard at some point soon.

Jeffrey 28:54

I think there was a little hidden silver lining here that is as bad as I think things are for the CTBT. I think it might be worse if Obama had gotten ratification of the CTBT, I think.

Anne 29:09

I think we'd be out of it now. No big trees, just some small norms. That's fine.

Jeffrey 29:15

Yeah. Yeah, so there you go, that's your happy note. So I'm glad that we haven't lost the CTPT yet. I like treaties, and if you like treaties, if you like arms control, if you like podcasts about arms control, if you just wanna make sure that Anne can afford flour for bread, you can go ahead and support the podcast.

By going to patreon.com slash ACW podcast for $5 a month, $3 if you're a student. You can support the podcast and you will get access to our members only Slack channel. Which is doing all kinds of wild and interesting stuff. Again, that's only $5 a month. It helps us keep producing the podcast at patreon.com slash acwpodcast.

Anne 30:16

Flower and health insurance. I thank you all.

Jeffrey 30:20

Anne?

Anne 30:22

Yes?

Jeffrey 30:24

Thank you for joining me. It's always a pleasure. And thank you for listening.