Jeffrey LewisNORK DATA: It was a DUD


I love the US Geological Survey.

They’ve published lat/long (41.294°N, 129.134°E) and Mb estimates (4.2) for the North Korean test.

There is lots of data floating around: The CTBTO called it 4.0; The South Koreans report 3.58-3.7.

You’re thinking, 3.6, 4.2, in that neighborhood. Seismic scales, like the Richter, are logarithmic, so that neighborhood can be pretty big.

But even at 4.2, the test was probablya dud.

Estimating the yield is tricky business, because it depends on the geology of the test site. The South Koreans called the yield half a kiloton (550 tons), which is more or less—a factor of two—consistent with the relationship for tests in that yield range at the Soviet Shagan test site:

Mb = 4.262 + .973LogW

Where Mb is the magnitude of the body wave, and W is the yield.

3.58-3.7 gives you a couple hundred tons (not kilotons), which is pretty close in this business unless you’re really math positive. The same equation, given the US estimate of 4.2, yields (pun intended) around a kiloton.

A plutonium device should produce a yield in the range of the 20 kilotons, like the one we dropped on Nagasaki. No one has ever dudded their first test of a simple fission device. North Korean nuclear scientists are now officially the worst ever.

Of course, I want to see what the US IC says. If/when the test vents, we could have some radionuclide data—maybe in the next 72 hours or so.

But, from the initial data, I’d say someone with no workable nuclear weapons (Kim Jong Il, I am looking at you) should be crapping his pants right now.

First the missile, then the bomb. Got anything else you wanna try out there, chief?


  1. ross (History)

    Forgive my ignorance, who should be crapping in their pants and why?

  2. ross (History)

    you also commented that 3.7 gives you a couple hundred kilotons, and yet you also say 4.2 yields around a kiloton, what am I not getting? Thanks

  3. Jeffrey Lewis

    Sorry, I wasn’t super clear in the first draft and made a typo — it was tons, not kilotons. Mr. Kim should be crapping his pants, as the value of his nuclear deterrent just dropped to zero.

  4. Muskrat (History)

    Doesn’t that also raise the issue of a possible bluff—a large chemical explosion? Can the boffins tell the shock waves apart?

    Bluffing with such a small yield would be a bad idea, but then it looks like NorK is full of bad ideas. Any speculation on what kinds of failure mode could produce such a dud?

  5. j house (History)

    It may be the first of several tests to come, and they will improve their capability (as well their missile capability).Give them another five years, unless there is some dramatic internal policy change.
    The Chinese have lost serious face here.Their credibility is seriously damaged.

  6. Haninah (History)

    I was thinking possible bluff, too. A few hundred tons to about 1kT, unlike, say 20kT, is within the ranger where it’s at least remotely conceivable that some huge amount of conventional explosives (even ANFO) could have been used to simulate/fake a nuke explosion. It’s possible logistically, of course – ye olde Divine Strake is supposed to weigh in at around 500 tons TNT equivalent – but I imagine that our intelligence community would be able to tell the difference between someone installing a nuclear device and someone installing two million pounds of cow pie and diesel. (Of course, some people were speculating the Norks would try to fake a test in the 20kT range, but I don’t even want to think about how much ANFO that would be – ridonculous).

  7. James (History)

    Before you celebrate the “failure” of a DPRK nuclear test, keep in mind that the Pak and Indian tests in 1998 were smaller than expected, but they still managed to join the nuclear club as de facto members.

    On that note, also remember that nuclear tests rarely consist of a single shot, and there are some indications already that this story may not be over:,23599,20553706-1702,00.html

  8. Amyfw (History)

    Jeffrey, I value your opinion. So, please triple check your math as I will be referring everyone to you when I start getting calls at work tomorrow.

  9. Geoff Forden (History)

    Congratulations for doing an analysis of the yield of NK’s nuclear test and thanks for putting up the reference material! However, I think you are somewhat premature in calling it a dud. The reference material you link to gives several examples of yield equations and the hard rock equation that you use is the most pessimistic. If the test took place in soft soil, such as the slopes of a mountain that was weathered, then the explosion does not couple nearly as strongly to the surrounding earth and the yield equation changes substantially. I would suggest giving a range of yields for the lowest to the highest. The lowest, with currently available data, uses the hard rock coupling with south korea’s estimates for quake strength: 0.45 KT. The high end uses the USGS estimate, 4.2, and the soft soil yield equation: 8.2 KT, which is hardly a dud.

  10. Claude Muncey (History)

    James, reports that the initial Pakistani tests came in a 40 KT—not bad for the first tries, twice nominal yeild.

  11. TigerChE PE

    In the research paper linked states that this equation is valid in the range of 5.3 kTon to 120 kTon. There are other relationships in the 52 kT to 1000 kT and > 60 kT. Extrapolation outside of valid range may not give accurtate results. It looks as if this equation will OVERSTATE the yeild at lower body wave magnitudes (richter scale readings)

  12. Richard

    The Pakistani claim to have detonated several nukes at once and some were only sub kt range! So there IS a country which starts off with a sub kt device. And one with ties to DPRK. And a tactical weapon makes sense on the south korean battle field…..SCARED ?

  13. BJR

    It might just be this guy stomping around again:“Catastrophic Valiant Kim-Chee Earthquake Stomp-Kick!”

  14. mark gubrud (History)

    If the test really was a fizzle, with a yield around a kT, that’s still quite a terror weapon, if it can be delivered. “Zero” is a pretty drasically low estimate of its deterrence value. To say nothing of NK’s sufficient conventional deterrent.

    I’m still expecting that the Bush gang will see this as creating both an opportunity and a necessity to strike Iran before the election. But a perception that the NK test was a “dud” might make enough of a difference there.

  15. liberal (History)

    James wrote, ”…keep in mind that the Pak and Indian tests in 1998 were smaller than expected, but they still managed to join the nuclear club as de facto members.”

    I thought India joined the nuclear club in 1974, not 1998.

  16. Formica (History)

    If in fact this test was a fizzle, what are the chances that they were actually testing a weaponized design, i.e. a boosted fission weapon based on AQ Khan’s provided expertise? Perhaps it was their intent to skip the Fat Man stage and go straight to trying for a useable warhead. A second test might indicate that the weaponized test was a failure, and that they’re going to now test a cruder, unweaponized big-hunk-of-plutonium weapon that doesn’t require such precise lenses/subcritical mass.

  17. ET (History)

    Assuming this was a real nuke, I wouldn’t discount the possibility of a subcritical device with an implosion squeeze. Both more valuable for purposes of constrainted plutonium supply and more impressive technically – especially for black market purposes.

  18. gtash (History)

    I want someone to explain what constitutes a “dud” in terms other than kilotons. Are you really saying the explosive used to compress the nuclear materials did ignite, but the necessary nuclear compression needed to force a chain-reaction did NOT take place?

  19. lawnorder (History)

    I posted an excerpt at kos, with links to here.Sorry if I quoted too much, but your server seems to be taking a beating with all the hits.

    My money is on NK faking it.

    Thnks for the informed commentary!

  20. jim (History)

    Actually the one person in the world who’s absolutely sure the test was a success is Kim Jong-Il.

  21. James (History)

    Why aren’t we also asking about the possibility of a fake test; one million pounds of TNT is not that costly or bulky if squirreled away over time. And please recall the 22Apr04 dynamite accident near the Chinese border with North Korea when “A Red Cross team sent to the scene has reported that 1,850 homes have been flattened and 6,350 seriously damaged.”

    I’m just saying…

  22. Grumpy Physicist (History)

    I suspect that both Pakistan and NK have been suffering from using a poor mix of Pu isotopes in their bombs. This is a result of trying to extract Pu from power-reactor fuel, that typically is irradiated for 12-18 months, rather than the 3 day irradiation that is used when a reactor is dedicated to producing bomb-grade Pu.

    (the timescale is determined by the halflife of 239Np; back in the Manhattan project, Hanford used to push fuel pellets through with a 3-day residence time in the reactor for the same reason)

    So now Pakistan is starting work on a reactor specifically for producing Pu. Oh joy.

  23. Robert Merkel (History)

    Why should Kim be crapping his pants? Regardless of the success or failure of this test (and a half-kiloton fizzle is still going to kill many thousands of people if launched into the centre of a densely populated city) nothing else has changed about the strategic situation.

    You know. The one that says that Seoul turns into a artillery free-fire zone, complete with chemical weapons, if our buddy Kimmy gives the word.

  24. sunbin (History)

    “Got anything else you wanna try out there, chief?”

    yes, wooden planes and tanks!

  25. Rocketman

    ‘DUD’ !!!!?????????I love how those GOP fans try to put a spin on everything, the DPRK dictator has always openly said he’s got a WMD program ( bioweapons, nuclear and chemcial )It ain’t a dud, anyone who says it to be a ‘dud’ clearly have no understand of nuclear tesing. Many of the British nukes at Operation Antler, the US Nukes at the Nevada test site, and the Pakistani Nukes produced very small yields (but would still be powerful enough to kill hundreds of thousands people).

    Perhaps when that dictator Kim finally gets to build his ICBM with all those cherry-on-top warheads he desires so much, so he can turn SanFranciso and Texas into radioactive craters we will finally be convinced this madman has got a few WMDs.

  26. Olly

    [snip] [snip]

  27. ploeg

    How ‘bout the strategic situation where the Chinese decide that Kim is too much of a pain in the tuckus for their own good, and decides to cut aid shipments and encourage a coup?

  28. Geoff Forden (History)

    I have to correct my earlier post on the yield assuming “soft earth.” In fact, while soft earth would produce a smaller estimated yield, I mistakenly used the formula for surface waves (as opposed to “body waves” that are detected far from the source). This had the effect of dramtically over estimating the yield.

  29. urizon (History)

    Is it possible that this was a conventional explosion, designed to look like a nuclear test? I can’t imagine that any government would have much trouble assembling and detonating 500 tons of plastique.

    Any reliable sources out there as to a radiological fingerprint / signature (sorry, I don’t know the proper vernacular for this)?

  30. Jon Kåre Hellan

    If you are in charge of developing in North Korea’s nuclear bomb, and have failed, what do you do? Tell Kim Jong Il? Or get hold of a lot of TNT and pretend that the bomb works?

  31. Mei (History)

    Everyone’s fizzled. The Yankies invented it and the Russkies, Brits, French, Chinese, Indians and Pakistanis have all joined fizzle club. Then you’ve got the US speciality when you get overshoot, the opposite of a fizzle – El Zzif – I don’t know what you’d call it. French and and Pakistanis must be relieved as up till now they held record for earliest fizzles. See my note on the history of fizzles at this link.

  32. abm (History)

    The ultimate price for this dud will be paid by Kim’s scientists. Which is probably a good thing for the rest of us.

  33. Lab Lemming (History)

    If you laugh at the USGS, then you are amused by computer generated jokes. The USGS computers pick seismic signals and determine magnitude and location automatically. If an event is greater than mag 5, it will also calculate a first moment tensor, but this event was not large enough for that to happen.