Was North Korea test a fake or a dud?
This diary on Kos looks at the seismic data and has a nice diagram of how a plutonium-based bomb works. (It's pretty hard to do.) It seems to say that the test "fizzled"; a fizzle being a nuclear detonation in which the compression didn't go exactly as planned and a lot of the plutonium got blasted away rather than compressed into the critical mass.
We should have some expert opinion coming in on Internal Monologue soon!
We should have some expert opinion coming in on Internal Monologue soon!
Comments
Mudanjiang, Heilongjiang, China (MDJ) recorded it first and computed a local (Richter 1935 method) magnitude of 4.2, a duration magnitude of 5.0, and a Nuttli method magnitude 3.8. The duration magnitude isn't useful (and the USGS discarded it) because this was not an earthquake, so the usual duration relationships don't apply.
Inchon, South Korea (INCN) came up with 4.1 by Richter 1935 method, and 3.7 by both Nuttli and duration. The duration magnitude was thrown out here as well.
Matsushiro, Nagano, Japan (MAJO) reported Richter 1935 method 3.6, which was thrown out as inconsistent along with a duration magnitude of 4.4. They contributed only a Nuttli method 2.8. That gives a clue that the Nuttli method is probably pretty inconsistent for this case, but the value was used in the average computation all the same.
Now, here's the meat of the USGS analysis. There are Gutenberg-Richter body wave values from a whole parade of mostly Russian stations: Yakutsk, Russia (YAK); Bilibino, Russia (BILL); Chiang Mai, Thailand (CHTO); Ala-Archa, Kyrgyzstan (AAK); Chkalovo, Kazakhstan (CHKZ); Zerendi, Kazakhstan (ZRNK); and College, Fairbanks, Alaska (COLA). What's interesting is that the Russian stations both had this as a very strong 4, with YAK at 4.7 and BILL at 4.8. They were the outliers on the strong side, and that could explain why the Russians have a bigger estimated yield, if they used only local seismic data. Chiang Mai and Fairbanks came in middle of the road at 4.2 and 4.4 respectively, and the Kyrgyz and Kazakh stations were lower at 3.8-3.9. Having so many distant 3.8+ readings seems to indicate that even if the USGS is a bit too high, 3.6 is probably quite a bit low.
Finally, look at the "azm" column, which shows the azimuth, or the straight-line direction to that particular seismic station from the epicenter. Interestingly, if you divide the stations into two groups, those generally "northward" with an azimuth between 0 and 90 or between 270 and 359, those stations contain all the highest magnitudes for that particular method, mb 4.4-4.8, ML 4.2, and mblg 3.8. Those to the south (azimuth 90-270) are all consistently lower: mb 3.8-3.9, ML 4.1, mblg 2.8-3.7. Distance seems to have less of an affect than azimuth. I have no idea what this means, but there are a few possibilities. It could just be a coincidence, because there's no proof that there's anything more than that going on. It could be that the topography and geology of the site influenced the readings to have this general bias. It could be that the seismometers themselves are prone to different errors in different parts of the globe. Or maybe it could be that more energy from the test went a particular direction. I really don't have the expertise to know if that's even possible.
As far as the fizzle vs. fake argument, I think I have to weigh in on the side of the fizzle. You could theoretically make an explosion this size conventionally, but it would be pretty nearly as hard as building a nuke to make it look believably like a nuke because you'd have to get the timing exactly right. I'm assuming the nuclear signature really is there, because no one is jumping up and down with the smoking gun proving a fake. Speaking non-scientifically, it wouldn't work politically to fake it. I've seen the analogy online that faking a nuke test would be like robbing a bank with a fake gun. This may be true, but the key to robbing a bank with a squirt gun is to not let anyone know its a squirt gun. But anyone who would resort to nuclear blackmail should know that the seismic signature of a nuclear explosion is very distinctive. Faking this with a relatively slow-burning chemical explosion would be like filling the squirt gun at the bank lobby water fountain, because it would quickly become apparent that your threat wasn't real. If Kim Jong Il did fake it, he's dumber than Bush.