Monday, October 09, 2006

Seismology of a nuclear test

The North Koreans claim to have tested a nuclear bomb at about 10:30 this morning (11:30am eastern Australia time). Can geology tell us anything about what actually happened? Let us see.

First stop, the comprehensive test ban treaty organization. In theory, they have a monitoring network designed to detect nuclear tests. Unfortunately, as a UN organization, their webpage is full of schemes and classifications and plans and protocols, and as of 6 this evening, there is no data, and no press announcement or other hint that anything happened.

The Geoscience Australia earthquake page is great for small quakes near Australia, but does not contain any information on any events today.

What about the evil, super-secret world-dominating Americans?

The USGS earthquake list shows a little orange square in NE north Korea. Clicking either it or the “M2.5 to 4+ earthquake list” gives the following quake listing:




MAG UTC TIME LATdeg LONdeg DEPTHkm Region
4.2 2006/10/09 01:35:28 41.294 129.134 0.0 NORTH KOREA

Clicking the appropriate entry gives the summary page for that earthquake

So, what can we see? First, the depth of the earthquake is listed as 0 km. Most of the earthquakes in this region are very deep (see “historical seismicity” under the “maps” tag and look at the color scale). These deep quakes happen in the downgoing slab, as the old pacific plate subducts beneath Japan. Thus they are quite different than the zero depth M 4.2 event from this morning.

What else can we determine? In theory, an explosion should have an all-positive first motion tensor. Unfortunately, first motion tensors only seem to be calculated for quakes larger than M 5.0, so that test will need to wait until the North Koreans detonate a larger bomb. It might be possible to put a DIY tensor together by looking at the raw data, but I don’t know where that is archived.

Similarly, I seem to recall from my deep, dark past that explosions have a higher P/S wave ratio than normal quakes. But once again, this will require having the S wave data, and I don’t know where to find that, either. Hopefully some of the real seismologists who occasionally visit the lounge can give us some hints.

The only conclusion that I can draw is that there was a M4.2 event on land in North Korea with a zero depth, which was dissimilar to other seismicity in the area.

9 comments:

CJR said...

New Scientist say that the USGS have confirmed that event as being a nuclear explosion. Maybe when they all get out of bed they'll give us more.

Lab Lemming said...

So far they've just moved the event from the "earthquake" to the "headline" column. But I'd really like to get to the data and show people what it means and how to interperet it, rather than just link to an expert opinion. Science is supposed to be repeatable, after all.

I found a japanese site for wave forms here:
http://www.fnet.bosai.go.jp/freesia/waveform/prev.html
But they seem to be running 3 days behind- only up to the 7th, from what I can figure.

yami mcmoots said...

I had a look at some of the data from here in Berkeley, but either the signal is too weak to see or I'm pulling something up incorrectly. However, I get the feeling that our glorious leader here at seismolab would chew the face off of any second-year graduate student who posted waveforms on her blog before the grownup seismologists have finished their analysis...

Lab Lemming said...

Is Berkeley between 103 degrees and 143 degrees from N. Korea? 'casue there's that P / PKP shadow zone, if I recall correctly. Anyway, I would think that chinese, japanese, or russian data would be best, but I don't know how to check.

The station arrivals are listed here:
http://neic.usgs.gov/neis/bulletin/neic_tqab_p.html
but I won't have time to look them up until after work.

Do you reckon it would be possible to make a good first-year lab out of all this once we find all the neessary data and software?

yami mcmoots said...

Nope, we're 78ยบ away. I tried Berkeley first just 'cause it's trivially simple for me to get data from BDSN... I can also download from GSN, but I haven't done it before so don't know quite how obnoxious it is.

What would you want to teach with this lab? I think it would make a great trick question on an exam...

Lab Lemming said...

As lab, it would be a good final question for a first-moment tensor plotting exercise. I actually put my Fowler textbook in my bag to bring home in order to have a play this evening, but it was such a crazy day at work that I just couldn't face recreational seismology this evening- more of a beer&TV sort of a day.

But western Geologist has some squiggles posted:
http://westerngeologist.blogspot.com/

Maria said...
This comment has been removed by a blog administrator.
yami mcmoots said...
This comment has been removed by a blog administrator.
Lab Lemming said...
This comment has been removed by the author.