,
7 tweets,
2 min read
Read on Twitter
I was wondering when a story like this would appear - although I am not sure it quite gets at the nature of the problem. Two IMS stations (Dubna and Kirov) reportedly stopped transmitting two days after the accident.
wsj.com/articles/russi…
wsj.com/articles/russi…
Two days is a long time. The bigger question, to me, is why did no stations in Russia detect the event? Here is the initial @ctbto_alerts release -- the screenshots contain the names of the four stations that detected the explosion on August 8.
All four (FINES, ARCES, HFS and IS37) are in Nordic countries (red pins). Using distance as a proxy for likelihood of detection, four stations in Russia might have been expected to also detect it. AFAIK, none did. 
Cautions: The event was very small. Distance may be an imperfect proxy. Further review may find that other stations did see it. And one Nordic station, Hamar, seems to have missed it. I don't think one can prove deliberate Russian malfeasance at this point.
If someone did attempt to compromise the data, we should condemn them -- but mostly for being stupid. The IMS is far too robust to be compromised by turning off a few stations. Trying just makes one look guilty. This was a test of the IMS. The IMS passed.
(Making guilty parties look guilty is an under-appreciated aspect of verification, especially in cases where there is an inspection regime. Imagine if, after an alleged nuclear test, a state turned off its stations. You no longer need the inspection.)
There are so many sensors in so many countries that even without the Russian data -- for whatever reason that data is missing -- the IMS still detected an absurdly small event with four stations in three different countries. That's an IMS success story.
