,
8 tweets,
4 min read
Read on Twitter
Joel is deeply confused. A short thread.
First, there are *many* images of Sanumdong from the period in question, not "a single snapshot." By my count: 3 high-resolution shots (February 22, February 24 from @DigitalGlobe and March 8 from @planetlabs) and daily 3 m images over the same period from @planetlabs.
Second, this statement is silly. If we see vehicles, cranes, etc. on one day, what would we conclude from their absence two days or two weeks later? That they were never there? This is called "object permanence." It's a big developmental step for infants.
As it turns out, @planetlabs has a lovely March 8 image showing that Sanumdong has returned to normal activity levels. Whatever work was underway has now been completed. It doesn't mean all the rockets the plant has produced in the past have ceased to exist. 
If the rocket headed to Sohae, we might not see it's arrival. North Korea covered both rail transfer points at Sohae, impeding our monitoring of the site. The mobile gantry building that North Korea is rebuilding is one of the structures that makes it hard to monitor Sohae.
While we might get other indications -- such as a notifications to the IMO or a statement from KCNA -- we might not. North Korea might try to hide the precise timing of the launch, something they did successfully in December 2012.
nytimes.com/2012/12/12/wor…
nytimes.com/2012/12/12/wor…
All of which to say, this is hard because we're trying to surface things North Korea probably isn't quite ready to share. We can all be more careful about expressing our confidence levels and should be open to new evidence. But the activity at Sohae and Sanumdong is striking.
So, my advice to Joel is stop telling people what "experienced analysts" do and do not think. You're not good enough at this to be a dick about it.
