Thread by @ArmsControlWonk

So, @annemp13 and I built an annotated tour of Russia's Novaya Zemlya nuclear test site. I wanted to talk about what has been one of my favorite all-time projects.
view.ceros.com/miis-edu/nucle…

It is REALLY hard to get good satellite images of Novaya Zemlya. First, the test site is really far north. The main support area sits at a latitude of 73°23'. (The arctic circle starts at 66°34'.) There isn't enough light half of the year and, when there is, its usually cloudy.

We couldn't even buy a recent picture for @geo4nonpro. We had to task an @ImageSatIntl satellite to take one for us. Pass after pass, the damn place was always cloudy. Eventually we got a composite from different days where there was enough light and not many clouds.

Later, our friends at @planetlabs started taking 80 cm color shots. They got LOTS of them. For the first time, we had lots and lots of good images of the site.

(Now that Russia is testing nuclear-powered cruise missiles on Novaya Zemlya, most companies are tasking satellites to take pictures of the area. I like to think @mhanham and I started a trend.)

But how do you know what's new or unusual? Establishing a baseline meant going back and doing archival work -- we wanted to identify when each and every adit was built. Which turned out to be a ton of work.

For one thing, Soviet-era historical sources don't always agree. These are two Soviet maps of the test site helpfully marked with every test. One problem. THEY AREN'T THE SAME. 😂

The seismic data, while helpful, isn't accurate enough for to link specific events to specific tunnels, but it does help with the general location of tests. Although this paper by Paul Richards is an incredible resource.
ldeo.columbia.edu/res/div/sgt/vi…

.@annemp13 and I started purchasing declassified satellite images from @USGS of the site dating back to the 1960s. Some of those pictures are just gorgeous.

By the way, we weren't the first people to try and do this. @nupinytt did some incredible and ground-breaking work two decades. Go ahead and do yourself a favor and read this report by Johnny Skorve.
nupi.no/en/Publication…

The Soviets did a massive test that caused a landslide that buried their test cam in rubble and created an artificial lake. Skorve discovered it at a time that it was still secret. You can see it clearly here in this @planetlabs 1-month mosaic.

Skorve was also able to point out features (spoil, rock slides, craters) created from nuclear tests by comparing modern images to an aerial reconnaissance image taken by the Luftwafffe in 1942. (It's not 2018 without Nazis.)

We also found Soviet-era films that are incredible. This one, НЗ технология подземных испытаний, is fricking incredible -- and the soundtrack is hella-creepy. My favorite clip starts at 18:30.

One of the interesting things is how many of the Soviet tests vented. It's strange because some of the Soviet documents talk about choosing this site because it offered better containment, but the primary advantage appears to have been remoteness.
pubs.usgs.gov/of/2001/0312/r…

We also took good advantage of social media. You'd be amazed what kind of sensitive national security information people post on line. Or then again, maybe you wouldn't.

So what? First, some people in the United States want to resume underground explosions. Trump ordered @ENERGY to be ready to do so in as little as six months. A quick look at Novaya Zemlya suggests Russia could follow suit if it chose to do so.
time.com/5128394/donald…

Russia, unlike the United States, has ratified the CTBT. But I don't think anyone expects the Russians to let a resumption of US nuclear testing go unanswered. @SecDef19 has made this point repeatedly -- and satellite images suggest he's right.
nytimes.com/2016/04/17/sci…

Second, the lack of public data about the test site has resulted in some intelligence failures. In 1997, the US démarched the Russians over a nuclear test -- that later turned out to be an earthquake. Oops.
eps.berkeley.edu/~rallen/pub/19…

A lot of the anxiety about Russian activities in the late 1990s was based on questionable photo-interpretation that civil society couldn't see or assess. Today, my colleagues @CNS_Updates could tell you that there are helicopters on site all the damned time.

Third, satellite images show us a pathway toward ratification of the CTBT. We have been repeatedly told that the US can't sign the CTBT because Russia might be conducting low-yield nuclear tests.
nytimes.com/2001/03/04/wor…

Satellite images show what we suspect is a dedicated subcritical testing facility constructed in recent years. That judgment is based on the new tunnel, support buildings that suggest continuous usage, and a (vague) announcement from the Russians.
nti.org/gsn/article/ru…

If so, this facility serves a similar function to the U1A complex in Nevada. To be clear: These facilities do not violate the CTBT. But they pose a challenge, because the US and Russia each thinks the other might be up to no good.

If the United States and Russia have comparable sites for conducting no-yield tests, then one might easily imagine designing transparency and confidence-building measures for these locations.

Some years ago, Frank von Hippel and Suzanne Jones wrote an article outlining a technical approach to monitoring no-yield testing. The technical approach is straightforward -- if you know where the events are to take place.
scienceandglobalsecurity.org/archive/sgs06j…

Our colleague @sbidgood has made much the same point, arguing to allow experts to observe activities at these locations and agree on a bilateral protocol if either country were concerned about activities outside of these two sites.
tandfonline.com/doi/abs/10.108…

Envisioning transparency and confidence building measures for subcritical testing in Russia is hard when you don't know what the site looks like. But when you do, suddenly a world of new possibilities appears. /END