True Detectives

May 3, 2018

The following is an excerpt from Middlebury Magazine

The treacly martial music plays in the background while Kim Jong-un, North Korea’s leader—the third Kim to rule his country, the world’s only hereditary Stalinist dictatorship—looks out on water. He is on a boat, bobbing up and down, dressed in a black peacoat and wide-brimmed hat. The mountains behind the sea are lavender; the sky is faded rose gold. It is dawn, or dusk, or nearly either. Suddenly, a missile pierces the choppy water. As it shoots upward, it looks like a warped sunrise, the engine’s blast reflecting over the surface of the water. It surges toward the cloud line, slices it, and rises toward the darkness.

Left to right: Dave Schmerler, Jeffrey Lewis, and Melissa Hanham

Left to right: Dave Schmerler, Jeffrey Lewis, and Melissa Hanham

The narrator on KCTV, North Korea’s official television network, is rapt. The test of the submarine-launched ballistic missile, or SLBM, appears to be a success. The launch is shown from multiple perspectives. Kim Jong-un smiles for the camera.

But here’s the thing: the test was a composite. It was doctored. A fake. And we know this not thanks to the U.S. government or its allies; or the International Atomic Energy Agency (IAEA); or the United Nations; but because of the keen detective work of the researchers and analysts in the East Asia Nonproliferation Program at the Center for Nonproliferation Studies (CNS) at Middlebury’s Institute for International Studies at Monterey (MIIS).

This group—led by Jeffrey Lewis, the program’s director—knows more about North Korea’s nuclear weapons program than almost anyone on this planet not currently working for a government agency. “Time and time again, the reason why people tell me that CNS is the ‘dream team’ for North Korea analysis is that we have people with skills across the board here that we brought together,” says Andrea Berger, a London-based CNS analyst who specializes in illicit finance networks. “That has resulted, I think, in some really special research.”

Berger’s estimation is widely shared across the nonproliferation community. “No group has influenced the public discourse on North Korea’s nuclear weapons program and progress more than Jeffrey Lewis and his colleagues,” says Vipin Narang, an associate professor of political science at MIT and prominent nuclear-weapons expert. “They figured out how to put all the pieces together with readily available tools, like nuclear detectives. Knowing the launch location for a test or one that failed used to be the monopoly of intelligences agencies. They single-handedly broke that monopoly.”

Lewis and company, moreover, helped revolutionize their field while possessing only a fraction of the tools available to the world’s spy services and an iota of these organizations’ human capital. The entire North Korea team at CNS consists of fewer than 10 experts. (This number excludes CNS’s graduate students, who provide key research support. I met one typically atypical student, Grace Liu, down in Monterey. When Liu, who is fluent in Chinese and Korean, isn’t engaging in geospatial analysis at CNS, she is a military intelligence officer in the National Guard, overseeing a unit of nearly 300 soldiers.) The CNS teams have no special access or trove of classified data from which to draw. They use entirely open-source methods—commercial satellite imagery, North Korean propaganda videos and photos, social media, business records, and other tools—for their work.

“Look at the shape of the mountains behind the missile,” says Melissa Hanham, a senior research associate at CNS, as she points to her computer screen. We’re watching footage from the aforementioned submarine-launched missile test. I am in Hanham’s office in Monterey on a sunny, cool February day. It’s a neat space—most certainly by academic standards—and I find the books and papers that threaten to encroach on this order a comfort in their own right.

There’s a framed North Korean propaganda poster on the wall, covered in sticky notes, with phrases like “Syria Chem” and “Ship Tracking” and “Pyongyang Biotech Institute” on them, reminders that now threaten to obscure this particular Socialist realist paean to the working man.

On another wall is a small, framed poster containing a tiny mounted piece of wire from the Demilitarized Zone, or DMZ, which buffers the border between the two Koreas. (The DMZ is the product of the 1953 armistice that concluded the Korean War; a formal peace treaty between the two Koreas was never signed, meaning they are still technically at war.) When I first walk into Hanham’s office, there’s Shoegaze or Dream Pop wafting through her computer speakers. It’s a disarmingly pleasant place to tackle one of the world’s most exigent security problems.

The East Asia nonproliferation team broke down this particular SLBM test in stages. First, says Hanham, one of her colleagues, Dave Schmerler, geolocated the mountains in the video from the missile test, using Google Earth to identify from above what the regime gave the rest of the world a glimpse of from below. Now CNS’s researchers had a pretty good idea of where the test took place. Schmerler—who is by his own colleagues’ lights a savant among savants—also happened to remember that he had seen some of the exact same footage (the part where the missile pierces the cloud line), during a 2014 North Korean Scud missile test. Had the North Koreans spliced footage from two separate tests together? Something, they knew, was off.

Another CNS colleague, Catherine Dill, separated the three clips of the missile launch (North Korea showed what appeared to be three separate shots of the same test), which allowed them to be played together, side-by-side. In order to compare these clips, she sped up or slowed down each frame to make sure they were advancing at the same pace, because the North Koreans had tweaked the speed of the videos. Dill discovered that one of the clips was a cropped and flipped version of another—in other words, these were the same shot, from the same camera, edited to appear as different. (North Korea is often “selectively transparent” about its nuclear program, says Joshua Pollack, a Washington-based CNS analyst, and open-source researchers can learn a lot more than might be expected from a country often described, inappropriately in his view, as a “hermit kingdom.”)

One of these three clips had more total frames of video than the two others did. What Hanham noticed was that, in the last few slowed-down seconds of the longest video, the entire missile becomes engulfed in flames. If you look closely enough, you can even see little chunks flying off the missile as it explodes after launch. “The video editor who put together this propaganda just left too many frames in,” says Hanham.

Still, says Hanham, to be sure that the North Korean test deviated from what a successful one would look like, they needed to compare it with the launch of a similar submarine-launched missile. Hanham and her colleagues believed that the North Korean missile was based on an older Soviet model. So they found an old Soviet test video and juxtaposed the two, sizing both videos to scale. In both, the missiles burst from the water, vapor spewing forth. But there was no hint of an explosion in the Soviet test. The North Korean test, they now concluded, was a failure.

Continue reading at Middlebury Magazine

Comments Are Closed