Castle Romeo test, Bikini atoll, March 27, 1954
North Korea pops second nuke; test is immediately detected
On May 25, North Korea set off its second nuclear bomb deep under ground. Once upon a time, the nuclear powers showed off their latest and greatest in awesome, earth-shattering, aboveground tests. But because the radioactive fallout from those explosions could cause cancer, most tests have been underground since the early 1960s.
That reduced fallout, but because a ban on nuclear testing is widely considered a key step in slowing the nuclear arms race, obscuring the tests had serious arms-control implications. After all, every nuclear nation (except probably Israel) has tested its bombs to ensure that they will not fizzle. And so, as President Dwight Eisenhower realized in the late 1950s when he tried to negotiate a halt to nuclear tests, banning tests could slow the nuclear arms race.
Now, with North Korea definitively in the "nuclear club," we have two questions: How are nuclear tests detected? And how easy would it be to cheat if and when a comprehensive test ban finally takes effect? Aboveground tests are visible to satellites and special microphones, and underwater tests can be heard by 11 strategically situated submerged microphones. But the first detection of an underground test comes from seismometers (AKA seismographs), instruments built to detect earthquakes, and then can be confirmed by the discovery of rare isotopes that are traceable to nuclear bombs.
A telltale shaking
Within minutes of North Korea's May 25 test, earthquake experts had seen ground shaking on a series of seismometers. Within hours -- well before North Korea began to boast of its achievement -- governments around the globe had been alerted to the explosion.
As the seismometer was invented in China in the year 132, it's fitting that some of the best data on the North Korean test came from an instrument in China. Seismometers are now so sensitive that they can detect footsteps and traffic -- and movements of the earth's crust on the other side of the world.
To understand a seismometer, imagine a ball suspended from a miniature gallows, then mentally tap the gallows sideways. Because momentum will momentarily hold the ball still, you can measure its motion relative to the apparatus, which will reveal the direction and size of your tap.
How a seismometer works
Now multiply the sensitivity and price by a few million, place the seismograph on solid rock a long way from traffic, mines, and other local sources of earth movement, hook it to a computer and a satellite transceiver, and you have the general layout for a seismograph that can detect earthquakes and explosions, whether chemical or nuclear.
Earth movement causes several types of waves, including pressure (P, or compression) waves, and shear (S) waves. Pressure waves are alternating waves along a line between us and the source, while shear waves are movements at right angles to this line.
If you stretch a slinky and move it toward and away from you, you get P waves. If you shake it side to side, you get S waves. Both P and S waves move through the body of the planet; surface waves (AKA Rayleigh waves) move at the surface, and take longer to reach the seismometer.
These distinctions matter, because earthquakes and explosions create different types of movement: An explosion pushes outward in all directions, creating a storm of pressure waves, while an earthquake, which largely involves earth slipping side to side, creates more shear waves.
Surfing that wave
Mine collapses are one seismic event that could be mistaken for an explosion, because both make a lot of pressure waves, but the seismograph line starts to fall with a mine collapse, while it rises after an explosion.
Eliminating mine collapses and earthquakes leaves you with an explosion, but is it chemical or nuclear? Here, size matters. A small nuclear explosion, like North Korea's 2006 test (estimated at 0.6 kilotons of TNT), could conceivably be faked with chemical explosives. But even if you manage to detonate 600,000 kilograms of high explosive all at once, digging the requisite hole would probably be visible to a spy satellite.
We asked Clifford Thurber, a professor of geophysics at University of Wisconsin-Madison, about seismic detection of nuclear blasts. Thurber, who has explored how seismic signals can be used to locate nuclear tests, says estimates of an explosion's size can be "shockingly variable," because the intensity of seismic signals greatly depends on the Earth's internal structure, which is "not a spherical ball of uniform material."
Thurber says seismometers can also locate explosions and earthquakes through a process of triangulation, much as a GPS receiver "finds" itself on Earth's surface. "We want to have stations distributed as well around the event as possible." Ten stations well distributed and close to the seismic source would probably be adequate, he adds, "but more typically, we would use 50 to 100 to establish the location and the size."
Vertical ground movement recorded at a seismic station in China
A history of rock
It also helps to have a seismic history of the region, says Won-Young Kim, a seismologist at Columbia University's Lamont Doherty Earth Observatory. Kim says he was at work in 2006, when he heard a computer beep that signaled some seismic event, somewhere. He knew that North Korea had recently announced its intention to set off a nuclear bomb, and had also seen the country's test sites in a newspaper map.
Within two hours, Kim and colleague Paul Richards had compared the seismic signal to other geologic events on the peninsula. "We used those as benchmarks, and looked at the character of the seismic signal," Kim told us. "It became quite obvious that the signal from North Korea was similar to an explosion, it had a lot of P waves, but not much S. That convinced us that there was more than an 80 percent chance this was nuclear."
Richards quickly told the newspapers that the seismic event probably came from North Korea's first nuclear test.
In 2009, the international system of seismometers was more complete, "and it worked a little better," says Kim. "In three years, we learned some lessons about how to look at things in that part of the world, relative to earth structure and the location of seismic stations." Although the monitoring network is still under construction, he says, "even the incomplete system was able to provide enough information."
Kim, who co-authored a recent article on nuclear detection (see #1 in the bibliography), says he's not privy to any secret military information, and provides expert commentary on nuclear detection as a public service. "We are just scientists. We do this kind of analysis to help verify a test ban treaty, which we believe is important for all human beings. We scientists and engineers developed the bomb, and now we are trying to ensure it will not harm anyone."
Yet, as Thurber notes, "seismic data by itself will only be suggestive, it will not be conclusive" in proving that an explosion was nuclear.
On the importance of detecting nuclear tests
Since the late 1950s, many experts have asserted that banning nuclear tests is an excellent way to control the nuclear arms race. In 1963, the Partial Nuclear Test Ban Treaty prohibited all aboveground tests, making monitoring more difficult.
The United States has signed and is observing the Comprehensive Test Ban Treaty (CTBT), but the Senate has not ratified it. The CTBT must be signed by North Korea, India and Pakistan before it can become effective.
In the meantime, the CTBT Preparatory Organization has networked 130 seismometers to detect underground explosions in its International Monitoring System.
Terry Devitt, editor; Nathan Hebert, project assistant; S.V. Medaris, designer/illustrator; David Tenenbaum, feature writer; Amy Toburen, content development executive