The Silence of the Bams.

Is it possible to hide a nuclear detonation?

In the predawn hours of July 16, 1945, an explosion rocked the desert of central New Mexico. The flash of light from the blast lit the sky statewide, and residents felt the shock wave as far as 160 miles away. U.S. Army officials first said a munitions storage area at the Alamogordo Bombing Range had accidentally exploded. The truth was revealed less than a month later, when U.S. planes dropped atomic bombs on Hiroshima and Nagasaki and ended World War II.

The 19-kiloton blast that had shaken residents of the American Southwest--the results of the 3-year, secret Manhattan Project--was the world's first nuclear test.

Since then, eight other countries--the former Soviet Union, France, Britain, China, India, Pakistan, Israel, and South Africa--are known to have successfully developed nuclear weapons. As many as 20 more countries, including so-called rogue nations such as North Korea and Iraq, have sought or may be seeking to develop nuclear weapons. Keeping tabs on who's got "The Bomb," by methods including monitoring seismic rumbles and traces of radioactive fallout, is one of the most critical elements of national-defense strategy.

What if rogue nations or anyone else could test nuclear bombs without the world knowing about it? As it turns out, that might not be so hard. If set off within a cavity of the right size and shape, even a moderate-size nuclear bomb blast might appear to be no larger than a dynamite explosion that miners use routinely to loosen large volumes of rock.

While scientists in some nations may be clandestinely developing nuclear weapons, other researchers are racing to improve methods of remotely detecting and monitoring nuclear tests. Such techniques, if successful, would deny a cloak of invisibility to groups attempting to covertly become a member of the nuclear club.

Detecting a nuclear explosion used to be relatively easy: bright flash, big boom, mushroom cloud, lots of radioactivity. Then, nuclear testing was literally driven underground. The 1963 Limited Test Ban Treaty, which more than 115 nations have adopted, prohibits nuclear tests in the atmosphere, underwater, and in outer space.

A 1996 follow-up to that treaty, the Comprehensive Nuclear Test Ban Treaty, would go even further and prohibit all nuclear explosions in any environment. Of the 161 current signers of this latest treaty, 76 have ratified it. Although the United States was one of the first in line to sign the comprehensive treaty 5 years ago, the Senate has yet to ratify the accord. Critics of the treaty believe that it's unverifiable. They cite concerns that a rogue nation could covertly develop nuclear weapons and test them and nobody outside that country would know about it.

That's why a reliable monitoring system for all nuclear tests is so crucial to the comprehensive treaty. As specified by the accord, an international monitoring network would comprise 321 stations scattered across the globe. Some 60 stations in the worldwide system would detect the minute, low-frequency variations in air pressure that might be associated with aboveground explosions. About 80 others would sniff the air for radioactive fallout, and 11 instruments deployed in the oceans would be alert for hydroacoustic pressure waves generated by explosions in or just above the water. The remaining 130 sensors would listen for seismic vibrations that might have been generated by underground explosions.

More than 100 of these four types of sensors are now collecting data, and as many as 60 new stations are scheduled to start transmitting data this year, says Steven R. Bratt of the Vienna-based Comprehensive Nuclear Test Ban Treaty Organization.

There are three big challenges to remotely monitoring possible nuclear tests and verifying the comprehensive treaty, says Anton M. Dainty, a geophysicist at the Defense Threat Reduction Agency (DTRA) in Dulles, Va. First, the sensors must detect a signal associated with a possible nuclear test. Then, the source of that signal must be located. Finally, what triggered the signal must be identified as either a natural phenomenon or a human activity.

Currently, DTRA is funding about 60 contracts for basic research and development related to these three challenges. Over the past 3 years or so, these contracts have totaled about $10 million per year. The lion's share of the money has gone to seismic research, Dainty adds. In particular, DTRA is now spending about $9 million on three different contracts to determine more accurately the source of earthquakes and large explosions in central Asia, northern Africa, and the Middle East.

The trick with seismic waves is distinguishing bomb blasts from earthquakes. Extremely deep sources of rumbles and those beneath the ocean are readily attributable to earthquakes. But seismologists must carefully analyze ground motions from shallow sources on land to know what caused them.

Seismic vibrations travel both in push-pull compression waves and side-to-side shear waves. Earthquakes and explosions distribute their energy differently between these two types of waves, a characteristic that can help scientists tell the sources of vibrations apart.

Other factors can help, too. Many detonations used in mining consist of a series of explosive charges fired sequentially. This so-called ripple firing provides a distinct seismic signal that positively identifies the source of the ground vibrations as humanmade explosions.

Such close analysis doesn't always do the trick, however, because the seismic signals from some natural phenomena can mimic explosions, says John R. Murphy, a Reston, Va.-based seismologist at SAIC, an engineering and research firm.

For example, rockbursts--the violent, accidental blowouts of rock under high stress that occur often in mining tunnels--can generate ground vibrations that can't be distinguished from those triggered by large explosions. Rockbursts are particularly vexing because they happen at relatively shallow depths, just where bomb developers might perform a clandestine nuclear test.

Even though scientists can distinguish between the ground waves caused by earthquakes and those from explosions, it's often difficult if not impossible to tell a nuclear blast from a chemical explosion, such as one caused by TNT. Murphy notes also that it's particularly difficult to analyze small seismic signals because the characteristics that distinguish one type of vibration source from another can be overwhelmed by background noise.

There are additional complications. Ground motions can appear small to detectors because the source is far away. However, says Murphy, such vibrations can also seem small because they come from a blast set off in a large, empty cavern. A bomb far from the walls of the cavern doesn't shatter the surrounding rock, a process that would send out distinctive seismic vibrations. The energy from the explosion is instead transferred gradually to the walls of the cavern, which then vibrate in ways that can mimic an earthquake. Scientists say that such explosions have been decoupled from the cavity they're in.

Not only does cavity-decoupling remove a nuclear explosion's fingerprint, but it also sends out ground vibrations much smaller than expected. For example, a cavity-decoupled nuclear explosion in a cavern of salt would seem as small as one-seventieth its actual power. In other words, someone conducting a clandestine nuclear test in a large enough cavity could make a 1-kiloton nuclear bomb appear as if it were only 14 tons of a chemical explosive.

The size of the cavity needed to fully decouple a nuclear explosion depends on the size of the bomb, the depth of the cavern, and the material into which the cavity has been mined. For a small nuclear explosion in hard rock, such as granite, a 20-meter-diameter spherical cavity would do the job, says William Leith, a geologist with the U.S. Geological Survey in Reston, Va.

A spherical cavity is more difficult to dig out than an oblong one. However, calculations say an oblong cavity of similar volume will cloak a nuclear explosion as well as a spherical cavern will, Leith says.

Digging large oblong cavities is quite doable. To house hydroelectric turbines, engineers in Japan and Indonesia have excavated such caverns in rock with unsupported roof spans of 35 m. The Chinese have built an underground aircraft hangar in rock whose ceiling spans 42 m.

The largest such unsupported cavern, excavated by the Norwegians to house an Olympic sports hall, has a 61-m roof span. By stringing together a series of such sports halls, it might be possible to excavate a cavern large enough to fully decouple a 10-kiloton nuclear explosion from its normal seismic signature, Leith says.

Hiding an excavation project of that scale from prying satellite eyes would be difficult, however. Engineers would have a hard time surreptitiously dumping all the rubble coming out of such a hole. For that and other reasons, Leith says, it would be much easier to cloak a nuclear test inside a large underground salt deposit.

At the temperatures and pressures found hundreds of meters underground, salt has the strength of concrete and is impermeable to most liquids and gases, says Leith. Furthermore, under extremely high pressure, salt will slowly deform and flow, which means the walls of a cavity would be self-sealing, preventing radiation leaks after the test.

Although salt can be mined just like rock, it's much easier to create a cavity by using a technique called solution mining. In this method of excavation, engineers drill an 8-to-10-inch hole down to the level at which they'd like the cavity. Then, they pump water into the hole to dissolve the salt, and pump out the brine that's created. All that's needed is an adequate supply of relatively fresh water--although even seawater will do, Leith notes--and a place to dispose of the brine.

Engineers can excavate huge cavities with the solution mining technique. That's how they've created more than 50 of the caverns that hold the United States' strategic petroleum reserve. On average, each of those cavities is about 60 m in diameter and 600 m tall--enough to hold 10 million barrels of oil or both towers of the World Trade Center in New York. The largest such cavern, however, is more than eight times that volume.

The equipment used to conduct solution mining is both simple and unobtrusive, no more than a few pumps and narrow pipes at ground level. Salt-mine excavators could easily hide most of that equipment in a small building. Typical solution-mining operations pump brine into large surface ponds. However, engineers conducting a clandestine nuclear testing operation could easily hide that part of the operation by pumping the millions of gallons of saturated brine into an underground aquifer, says Leith.

Solution mining is both quick and inexpensive. By pumping about 600 cubic meters of water each hour, engineers can excavate a cavity of up to 400,000 cubic meters in a year's time. It would only take about $2.3 million and a little more than 3 years to mine a 1.3-million-cubic-meter cavity in this way. That's the size needed to fully decouple a 20-kiloton nuclear explosion. This cost is a pittance compared with the overall budget of a nuclear-weapons program, says Leith.

An aspiring nuclear power could hollow out a salt cavity and then fill it with oil to keep it from collapsing before a bomb is ready for testing. It's conceivable that within just a few weeks, workers could empty an already excavated salt cavern detonate a nuclear explosion inside, and then refill the cavity.

Such a scenario is conceivable, say critics of the Comprehensive Nuclear Test Ban Treaty. Senators Trent Lott (R-Miss.) and Jesse Helms (R-N.C.) have expressed concerns that the treaty's network of sensors may be inadequate to detect all nuclear tests, especially those that are cavity decoupled. Indeed, then-Senate Majority Leader Lott contended during the Clinton Administration that the President's push for the treaty actually accelerated the weapons programs of India and Pakistan, both of which tested nuclear devices underground early in 1998.

SAIC's Murphy points out that the network's 100 or so sensors that are now operational can identify as natural phenomena only 65 percent of seismic events of magnitude 3.5 or higher. Determining the cause of smaller seismic events is even less certain, he adds.

Formal complaints under the comprehensive treaty can be supported only by evidence from the treaty's official network. On the other hand, nations can independently look for signs of nuclear violations using a variety of instruments and analysts beyond the network, says Gregory E. van der Vink, a geophysicist at Princeton University. These include satellites, spies on the ground, and the regional networks of seismometers operated by governments, universities, and other organizations around the world.

More than 16,000 seismometers are installed in networks worldwide, so usually one of these instruments will be closer to any clandestine test than any of the 130 seismometers dedicated to treaty verification will be. That proximity means that the unofficial test-monitoring devices might have a higher probability of finding out whether the source of the vibrations is natural or the result of human activity. In August 2000, for instance, a network of seismometers in Scandinavia recorded seismic pulses that helped scientists determine that the Russian submarine Kursk sank as a result of explosions, not a collision with another vessel (SN: 1/27/01, p. 53).

Van der Vink is convinced that more-open sharing of data from such instruments outside the official network could play a vital part in monitoring the test-ban treaties. Rapid access to such data via the Internet could immediately alert nations worldwide to suspicious seismic vibrations, which then could be further analyzed with data from the official treaty-verifying network.

"It's quite likely that if evidence of a clandestine nuclear test is ever discovered, it will first come from data collected by instruments that were never specifically intended for nuclear monitoring," van der Vink notes.