Abstract
Some argue that a nuclear test ban is not verifiable. Meanwhile the test ban treaty's monitoring system of 321 stations, now well on the way to completion, is looking good.
For almost three years, representatives from more than 40 nations met in a cavernous chamber of the little-known Palais des Nations in Geneva, negotiating the Comprehensive Test Ban Treaty (CTBT). They labored over membership obligations, rules for entry into force and compliance, and the politically sensitive question of verification.
The Group of Scientific Experts, which had spent 17 years considering the technical issues that would be involved in verifying a nuclear test ban, however, was able to provide the necessary expertise for the design of a modern, comprehensive monitoring system based on experience in seismic monitoring. This was supplemented by expert work in ra-dionuclide, infrasound, and hy-droacoustic monitoring.
The result is an unprecedented verification regime that provides information that no single country can collect on its own. It is designed to provide the international community with a high degree of confidence that clandestine nuclear explosions will be detected.
Unlike other arms control agreements–the Chemical Weapons Convention, the Intermediate-Range Nuclear Forces Treaty, and the Strategic Arms Reduction treaties–this treaty, adopted by the U.N. General Assembly on September 10, 1996, does not rely on an on-site inspection regime as the only, or even the major, means of verifying compliance. Instead, the treaty will depend heavily on a major new creation, the International Monitoring System (IMS)–a globally linked network of 321 monitoring stations located in more than 90 countries. That system will utilize three related acoustic waveform techniques–seismic, hydro-acoustic, and infrasound–and two radionuclide techniques for the detection of airborne radioactive particles and gases.
Mules carry equipment to the monitoring site on Juan Fernandez Islands, Chile. At left, the Dundas, Greenland, monitoring site.
The negotiators in Geneva knew it would be critical to track both the shock waves from an explosion and the nuclear debris. Both would be generated regardless of the environment–solid earth, ocean, or atmosphere–in which a nuclear explosion occurred: Shock waves transform into acoustic waves some distance from an explosion; nuclear debris migrates in the atmosphere, and to a lesser extent in ocean water and even solid earth.
The acoustic waves move very fast. This is especially true for the seismic waves, which may reach monitoring stations within minutes after an event. In contrast, the movement of air masses in the atmosphere is slow. It may take weeks before radioactive debris reaches a monitoring station, so a long time is needed for data collection and analysis. And although attempts to trace the atmospheric transport path back may not contribute to locating an event, the detection of radioactive debris in the atmosphere remains the most conclusive evidence–the “smoking gun”–of a nuclear explosion.
Borovoye, Kazakhstan.
During the treaty negotiations many other technologies were proposed, including satellite-based, optical, and electromagnetic pulse-detecting techniques. But high costs and political arguments prohibited their inclusion.
The International Data Center
Data from all the monitoring stations around the world will be sent continuously and in real time to the International Data Center in Vienna, Austria. The center will produce compilations, usually referred to as “bulletins.” Treaty members may receive both raw and processed information.
The center's challenge is to scrutinize and distill all the data to capture any indication that might be deeply buried in background noise, indicating that an “event”–an earthquake or an explosion–has happened somewhere.
A distant event's imprint can be found in the data at individual monitoring stations using well-established methods and algorithms, including digital filtering to enhance signals within a particular frequency range and so-called beam-forming to improve the signal-to-noise ratio for signals recorded on array stations.
Listings of detailed information about each detected signal–the exact time, the direction of approach, and the speed at which the signal traverses an array station–are not usually very useful from any individual station. But when they are combined with the detection lists from all the other stations in the network, they can be recognized and grouped together as having a common origin. This is a complicated, non-linear process closely linked to the estimation of the location of an event. Using the exact time of arrival at each station of the various seismic wave types originating from the event, as well as knowledge about the speed at which seismic waves travel through the earth's interior, one can derive the event's location and time of origin, an estimate of the depth at which it occurred and its magnitude, as well as uncertainties associated with those parameters. The same procedure, in principle, is used to locate events from hydroa-coustic and infrasound signals.
Solving sticky problems
Most procedures for locating earthquakes and underground explosions using data recorded globally are based on the assumption that the Earth is a homogeneous sphere, and that the speed of seismic waves varies only with depth. But experience has shown that speeds vary in the upper mantle and crustal layers of the earth, producing time differences for certain types of waves of as much as five to 10 seconds. These variations in travel times can introduce substantial errors, throwing off the location of an event by as much as 100 kilometers.
This bias must be removed by calibration based on empirical observation. The treaty recognizes the need for calibration, and the international scientific community is engaged in this important work. Chemical detonations, for which the exact firing times and coordinates are known, are being recorded at various monitoring stations. As calibration data is gathered and incorporated into databases, the networks will be able to pinpoint locations.
Processing data
The data center will generate lists of events using a combination of automatic and manual analysis. Because the incoming data stream is vast, the system must be automated to the greatest extent possible. However, the well-trained human eye is crucial to check that no important piece of information is overlooked.
The center will have a staff of about 100 persons to support its functions, including the production of a daily “event bulletin,” which is expected to report an average of 100 or more earthquakes per day, in addition to a similar number of explosions carried out in mining, quarrying, and construction work. Because such a global network has never before been in operation, the number of events defined from infrasound and hydroacoustic observations is as yet unknown. Ultimately the bulletins will include an integrated analysis based on the fusion of data from all technologies. Bulletins, as well as raw data, will be made available to all treaty members.
The data center will also offer “screened” bulletins, which will not include events considered to be natural (earthquakes, meteors) or non-nuclear, man-made phenomena (mining, sonic booms). Screening requires increased knowledge of how natural events will be recorded by the monitoring system. The screening process should reduce the potential number of events reported daily from several hundreds to a handful.
Bulletins will provide estimates of agreed parameters only; they will not explicitly address or identify the nature of events.
Producing these reports will present the data center and its staff with a range of challenges. Perhaps the main challenge will be to make sure that events are not missed while keeping the reporting of non-events to a minimum. Rejecting spurious events, the appearance of which is inevitable in a largely automated system, is one of the key functions of the skilled human analyst.
High-tech times
The CTBT's verification system takes full advantage of modern technology. All the machinery–global monitoring stations and a data center for centralized data processing–draws heavily on recent technological advances in global communications and data processing.
Only 15 years ago, the Group of Scientific Experts used telex lines with data transmission rates of only 50 bits per second for the international exchange of seismic data. Today's affordable global satellite communications system speeds things considerably.
Radionuclide detection
The treaty calls for 80 radionu-clide stations capable of detecting even minute levels of radioactive particles. High-volume air samplers will force air through filter paper and collect aerosol particu-lates spread by the wind. New filters will be installed every few days, and analysis of filters will take place routinely. Local analysis will be supplemented when necessary by analyses at 16 certified laboratories specified by the treaty. This process determines the precise radionuclide composition, or “fingerprint,” of the particles.
Because particulates may be washed out of the air by rain, the negotiators in Geneva decided that noble gas detectors should also be deployed at radionuclide stations. These detectors absorb mainly xenon isotopes on an active, finely divided charcoal bed or filter. After collection, the gases are carefully extracted at higher temperatures and analyzed. This technique also detects underground nuclear explosions that leak radionuclides to the surface.
The group's early data exchanges were limited to parameter data (arrival times, amplitudes, and other attributes of the signals) derived from national analysis of much more voluminous waveform data.
A view from near the monitoring site in Saudi Arabia.
The verification system can now transfer all data available from 321 monitoring stations and 16 radionu-clide laboratories to the data center in Vienna for fully integrated processing and analysis. One station alone produces as much as 50,000 bits of data per second. In order to collect data from 337 sites and send it to today's 164 treaty signatories (out of the world's 193 states), there will be a potential 530 independent globally distributed satellite (and other) communication links. The VSAT (very small aperture terminal) satellite communications system now being installed around the world to serve treaty purposes will transfer 10 gigabytes of data to the center a day, and send varying amounts of data to the signatories, depending on their individual needs. The cost of communications is estimated at about $9 million a year–or about 10 percent of the annual budget of the fully implemented verification system.
The data center needs tools for archiving and quickly retrieving current and past monitoring data. The center will have a database of 125 terabytes (125,000,000,000,000 bytes) of information available on line. This technology is commercially available today, and the speed of current computer workstations is also adequate, even in circumstances of very high signal detection rates, originating, for example, from earthquake swarms or a series of explosions from volcanic eruptions.
On-site inspection
The assessment of compliance or noncompliance is a political process. Information provided by the international verification system is an essential input, but it will not be the only one. Individual treaty parties have varying capabilities to analyze raw data, and some will also make use of “national technical means,” such as satellite observations and information from their own national monitoring networks.
If a state party to the treaty is concerned about another state party's possible noncompliance, it has the right to engage both the treaty organization and other members, including the possible violator, in an effort to find out more about a suspicious event. If after consultations the state party is still concerned, it may request an on-site inspection of an area not to exceed 1,000 square kilometers. At least 30 of the organization's 51 Executive Council members must vote for such an inspection.
Wave detection
Both underground explosions and earthquakes are detected and identified by the seismic waves they produce. Solid earth is a good medium for long-range seismic wave propagation. Highly sensitive seismometers pick up the waves through the ground vibrations they produce; a deflection on the order of one nanometer (one millionth of a millimeter) is significant in this context.
The test ban treaty's seismic network will include 50 primary and 120 auxiliary seismic stations distributed over the land masses of the globe; these stations will detect underground nuclear explosions of extremely low yields. The 50 primary stations will provide continuous data; auxiliary stations will respond automatically upon request for additional data. This highly sensitive system is expected to detect as many as 50,000 earthquakes each year.
Lanzhou, China.
The primary stations include many technically advanced seismic arrays–with sensors arranged in a defined pattern over an area ranging in size from a few kilometers to as many as 50 kilometers. The array is, in fact, a steered antenna that provides information on the direction and speed of the incoming signal.
The seismic network will have an event detection and location capability of magnitude 3.5 for most parts of the world, which is generally considered to be adequate to detect a fully contained underground nuclear explosion with a yield of a fraction of a kiloton.
The treaty calls for 11 ocean-monitoring stations. Six underwater hydrophone systems, one in each of the large ocean basins, will detect any significant explosion that occurs in the water. Because the speed of sound is slower in the mid-depths of the ocean (known as the “SOFAR,” or sound fixing and ranging channel), events in the oceans can be readily detected. Experiments have shown that chemical explosions of only a few kilograms can be recorded thousands of kilometers away.
The other five oceanic monitoring stations are island-based seismic stations designed to pick up “T-phases”–acoustic signals that travel in water but convert to a seismic signal when they encounter steeply sloped coastlines.
Acoustic monitoring of the atmosphere detects “infrasonic” or low-frequency (less than 20 Hertz) waves that are inaudible to humans.
Immediately following an explosion in the air, the shock wave is a very steep, high-frequency, non-linear pressure change. As it dissipates, however, the atmosphere absorbs the higher frequencies, but the low-frequency signal remains and may travel thousands of kilometers without noticeable attenuation.
The atmosphere acts like a gigantic wave guide for the lowest frequencies. Other sources of infrasound are mining explosions, meteors, or bolides (exploding fireball meteors), and sonic booms from supersonic airplanes, which may produce false alarms. Infrasound waves may also be generated when strong winds cross mountain ranges.
The treaty calls for 60 infrasound stations, evenly distributed around the globe. Each station consists of an array of at least four microbarometer sensors placed within 1 to 3 kilometers. In order to minimize the effects of pressure variations caused by turbulent winds, each of the detectors is equipped with a noise reducer, a device that integrates the pressure variations over an area of more than 2,500 square meters without affecting the coherent low-frequency sound wave.
Although all three environments–solid earth, ocean, and atmosphere–share some of the same properties with respect to wave propagation, in detail they are vastly different. The fastest propagation is observed in the interior of the earth, where velocities approach 14 kilometers per second. Signals spread three dimensionally and, due to regional and local differences in geophysical properties, signal characteristics vary significantly over the globe. In the ocean, the hydroacous-tic signal travels at a speed of about 1.5 kilometers per second, with small differences depending on temperature variations. Two-dimensional spreading in the SOFAR channel leads to a significantly slower decline of the signals over distance. Thus, almost two-thirds of the earth's surface can be monitored with a very small number of stations.
Teheran, Iran.
In the atmosphere, the velocity of acoustic signals is about 0.3 kilometers per second, with considerable variation depending on meteorological conditions. These conditions also influence signal properties.
The local effects a visit might uncover vary from seismic aftershocks to “spall”–the local up-throw of material caused by the shock wave hitting the surface. Radioactive gases, notably xenon, may be detected if an explosion has not been fired deep enough to prevent leaking along cracks, faults, or holes used during the test. Another distinct local expression of a nuclear explosion is a cavity at the site of the explosion. The cavity may collapse, leaving a crater at the surface. In explosions at sea or high in the atmosphere, radioactive debris remains and is only slowly transported away, depending on the strength of ocean currents or wind. Atmospheric explosions at low altitude over land leave a long-lasting imprint.
By its very nature, an on-site inspection will be intrusive, including visual inspection and photography, measurements of radioactivity at and below the surface, seismological monitoring for aftershocks, the search for an underground cavity by active seismic surveys, magnetic and gravitational field mapping, and, as a last step, drilling to obtain radioactive samples.
If a member is found to be non-compliant, the Conference of States Parties may recommend collective measures in conformity with international law and the Conference or the Executive Council may bring the issue to the United Nations.
Where things stand
The Preparatory Commission–the precursor to the organization that will be in charge of implementing the treaty–has been established in Vienna. The cost of the verification system has been estimated at $225 million for equipment and installation, with annual operating and maintenance costs of about $85 million.
Building the global verification system is not an easy task. It involves a complex technical system of global dimensions in a political environment. The commission has been working for four years and had a budget of $84 million in 2001. The International Monitoring System is about half complete, and efforts are well under way on stations around the globe, in close cooperation with host states. About 125 stations, often in sites as disparate as the Antarctic and Iran, are fully installed. The Provisional Technical Secretariat has more than 250 employees, many of them well-known scientists, at its headquarters in Vienna.
At Windless Bight, Antarctica.
The commission is drafting agreements, technical specifications, data authentication and certification procedures for monitoring stations, as well as operational manuals, training programs, and procedures for on-site inspections, along with all the other requirements of establishing a complex new organization.
The challenge is not in any individual component; what makes this system unique is its global reach. It will be highly integrated, from the measuring stations to the centralized analysis. It is the first time that an international verification system has been built using stations in all climates of the world, from polar regions with high winds and low temperatures to hot and dusty deserts, and even to humid rainforests. When completed, monitoring stations will be established in 92 countries with different legal systems, cultures, and ways of approaching technical questions. An added difficulty is that most of the equipment is being installed at quiet locations, far from the infrastructure found in the noisy environments of major cities. Building the monitoring system is a challenge in its own right: High-tech infrasound stations are carried on the backs of donkeys down dirt roads in the Chilean desert; hydroacoustic cables are laid on the floor of the Antarctic Ocean, accessible only a few months of the year.
Laying cable in the Indian Ocean for a monitoring installation on Diego Garcia.
A golden opportunity
Science and scientists have played important roles in creating the conditions necessary to achieving a verifiable test ban treaty as well as shaping the actual verification arrangements, the most elaborate of any multilateral arms control and disarmament treaty to date. The CTBT is a good example of how science and technology can be used to enhance global security.
Some scientific issues, like the precise location of seismic events through the calibration efforts discussed earlier, are being addressed. Another issue is the proper categorization of the 100 or more seismic events observed every day. It is essential to avoid unfounded suspicion about naturally occurring events. The screening process requires an understanding of seismic sources in areas of interest and of the transmission properties along the path from source to station. This issue is high on the scientific agenda. Additionally, there has never been a global network of infrasound or ra-dionuclide monitoring stations, and research is under way on how to analyze and interpret data from these treaty networks. It is essential to improve our understanding of these observations and the events behind them, and to train scientists who can support the future operation of the verification system.
Although the major purpose of the monitoring system and the data center is to serve the needs of the treaty, the negotiators realized that the facilities and the data they provide could be extremely valuable for other scientific purposes. The Preparatory Commission is discussing ways to make the data available to the scientific community.
The system could provide basic data for a large variety of scientific studies within each of the monitoring system's four geophysical disciplines. While its raw and processed data could easily be used at existing institutions around the globe, it is tempting to consider creating a global meeting place for geophysical scientists. An international scientific center, using data generated by the treaty's verification activities, would also be a long-term asset to the treaty organization, providing scientific input to help maintain an up-to-date verification system.
Toward entry into force
The Preparatory Commission will continue to operate until the treaty enters into force. But when will that happen? The answer to that question is elusive. The treaty cannot enter into force until the 44 states that possess nuclear power or research reactors have ratified it. So far, 31 of the 44 have done so. Ratification by the United States and China is still pending, and India, North Korea, and Pakistan have not signed.
Article XIV of the treaty provided for a conference of ratifiers three years after it was opened for signature. A first conference, held in October 1999 in Vienna, discussed “what measures consistent with international law may be undertaken to accelerate the ratification process in order to facilitate the early entry into force of this Treaty.”
Tel-Alasfar, Jordan.
Ratifiers and signatories gathered again in New York in November 2001 to urge countries to sign and ratify the treaty at the earliest possible date and to consider the treaty's entry into force.
Signatories continue to respect the basic obligation of the treaty. An effective global verification system is being established with the support and close cooperation of member states, and provisions for practical implementation of an on-site inspection regime are being worked out.
It is difficult to understand why some signatories are not taking the steps necessary to bring the treaty into force. On the one hand, they are observing the basic obligation not to test, but on the other, they are choosing not to take advantage of the comprehensive verification regime that will be available at entry into force.