The purposes of verification are detecting violations, or possible violations, of a treaty, thereby providing early warning of any threat to the state's security arising under a treaty regime; deterring violations of an agreement by increasing the risk of detection and complicating any attempts at evasion; acting as a domestic and international confidence-building measure in the viability of an arms-control treaty; and providing data for the presentation of arms-control issues to the public (Lyddon and Lingwood, June 1988).
The utility of any arms-control or disarmament treaty will depend to a large extent on the effectiveness of its verification provisions. Verification is, therefore, a crucial issue. It is also a highly technical issue, often relying on sophisticated sensors on board satellites or complex arrays of seismic monitoring stations.
Just how complex verification can become is shown by the proceedings at the current negotiations at the forty-nation Conference on Disarmament in Geneva on a comprehensive ban on chemical weapons, designed to prohibit the development, production, stockpiling, acquisition, retention or transfer of chemical weapons, including lethal and incapacitating chemicals and their precursors. The verification of this treaty will probably have to deal with: obligations to be checked by systematic international on-site verification; the destruction of stock-piles of chemical weapons by continuous monitoring with on-site instruments and the continuous presence on site of international inspectors; and the destruction of chemical-weapon production facilities by monitoring with on-site instruments and periodic international on-site inspections.
In addition, there may be provisions for some mandatory routine international on-site inspections of the destruction of chemical-weapon stocks; the verification of non-production at declared facilities; and
161
challenge inspections to deal with cases of suspected non-compliance which may not have been revealed by regular inspection of declared facilities. Any buildings, other than domestic habitations, may - in theory at least - be subject to inspection.
The verification procedures will require novel technical equipment, particularly for the continuous chemical analysis, by spectrometry and other methods, of effluents from various parts of chemical plants. The appropriate technologies exist, but suitable monitoring equipment is under development.
The verification requirements for a chemical-weapon treaty are comparable with, or more complex than, the safeguards procedures for verifying the provisions of the NPT, elaborated by the IAEA. Although confidence in these procedures has been shaken by Iraq's undetected violation of NPT safeguards, they have, over the past fifteen years or so, acquired considerable credibility with the parties to the NPT. The IAEA is a large agency, employing a large international staff, including some 200 professionally trained inspectors, and having an annual budget of $100 million. A global chemical-weapon treaty will probably require the establishment of another verification agency, comparable in size to the IAEA, to verify it.
A number of arms-control treaties - including the Antarctic Treaty, the Partial Test-Ban Treaty, the SALT I and SALT II Treaties, the Threshold Test-Ban Treaty, the Intermediate Nuclear Forces Treaty and the START Treaty - contain an explicit prohibition of interference with 'national technical means', thereby legitimizing these techniques for collecting information about other countries' activities. National technical means are, and are likely to remain, the most important verification techniques.
The technologies used for arms-control verification are essentially the same as those used to gather military intelligence. These currently rely mainly on photographic cameras carried on board satellites; the monitoring of radio and radar communications generated by military activities, an activity called electronic intelligence (ELINT) and normally performed by electronic equipment carried on satellites; the collection of data by other satellite-borne sensors, particularly infra-red sensors; and the seismic monitoring of underground nuclear tests. Clearly, the family of sensors carried on satellites - including photographic cameras, television cameras, multi-spectral scanners, radi-ometers, microwave radars, gamma-ray detectors, X-ray detectors, and detectors of electronic and communications signals - are crucial in verification activities.
162
These national technical means are supplemented by information collected by: on-site inspections; espionage (the clandestine collection of information by human agents operating on the territory of other countries); and audits and accountancy to check the production of materials controlled by arms-control treaties. The START Treaty, for example, allows for nine types of on-site inspection: baseline data inspections, data update inspections; new facility inspections; suspect site inspections; re-entry vehicle inspections; post-exercise dispersal inspections; conversion or elimination inspections; close-out inspections; and formerly declared facility inspections (US Arms Control and Disarmament Agency 1991), But photographic reconnaissance satellites - like the American KH-11 (Keyhold) and the Russian Cosmos satellites - are today's intelligence work-horses and are likely to remain so for some time, although satellites carrying infra-red sensors also play an important role.
There are two types of photo-reconnaissance satellite activities. One is to scan a large area of territory with a wide-angle low-resolution camera. If there is something suspicious on the photographs, it is investigated with a high-resolution ('close-look') camera. Some photo-reconnaissance satellites are devoted to just one activity, but current American KH-11 satellites carry both close-look and wide-angle cameras. These satellites typically have perigees of about 250 kilometres.
The resolution of the best high-resolution cameras used on reconnaissance satellites is a closely guarded secret. But, when the atmosphere is reasonably clear and the weather is fine, resolutions of 10 centimetres or less are probably achieved at orbital altitudes of about 250 kilometres (Krass 1985).
With a resolution of 10 centimetres, the following are among the objects that can be precisely identified and described in considerable detail on satellite photographs: bridges, radar sites, supply dumps, troop units, airfield facilities, rockets and artillery, aircraft, command and control headquarters, surface-to-surface and surface-to-air missile sites, surface ships, vehicles, land minefields, ports and harbours, coasts and landing beaches, railway yards and shops, roads, and surfaced submarines (Reconnaissance Handy Book 1980). Photographs taken with high-resolution cameras show impressive detail and are extremely useful for verification purposes. In the words of William Colby, the late director of the Central Intelligence Agency: 'You can see the tanks, you see the artillery, but you may not quite see the insignia on the fellow's uniform' (Colby 1979).
Nevertheless, photo-reconnaissance is limited by cloud cover and
163
poor light. And, of course, objects that are underground, underwater, inside buildings, or otherwise hidden, cannot be photographed with light. These difficulties can to some extent be overcome by using infrared and radar images.
The earth's atmosphere is transparent to infra-red radiation. Infra-red of some frequencies can make developable some types of photographic film and is called photographic infra-red. Although photographic infrared can penetrate haze to a useful extent, cloud cover still seriously limits its use in satellite photography. And so does darkness. But its use does considerably improve contrasts.
All objects in nature emit, absorb or reflect radiation in a unique way. Objects can, therefore, be identified by detection and measurement of their characteristic radiations. Camouflaged objects, for example, create a spectral response different from their surroundings, and differences in photographic infra-red reflectance can improve photographic contrasts enough to show up camouflaged objects.
Infra-red can also be used to detect radiation emitted by warm and hot objects with infra-red detectors. Although these can be made very sensitive, infra-red images from satellite-borne detectors have much worse resolutions than photographic ones, mainly because of larger atmospheric diffraction effects. In fact, satellite infra-red resolutions are typically a hundred times poorer than those for visible light. Nevertheless, infra-red imagery has important potential uses in verification. It could, for example, be used to detect the movement of vehicles at night, and camouflaged or otherwise hidden activities.
High-resolution images of objects on the ground can also be obtained by synthetic aperture radars (SARs) carried on satellites. A ground-based radar usually has a large antenna to obtain a good resolution, But to be practicable in outer space a radar must have a relatively small antenna. An SAR uses a cunning method to get a good resolution with a relatively small antenna.
It uses the motion of the antenna relative to the ground to increase the effective length of the antenna. As the antenna, equivalent to the aperture of a camera, moves with the satellite in its orbit, the radar signals reflected back from objects on the ground are received by the antenna and recorded; they are later analysed by a data-processing system. The maximum length of the synthetic aperture is the length of the satellite path along which the moving antenna receives the
164
reflected radar signals from a given object; the effective length of the antenna is thereby increased.
The resolution of an SAR is typically some ten times worse than that from optical photographs, but SAR pictures can be obtained through the thickest clouds at any time, day or night. Another advantage of SAR, compared with optical and infra-red sensors, is that its resolution is independent of the distance between the antenna and the distance of the object of interest. SAR satellites can, therefore, be operated in high enough orbits to give them long lives.
Some reconnaissance satellites use charged-couple devices (CCDs), semi-conductors sensitive to light, in linear arrays of detectors (Jasani and Barnaby 1984), The arrays are used in place of a film. The CCDs store an electric charge in the pixels (picture elements) that is proportional to the intensity of the light. After exposure, the charge at each pixel is read directly by computers and transmitted directly to a ground station. The use of films, and their time-consuming development, is thereby eliminated.
The CCD is an important new technological development that makes easier the collection of intelligence in real time, an extremely useful capability for some verification purposes. The resolution of the images obtained with an array of CCDs is reportedly as good as photographic films.
The Conventional Armed Forces in Europe (CFE) Treaty, reducing conventional weapons in Europe, includes very large numbers of weapons - armoured vehicles, aircraft, helicopters and artillery guns - which are relatively small and very mobile. The treaty stipulates that the countries of the old Warsaw Pact alliance, on one side, and NATO countries, on the other side, are each limited to 20,000 tanks, 30,000 armoured combat vehicles, 20,000 artillery pieces, 6,800 combat aircraft and 2,000 attack helicopters.
The verification of the CFE Treaty is mainly by intrusive inspection, Each country (twenty-two involved) is obliged to give the others detailed information annually about the size and whereabouts of its forces in Europe. Each country can check this information by sending teams of inspectors to other countries. Inspections are to check that no more weapons are being held at military site than have been declared as being held there. Surprise visits, called challenge inspections, can be made to check that weapons are not held anywhere else,
165
including non-military sites. Russia alone has to be ready to receive about 200 inspections a year.
In addition to the normal national technical means, there are a number of technological systems which can be used to verify treaties controlling conventional arms, whether in Europe or in any other region. Flying military aircraft can be monitored by existing systems, such as Airborne Warning And Control System (AWACS) aircraft.
Systems such as the Joint Surveillance Target Attack Radar System (Jstars) can detect, identify and track moving columns of tanks, armoured personnel-carriers and other vehicles. Jstars is an airborne system using side-looking radars able to observe objects within a circular area of a 150-kilometre radius.
An individual land vehicle, such as a tank or an armoured combat vehicle, can be provided with a token containing biological material from an individual animal or human which could be genetically unique and forgery-proof. In addition, the token could be equipped with a hologram, extremely difficult to replicate and which could be checked on the spot. Tokens could also be provided with integrated circuits broadcasting a signal on interrogation or sending telemetry signals with a tamper-resistant code.
Storage areas for armaments can be monitored by infra-red sensors, able to monitor the dimensions of vehicles, and video sensors; buried sensors, particularly seismic sensors; and X-ray sensors to monitor exits (Dean 1989). The verification of the Intermediate-range Nuclear Forces (INF) Treaty already includes this sort of storage-area monitoring.
Buried strain-sensitive cables can be used to register movements of vehicles crossing it: passive infra-red scanners to distinguish between people and vehicles, count numbers of persons or vehicles passing, and register direction and speed; and miniature seismic intrusion detectors can be used to detect vehicles at distances of up to 500 metres.
Verification may also involve detecting intruders into monitored areas. Active-beam alarms systems, in which intruders break the beam, use infra-red and microwave beams, but low-level laser beams may become the favoured system because the beam is finer and less affected by adverse weather conditions. For detecting intruders, however, microwave systems have the advantage of relatively wide coverage area, which makes it difficult to bridge the system. Microwave beams, which are very effective detectors, generally have longer ranges than infra-red or laser beams - up to 200 metres compared with 100 metres.
Fence detection systems typically use acoustic cable in which a treated coaxial cable attached to the perimeter fence acts as a microphone. A signal is generated when an intruder tries to get over the
166
fence or cuts the cable. A more sensitive system is a suspended wire, mounted on the top of the perimeter fence on insulators, which uses capacitance to detect the proximity of an intruder's body. A similar system is used to monitor the entire length of the Hong Kong border fence.
Perimeter, and other, security systems can be backed up with closed-circuit television. Slow-scan television can be used to send video signals from remote and isolated sites via telephone lines to a central control station. A two-way telephone link can command the cameras to pan, tilt or zoom, and a permanent video record of events at the remote site can be obtained if an alarm is triggered.
One purpose of future treaties controlling conventional arms will be to reduce the risk of surprise attack. The rapid analysis and communication of data will, therefore, be a crucial element of the verification of such treaties.