¿ìè¶ÌÊÓƵ

IN DECEMBER 1987, after signing the treaty banning intermediate-range
nuclear missiles, Mikhail Gorbachov announced to a surprised world that
Soviet scientists had perfected an amazing device. The instrument, he said,
could detect nuclear warheads on a ship from a helicopter hovering overhead.
Western scientists turned to the premier’s scientific advisers for an explanation.
It appears that Gorbachov was the victim of wishful thinking although his
statement did contain a kernel of reality. Radiation detectors can reveal
the presence of nuclear warheads but only if the weapons are not shielded
by heavy layers of metal. The devices might detect a nuclear weapon on the
deck, but not inside a ship.

Gorbachov’s enthusiasm for technology that can simplify the verification
of arms control treaties, such as the one for intermediate-range nuclear
forces (INF), has been contagious. ¿ìè¶ÌÊÓƵs in East and West are turning
their attention to the art of keeping track of nuclear weapons. One aid
in identifying nuclear weapons is the radiation they discharge. The raw
materials of the bomb are naturally radioactive, releasing neutrons and
gamma-rays as they decay. Emissions are characteristic of different materials;
warheads made with plutonium emit large amounts of neutron radiation, for
instance, while uranium-235, which can be used instead of plutonium in nuclear
weapons, releases very little.

In the US, many of these scientists are working at the Department of
Energy’s laboratories, where nuclear weapons are designed in the first place.
‘We know what a warhead looks like; we know what we’re looking for,’ said
David Dorn, who works on verification programmes at Lawrence Livermore National
Laboratory, California. The DOE is spending $110 million on nuclear detection
and verification research this year, 40 per cent more than it spent two
years ago.

¿ìè¶ÌÊÓƵs at the DOE’s laboratories are exploring a wide variety of
techniques. Tamperproof seals might show that the contents of a container,
once inspected, have not been changed. Tags that cannot be duplicated, when
attached to missiles, could serve the same purpose as licence plates on
cars: if a missile is sighted without a valid tag, it is obviously violating
the treaty . Radiation detectors, such as those that Gorbachov mentioned,
might be used to count warheads within a missile’s nose cone, or to distinguish
between non-nuclear and nuclear cruise missiles.

These detectors, however, illustrate the difficulty of finding verification
tools that are acceptable to all sides. The INF treaty requires American
inspectors to tell the difference between SS-20 missiles, which are banned,
and SS-25 missiles, which are not. The missiles are outwardly very similar,
but the SS-25 has an extra rocket stage and carries three warheads. The
SS-20 has only one warhead. Theoretically, inspectors could tell the missiles
apart by examining their nose cones with powerful X-rays. Computers can
transform the X-ray data into a three-dimensional image of the warhead called
a tomogram, just as computer-aided tomography, known as CAT-scanning, produces
an image of the human brain. Both sides rejected that idea because it would
disclose the design of the nuclear weapons. The X-rays might also reveal
other devices, such as decoys, contained in the nose cone. The Soviet Union
also vetoed equipment that would measure gamma-rays emitted from the warheads.
Gamma radiation can reveal sensitive details about the materials used in
a nuclear weapon, such as the isotopic composition of plutonium.

The two sides settled on a ‘passive’ device that simply detects emissions
of neutrons. The instrument measures neutron radiation from the warheads
at several points in a circle around the canister that contains the nose
cone. The radiation from the single warhead of an SS-20 is symmetrical;
the radiation from the SS-25 rises and falls as the detector moves between
its three warheads. Neutron detectors are not foolproof counters of warheads,
however. Carefully placed radiation shields and extraneous sources of radiation
within the nose cone can fool them. Warheads can also be designed to emit
very few neutrons or gamma-rays. This could be achieved by using only highly
enriched uranium, surrounded by layers of beryllium and tungsten. A more
reliable way to detect warheads that use only uranium-235 is to use an ‘active’
device that fires a pulse of neutron radiation into the missile. The neutrons
would cause fission within the uranium, releasing a burst of neutrons and
gamma-rays that can be detected by another device. Inspectors of the International
Atomic Energy Agency (IAEA) use this technique, called neutron interrogation,
to ensure that uranium and plutonium at civilian nuclear plants are not
diverted to military stores. They check to see that the amount of such materials
corresponds to the records kept at the plant.

The US has ruled out neutron interrogation for verification because
it would reveal sensitive information about the design of a warhead. There
are ways around this problem. Radiation detectors could be made less efficient;
designers could limit the sensitivity of the instruments, for instance.
Computers that record and analyse the radiation could also be programmed
to filter out and discard sensitive data.

Government officials are still uncertain about how much intrusion to
tolerate into the secrets of their nuclear weapons. They are aware that
the armed forces are reluctant to allow civilians too close to their arsenals,
but the balance continues to shift toward greater tolerance. ‘Look, the
Soviets know how to build a nuclear warhead,’ said John Harvey, who also
works on verification programmes at Lawrence Livermore. ‘The sort of information
that they might get about our weapons probably won’t help them improve their
designs. I’d like to err on the side of being able to tell if the Soviets
are cheating.’ Even though the use of X-rays to detect nuclear weapons has
been officially rejected by the government, scientists at the DOE’s laboratories
continue, with limited funding, to explore that technique. ‘We should plan
for the possibility that things will change, because you can’t come up with
these technologies instantaneously,’ said Harvey.

In experiments at Lawrence Livermore and at Argonne National Laboratory,
Illinois, scientists directed a narrow beam of gamma-rays at shipping containers
loaded with mock-ups of nuclear and non-nuclear sea-launched cruise missiles
(SLCMs). The beam passed over the missile’s guidance system, warhead and
rocket propellant in turn. Heavy elements, such as the uranium or plutonium
of a nuclear warhead, blocked much of the gamma radiation. Lighter elements,
such as the nitrogen and oxygen used in high explosive warheads, let the
beam through almost unobstructed. The graph of the beam’s attenuation provided
a ‘signature’ that clearly distinguished a nuclear missile from a non-nuclear
one. Such devices, set up at naval ports, could aid the identification of
nuclear cruise missiles. But the American government has decided that verifying
limits on SLCMs would be impractical because the missiles are so small,
easy to conceal and there are so many of them: the US is planning to deploy
about 4000 SLCMs. It says these missiles could be produced, maintained and
deployed in secret, unless inspectors were allowed to examine ships and
submarines around the world. The US Navy has vetoed that idea.

Some government scientists involved in verification research say that
the Department of Defense and the Arms Control and Disarmament Agency (ACDA)
of the State Department have tried to stop research into other areas of
verification as well. An evaluation of ACDA’s research programmes, carried
out by the agency’s own inspector general in March, criticised its ‘inhospitable
climate toward research’. Even when directly assigned to look for ways to
verify a ban on chemical weapons, for example, ACDA refused to devote any
resources to the issue, believing the ‘verification studies would add little
to existing knowledge’, according to the inspector general’s report.

Independent scientists have rushed in where the government’s specialists
fear to tread. In Britain, they are led by the Verification Technology Information
Centre, which coordinates work by 31 scientists at various British universities.
Much of their research is aimed at verifying non-nuclear disarmament, or
a ban on nuclear tests. Similar initiatives have begun in other European
countries, including West Germany and the Netherlands.

Groups of scientists in the Soviet Union and the US have joined forces
to propose verification schemes for more ambitious disarmament measures.
Earlier this month scientists from the Soviet Academy of Sciences and the
US Natural Resources Defense Council carried gamma-ray detectors aboard
a Soviet cruiser in the Black Sea, even though the USSR banned their use
to verify the INF treaty. The goal was to see whether the instruments could
detect several SS-N-22 missiles carried in launchers on the deck. Soviet
scientists from the Kurchitov Institute, a research centre on nuclear energy
in Moscow, also joined in the exercise.

The Federation of American ¿ìè¶ÌÊÓƵs and the Committee of Soviet ¿ìè¶ÌÊÓƵs
Against the Nuclear Threat began a series of studies on verification in
1987. The FAS-CSS project is chaired by Frank von Hippel, a physicist at
Princeton University, and Roald Sagdeev, chairman of the Soviet Academy’s
Committee on International Security and Arms Control. The scientists have
produced papers on how to verify a number of disarmament measures that the
American government has rejected as unverifiable.

¿ìè¶ÌÊÓƵs in the US have done most of the technical work on these studies,
but they have little influence on government policy. Their Soviet colleagues,
on the other hand, occupy positions of political influence, and are known
to advise Gorbachov on matters of arms control. ‘People like Sagdeev, Yevgenii
Velikhov (vice president of the Soviet Academy of Sciences), and Andrei
Kokoshin (deputy director of the Institute on the US and Canada) have had
a real impact on the Soviet leadership,’ said von Hippel.

Sagdeev has proposed banning reactors from satellites orbiting the earth,
but not from missions into deep space (‘Hungry for power in space’, New
¿ìè¶ÌÊÓƵ, 8 July). Any reactors in orbit could be detected easily by monitoring
the infrared radiation or gamma-rays that they emit. In fact, gamma-ray
astronomers in the US inadvertently detected reactors aboard several Soviet
military reconnaissance satellites, called RORSATs, when their observations
were distorted by gamma radiation from the reactors.

Soviet scientists are investigating ways to verify limits on the brightness
of ground-based lasers. Very powerful lasers could be used as anti-satellite
weapons. According to one preliminary study, sensors located up to 1 kilometre
from a laser source could monitor the brightness of the laser by measuring
the light scattered from dust and moisture in the area.

Much of the research by the FAS-CSS group is devoted to ways of monitoring
a reduction in the superpowers’ military stockpiles of plutonium and highly
enriched uranium. To succeed, techniques will be required to verify that
reactors producing plutonium and plants enriching uranium are shut down.
Inspectors will also need to verify that fissionable materials from warheads
have been removed from military stockpiles. Although the DOE’s scientists
pride themselves on their ability to work on verification without regard
to political constraints, they agree that verifying a cutoff of plutonium
production is not a popular topic of research. ‘We don’t think it will happen,
so we don’t put much effort into it,’ said Paul Stokes, a specialist on
verification technology at the Sandia National Laboratories, New Mexico.

Balancing the books

Theodore Taylor, a former designer of nuclear weapons at Los Alamos
National Laboratory, New Mexico, has outlined a method of verifying the
destruction of nuclear warheads. Taylor’s proposal does not foresee inspections
inside the plant where warheads are dismantled. Instead, it relies on measures
to verify that real nuclear warheads are transported to that plant, and
that all fissionable materials are accounted for when they leave the plant.
Highly enriched uranium and plutonium that are removed from the warheads
could then be placed under a system of safeguards, says Lawrence Scheinman,
a former official of the American State Department and of the IAEA. Referring
to the continuing strategic arms reduction talks (START), he said: ‘It doesn’t
make sense to have a START agreement that allows materials to be reworked
into other warheads. Here’s an opportunity for the superpowers to develop,
on their respective territories, international plutonium storage sites that
are verified by the IAEA.’ He adds that Soviet officials are considering
allowing the agency’s inspectors to verify the shutdown of several plants
that produce weapons-grade uranium.

Until the INF treaty, when the US and the USSR started to allow inspectors
to visit each other’s nuclear plants, the superpowers tried to detect each
other’s cheating by taking pictures from satellites. Even though on-site
inspections are becoming routine, satellites are still crucial for verification:
they can ensure that nuclear weapons are not being built or used in new,
unsuspected locations.

The US is launching improved spy satellites. Three of the space shuttle’s
first four launches, since it returned to space late last year, have added
links to the military’s surveillance network. Last December, a radically
new type of spy satellite flew into orbit aboard the space shuttle Atlantis.
It uses radar to produce high-quality images of the Earth’s surface. The
radar’s key advantage over previous sensors is its ability to see through
the clouds that generally cover much of Europe, since radar signals are
not affected by clouds. The satellite, code-named Lacrosse, was ready for
launch nearly two years ago. It was stranded on Earth when shuttle flights
were suspended in 1986 after Challenger exploded soon after takeoff.

Similar instruments, known as ‘synthetic-aperture’ radars, have been
carried aboard civilian satellites to study the Earth’s geological formations
and ocean currents. The spy satellite’s radar, however, can detect objects
much smaller than instruments built for scientific missions. Scientific
satellites cover a wide swath of territory, and the smallest element of
the image these satellites produce is typically 20 or 30 metres across.
According to John Pike, associate director for space policy of the Federation
of American ¿ìè¶ÌÊÓƵs, Lacrosse can detect objects as small as 1 metre
across. That level of detail is necessary to identify items such as mobile
Soviet missiles.

Traditional photoreconnaissance satellites that take pictures using
visible light and infrared radiation continue to be used as well. The latest
of these instruments, called the KH-12 satellites, will be launched aboard
a Titan IV rocket later this year. Such satellites use a powerful telescope
aimed at the Earth to gather detailed data. Images are recorded electronically
and transmitted to Earth. The KH-12 probably bears a strong resemblance
to the Hubble Space Telescope, which NASA will launch next year to observe
the stars. Both telescopes were built to fit inside the shuttle’s bay. If
the primary mirror of the KH-12 is the same size as Hubble’s, and no one
is saying precisely, it will be able to detect objects on the Earth 12 centimetres
across.

The third crucial part of America’s renovated surveillance network is
the Tracking and Data Relay Satellite System. TDRSS is generally described
publicly as the means by which shuttle flights and a variety of scientific
instruments communicate with the ground. But the satellites also relay a
stream of data from American spy satellites. TDRSS is especially critical
for Lacrosse, since synthetic-aperture radars generate a huge flood of data.
A radar system capable of detecting objects as small as half a metre across
could produce raw data, in the form of digital bits, at a rate of many billion
bits per second. This is far beyond the capacity of any existing communication
links in space. The two high-capacity channels aboard TDRSS can each relay
300 million bits per second.

By carrying computers onboard the spacecraft, and using them to reduce
the raw data instantly to images, the data stream could be compressed to
a size that TDRSS can handle. But that would require one of the world’s
largest supercomputers. A more likely way of getting around the data bottleneck
is for Lacrosse to operate intermittently, storing bursts of data on recorders.
The data could then be transmitted at a slower rate through TDRSS to the
ground.

Detecting smaller objects is no longer the key to more effective spying
from space. The greatest technical challenges now lie in programming high-speed
computers to unearth valuable information buried in the mountains of data.
Computers can search the data from a wide area, looking for the electronic
signal that matches the known echo from a Soviet missile launcher. But while
the idea of teaching computers to recognise an object is simple, the practice
is hard. One reason is that such mobile objects are often hidden behind
trees and under cover. According to Edward Aldridge, former secretary of
the US Air Force: ‘We’re still five years away from the point where some
data come in and ring a bell and say that I’ve got a target X in location
³Û’

Many other devices that could help verify arms control treaties exist
only as sketches, or as theoretical models. Turning them into working technologies
will take years, perhaps decades. Funding will be needed, but even more
important is political support. Leaders will have to approach the difficulties
of verification as problems to be solved, rather than convenient excuses
to avoid arms control.

Dan Charles is a science and technology writer based in Washington DC.