杏吧原创

IT DOESN鈥橳 look like a battleground. Skyscraping orange cranes frame a postcard-ready panorama: the dark waters of Puget Sound, the snowcapped Olympic Mountains and the dramatic skyline of Seattle, Washington. Sitting in the back seat of a government-issue truck, two young officers of the US Customs and Border Protection service ignore this vista. Instead they track the spectral outlines playing on an oversized video monitor and the cryptic lists scrolling down a laptop computer mounted between them.

The officers are the last line of defence in the American government鈥檚 war on terror, as waged at the country鈥檚 seaports. Each day, at Seattle, New York and any number of other ports, officers watch as scores of 40-foot shipping containers pass by on trailer trucks bound for all corners of the US. They aim to prevent a disaster: a nuclear warhead or weapons-grade material being smuggled in.

Seattle鈥檚 container port is the closest to a major US city centre, and in recent years has been one of the fastest growing ports in North America. From the top of the officers鈥� truck, a bracket extends like a giant caliper above the containers. It is one of 178 Vehicle and Container Inspection Systems, or VACIS, working across the US 鈥� the most advanced piece of technology now widely deployed to fight nuclear terrorism.

As each container passes by, a shutter opens and gamma rays from radioactive caesium-137 or cobalt-60 mounted on the bracket鈥檚 outer wing are beamed through the container to a screen mounted on the side of the officers鈥� truck. This captures something resembling an airport X-ray image, only much coarser, which is transmitted to the video screen inside. One agent fiddles with the controls 鈥� smoothing, sharpening and translating shades of grey into a palette of psychedelic hues to better distinguish shapes. The other scrolls through the manifest for each container鈥檚 contents. They check whether the images on one screen match the list on the other, or whether closer inspection is needed.

Reading the images is an exacting, if not necessarily exact, business. One container presents a uniform dark grey image. 鈥淔rozen seafood,鈥� one agent says. A darker patch would indicate something denser 鈥� a cache of uranium, say, or lead shielding. These 鈥渁nomalies鈥� are flagged for closer examination and an agent outside scans the container with a hand-held radioisotope detector, similar to a Geiger counter. In 2 per cent of cases, inspectors open the box, sometimes at a more secure location. Another container shows up on the screen with a stippled pattern. Nike shoes, the manifest reads. 鈥淚 can鈥檛 see individual shoes,鈥� says the operator, 鈥渏ust the pattern. I鈥檝e learned to recognise them.鈥� The next has vertical stripes of varying width. 鈥淐andle holders, electrical lamps,鈥� the second agent reads from her manifest. Her partner flips through his colour palette, squeezing the last bit of information from the image. 鈥淪orting it all out, that鈥檚 where experience comes in,鈥� he says.

And that鈥檚 the weakness of VACIS: how well it works depends on the judgement and image-hacking skills of its operators. In Seattle, they analyse 40 to 150 shipping containers a day, with each inspector using their own methods; on another visit by New 杏吧原创, the agent scarcely tinkered with the images at all. Before starting the job, they receive two weeks鈥� training. 鈥淚t takes doing this maybe three or four months on a daily basis to really feel comfortable,鈥� says one.

Officials in the US Department of Homeland Security will feel more comfortable if and when technologies now in the pipeline provide more automated and precise ways to spot smuggled nuclear weapons. In April 2005 they established a Domestic Nuclear Detection Office (DNDO), with an annual budget of $56 million, to tie together efforts scattered across many agencies. This June, in a breakthrough meeting at Lawrence Livermore National Laboratory in California, DNDO officials announced funding for some 40 advanced nuclear-detection systems that won鈥檛 be up and running for three or more years.

VACIS was developed in the 1990s for safeguarding nuclear material in the post-Soviet states. It represents one of two approaches to detecting contraband: optically scanning for suspicious shapes and opaque materials. The other is to detect radioactive materials. Here the standard technology is radiation portal monitors, which use polyvinyltoluene plastic sensors that produce visible light in response to high-energy gamma rays and neutrons. These devices 鈥� 842 of which are stationed at US ports and borders, including several at the Seattle pier 鈥� are set alongside traffic lanes to scan passing vehicles and containers. Their incongruously cheery yellow pillars are becoming familiar fixtures, like metal detectors and X-ray baggage scanners.

鈥淩adiation portal monitors are least effective at detecting highly enriched uranium鈥�

Portal monitors are least effective at detecting highly enriched uranium, one of the hardest bomb materials to track. Its fissile component, uranium-235, does not emit many neutrons or high-energy gamma rays, and its low-energy gamma rays can be easily shielded with lead plates. So it鈥檚 up to VACIS to spot any shielding. Confusingly, portal monitors are also distracted by benign positive readings. Many naturally radioactive goods can set them off 鈥� ceramics, cat litter, road salt, smoke alarms, medical isotopes, even the potassium-40 in bananas.

One improvement is already in production and should be installed early next year: spectroscopic portal monitors which use sodium iodide crystals. Unlike plastic sensors, sodium iodide registers the distinctive gamma-ray signatures of isotopes 鈥� so no more masking plutonium-239 with bananas. But the crystals must be kept away from water or they will disintegrate, and they can crack when shaken or exposed to rapid temperature changes. Richard Kouzes of the Pacific Northwest National Laboratory in Richland, Washington, notes that another material, semiconducting germanium, can provide about 40 times the spectral resolution of sodium iodide, allowing finer distinctions between isotopes. It is much more expensive, however, and must be supercooled.

So while some researchers want to improve 鈥減assive interrogation鈥�, detecting radiation that is naturally emitted, others are pursuing 鈥渁ctive鈥� approaches (New 杏吧原创, 22 October 2005, p 30). The best-established forms of active interrogation entail bombarding a container with neutrons, as spearheaded by Dennis Slaughter at Lawrence Livermore, or with X-rays, as led by James Jones at the Idaho National Laboratory. This provokes partial fission in materials such as plutonium and uranium-235, which then emit easily detected high-energy gamma rays or neutrons. 鈥淭he idea goes back to my 35-year-old thesis, which only two people have read: me and my PhD adviser,鈥� jokes Slaughter. 鈥淚t never occurred to me I鈥檇 get anything out of it.鈥�

Neutron gobblers

The two active approaches share a special characteristic: the neutrons and high-energy gamma rays they elicit can pass through lead and other shielding. What鈥檚 more, neutron interrogation can identify conventional explosives and chemical agents and is already being used in at least one US border station to search for drugs. However, both methods have a conspicuous drawback: hydrogen atoms absorb neutrons, and goods made from organic materials, such as clothing, food, plastics and wood, are loaded with hydrogen, as are water and petroleum. The hydrogen could block both neutron beams and the neutrons released by X-ray interrogation.

Slaughter and his colleagues may clear this hurdle with the help of a small Russian-designed particle accelerator called a radio frequency quadrupole. 鈥淲ith the accelerator, we can produce a much brighter neutron beam and concentrate the neutrons in a forward direction,鈥� he says. This beam can detect a 600-gram sample of enriched uranium, a tiny share of the quantity needed to build a bomb, through organic material denser than the average container load. Still there are trade-offs: the deeper and denser the surrounding material, the larger the sample required for interrogation to work.

The developers of a third active technology, nuclear resonance fluorescence imaging, need not worry about such trade-offs. William Bertozzi, a professor of physics at the Massachusetts Institute of Technology, has designed a system that shoots an X-ray beam spanning a range of energy levels into a sample. The various elements and isotopes, radioactive or not, respond to the rays by emitting photons with distinctive fluorescent signatures. The X-rays and photons can be stopped by lead, but travel easily through organic materials. This technology might complement neutron beams: each can see where the other cannot.

So far so good, except that they work by beaming radiation which could endanger any workers or stowaways. Slaughter and Bertozzi think they can get strong enough readings while staying within safety guidelines. However, as with VACIS, drivers must leave their vehicles before screening, slowing the process. Slaughter recommends neutron interrogation as a second tier, to investigate containers that VACIS flags as suspicious. This would obviate some fears about radiation levels, as any stowaways would already have been spotted by the imager and removed.

The creator of another nuclear detection technology, Los Alamos National Laboratory physicist Chris Morris, doubts his rivals鈥� safety claims: 鈥淭he doses you need to penetrate cargo containers are large, whether you use neutrons or gamma rays,鈥� he says. 鈥淚f they tell you they鈥檙e not, they are overselling their technology.鈥�

Morris鈥檚 approach, known as muon interrogation, carries no such risk: it draws its power literally from the stars, using the cosmic rays that are constantly bathing the Earth. When the resulting charged particles, or muons, strike weighty elements such as lead, uranium and plutonium, they are deflected. It is this deflection that can indicate a suspicious item is present. Morris sees the technique first as a means of screening the 120 million motor vehicles that cross US borders each year, many of which receive only portal monitoring or no scanning at all, and second for the ports. Like VACIS, muon interrogation scans for density rather than radioactivity. Unlike VACIS, it renders a three-dimensional image.

The detectors Los Alamos uses are thin-walled aluminium tubes filled with a mixture of argon and ethane gas. Gold-plated tungsten wire runs down their cores, carrying 2000 volts of electricity. When muons pass through, they ionise the gases, leaving electrochemical tracks that enable a computer to locate them to within a quarter of a millimetre. The tubes are arranged in pairs of narrowly spaced arrays, each on opposite sides of the object to be screened (see Diagram). The first pair of arrays captures a muon鈥檚 original path while the second captures its trajectory after any deflection. Plot the two and you can find the intersection point at which a dense mass made the muon veer. Assemble these points in sequential planes and you get a 3D picture.

鈥淵ou can visualise a nuclear weapon in a van in 30 seconds, while you鈥檙e looking at passports鈥�

Precise tracking of each muon is possible because they are so rare, just one per minute per square centimetre at sea level. That raises a contentious point. 鈥淢y feeling is it鈥檚 impractical,鈥� says Kouzes. Others agree, saying there just aren鈥檛 enough muons out there to provide a good image in a reasonable amount of time. 鈥淭hat statement is idiotic,鈥� replies Morris. 鈥淚n 30 seconds, you can clearly visualise a nuclear weapon in a van, in the time you鈥檙e looking at their passport.鈥� As for image quality, Morris claims 鈥渞esolution on the order of VACIS鈥�. That may not cheer the customs operators straining to decipher scans on Seattle鈥檚 dock, but muons would provide 3D imaging and deeper cargo penetration 鈥� through even a metre of steel. Also, agents need not wait for the driver, passengers or livestock to get out before starting the scan.

Muon interrogation may be dazzling in theory, but it鈥檚 further from completion than the other systems. Los Alamos has built a small prototype and simulated it on a computer. Now Morris鈥檚 team is doing more extensive computer modelling and plans to complete a full-scale prototype suitable for scanning a large van in about 18 months.

None of these monitors, however, can cope with one potential nuclear transporter: the large bulk-cargo carrier. Oil tankers, many of which come from the Middle East and other volatile regions, remain the ultimate border-security puzzle. Their enormous size and dense contents, rich in neutron-gobbling hydrogen, may be impenetrable. Slaughter says any nuclear device would probably be attached to the hull rather than suspended inside a tanker, so running a neutron beam over the exterior should locate it. Morris has another idea: float a mini-submarine in the oil to get a radiation reading. Researchers are exploring still-secret schemes for mounting particle accelerators on barges to conduct scans of entire ships. However, these might release dangerous amounts of radiation and still fail to penetrate a hold full of petroleum.

Whatever detection arsenal the US government decides on is likely to incorporate several technologies. Yet all these approaches share a fundamental weakness: they are fixed systems that must take readings in seconds, perhaps a minute, if they are not to sabotage international commerce. Critics also contend that the basic approach is flawed; terrorists might just smuggle contraband overland or onto a deserted beach or airstrip, as drug traffickers do, and carry it to their target in a van or car, as in the first World Trade Center bombing in New York.

James Carafano, a national security fellow at the Heritage Foundation in Washington DC, dismisses the container effort as an 鈥渋nevitably imperfect response to an unlikely threat鈥�, while other more likely threats, such as conventional explosives or small nuclear weapons brought in overland, get ignored. 鈥淭o be relatively useful,鈥� argues Carafano, 鈥渁ny mass-screening technology must have low rates of false positives and negatives, low human capital and infrastructure costs and no serious privacy or health concerns. By and large, those technologies just don鈥檛 exist.鈥� America鈥檚 top nuclear laboratories, however, are determined to prove him wrong.