Such a strike is already one of the "low probability-high consequence events" that the U.S. Department of Homeland Security's new risk-based strategy is designed to direct more resources to combating.

But a careful review of the evidence suggests that there are technical obstacles to such an attack that are insuperable, for the time being at least, by the only terrorist organization seriously interested in staging one -- Osama bin Laden's al-Qaida.

And the two scientists themselves acknowledge there is no evidence that any terrorist group currently possesses the technical expertise necessary for a nuclear effort.

Bin Laden "perhaps has yet to find his Robert Oppenheimer," write Peter Zimmerman and Jeffrey Lewis in "Foreign Policy," the journal of the Carnegie Endowment for International Peace. Oppenheimer was the scientist who led the Manhattan Project, the secret U.S. effort at Los Alamos, N.M., to build an atomic weapon.

Zimmerman and Lewis are widely respected, and Zimmerman, now an academic, was formerly chief scientist both of the Senate Foreign Relations Committee, and the Arms Control and Disarmament Agency -- before its merger into the State Department.

Their piece imagines a year-long effort, undertaken at an isolated ranch property, by a team of 19 terrorists -- the same number who carried out the Sept. 11 suicide hijackings.

The 19, all technically-trained specialists, would be divided into teams dealing with physics, engineering, metallurgy, machining, ballistics, electronics and procurement.

The weapon they would build is a "gun-assembled" uranium bomb, like the one code-named "Little Boy," built by Oppenheimer's Manhattan Project, and dropped on Japan in 1946.

The authors call such a bomb "conceptually simple," pointing out that the technology "is more than 60 years old," and adding that "It is perhaps easier to make a gun-assembled nuclear bomb than it is to develop biological or chemical weapons."

They add that, at a cost of $10 million, the bomb, which could kill up to 100,000 people, would be an extremely cost-effective attack.

"In strictly commercial terms ... for a cost of ... about $100 a murder, it would be a bargain."

But some experts that United Press International spoke to -- whilst stressing their respect for the authors -- expressed deep skepticism about their argument.

Arms control expert Milton Leitenberg of the University of Maryland called the scenario "super-optimistic," and said the authors had glossed over the difficulty of finding the kinds of highly qualified experts the project would need.

"Yes it's conceptually simple," he said, "And that's where the simplicity ends."

The tiny size of the team -- the Manhattan Project had a staff of three thousand -- meant that in every one of a dozen or more expertise categories "you would have to find someone with the absolute optimal skills."

"How does that kind of organization find those kinds of people, in the real world?" he asked.

"Historically, al-Qaida has never had anyone at that level who was prepared to help them in that way," he said, adding that al-Qaida's unconventional weapons development efforts in Afghanistan had amounted to very little.

The Japanese cult Aum Shinrikyo -- who managed to make sarin nerve gas in the 1990s -- had a nuclear effort, too, said Leitenberg, adding "It was a shambles."

"These people don't pop up like jack-in-the-boxes," he said of the experts in more than a dozen fields required for such a project.

George Smith, a veteran weapons analyst, pointed to a paper prepared for a more specialist audience in 1987 by the Nuclear Control Institute. He said the authors -- including J. Carson Mark, who led the Theoretical Division at Los Alamos National Laboratory for nearly three decades -- had addressed detailed technical issues not covered in the "Foreign Policy" article.

"I think in writing for a general audience, (Zimmerman and Lewis) elided some of those technical issues," he told UPI.

In particular, the institute paper addresses the difficulty of preparing the fissile material -- uranium in the case of the "gun-assembled" device -- and the large amount that would likely be required for a successful fission reaction to be initiated.

"They would have to acquire more material than is to go into the device, since ... considerably more material is required to work with than will appear in the finished pieces," says the paper.

Moreover the institute paper points out that with a relatively crude gun-type device, there is a good chance of a problem known as pre-detonation, where the nuclear chain reaction starts too soon, resulting in what is called a "fizzle yield" -- an explosion equivalent to just a few thousands pounds of dynamite, as opposed to the hundred thousand pounds of a successful 10 kiloton nuke.

Overall, Leitenberg said their account "omits real consideration of at least a dozen points in the process where something could, and very likely would, go wrong that would bring the whole project to an end."

Zimmerman and Lewis do acknowledge, "Our scenario does not suggest that terrorists would find building a nuclear weapon either easy or inexpensive."

"The most important obstacle remains the difficulty in acquiring enough nuclear explosive material to build a bomb," they add.

But they argue that "No one really knows how much highly enriched uranium there is in the world, or how close the wrong groups are to getting the right amount."

In reality, however, getting even the "right amount" is unlikely to be enough, given what we know about the capabilities of the 'wrong groups."

Related Links
Learn about nuclear weapons doctrine and defense at SpaceWar.com