VIENNA – A leaked diagram suggesting that Iran is pursuing a nuclear weapon is scientifically flawed, diplomats working with the U.N. nuclear agency conceded Friday. However they insisted that it still supports suspicions that Tehran is trying to build a bomb, especially when combined with other documents that remain secret.
The Associated Press reported Tuesday, citing the document leaked by officials from a country critical of Iran's atomic program, that it indicated that Iranian scientists had run computer simulations for a nuclear weapon that would produce more than triple the explosive force of the World War II bomb that destroyed Hiroshima.
The diagram showed a yield of 50 kilotons. But subsequent criticism of the AP's report showed that result was widely inaccurate. Instead, the yield of the hypothetical weapon was much higher and hugely greater than any bomb ever produced — meaning it was next to impossible that Iran was contemplating such a weapon.
Nuclear scientists Yousaf Butt and Ferenc Dalnoki-Veress, writing on the website of the Bulletin of Atomic Scientists, sharply criticized the AP's report and described it as containing a "massive error," adding that it was "either slipshod analysis or an amateurish hoax." They said the "level of scientific sophistication" to produce the diagram "corresponds to that typically found in graduate or advanced undergraduate-level nuclear physics courses."
But a senior diplomat familiar with the probe of Iran by the IAEA told the AP on Friday that the agency suspects that Iranian scientists calculating a nuclear yield intentionally simplified the diagram to make it comprehensible to Iranian government officials to whom they were presenting it. He said that when the right data are plugged in, the yield is indeed 50 kilotons. The diplomat, who is considered neutral on Iran's nuclear program, spoke only on condition of anonymity because he was not authorized to divulge intelligence.
As seen by the AP, the data on the left-hand vertical side of the diagram is listed in kilotons per second. But David Albright, whose Institute for Science and International Security is used by the U.S. government as a go-to source on Iran's nuclear program, told the AP that when that legend is substituted with another — joules per 10 nanoseconds — the yield comes out to around 50 kilotons.
When told of Albright's calculations, the senior diplomat confirmed that the agency thought they were correct. He also said the agency had a spreadsheet thought to have been drawn up by Iranian nuclear scientist Majid Shahriari that contained the data needed to produce a nuclear yield of 50 kilotons. He said Shahriari is also believed to have produced — and then altered — the diagram, which he said was one of several held by the IAEA showing such yield calculations.
Iran denies any interest in nuclear weapons and accuses the United States and Israel of fabricating evidence that suggests it is trying to build a bomb.
The senior diplomat said agency investigators realized the diagram was flawed shortly after they received it last year but believe it remains important as a clue to Iranian intentions.
He and a second diplomat said other classified material held by the IAEA supports concerns that the graph may be part of a past Iranian effort at developing nuclear weapons. The second diplomat comes from a country suspicious of Iran's nuclear intentions but not from the nation that shared the diagram with the AP. He too spoke only on condition of anonymity because he is not authorized to disclose secret information.
The diagram was disclosed to AP in an attempt to bolster arguments that Iran's nuclear program must be halted before it produces a weapon. The officials who leaked it provided what they described as a computer model of blast calculations only on condition that they and their country not be named.
In its report, AP said the curve of the diagram peaks at just above 50 kilotons at around 2 microseconds, reflecting the full force of the weapon being modeled. The bomb that destroyed the Japanese city of Hiroshima during World War II had a force of about 15 kilotons.
But experts said that if the figures and units on the diagram were correct, the blast produced would be in the order of nearly 2 million kilotons — far greater than yield given on the graph. It would also be much larger than the biggest nuclear weapon ever built — the Soviet Union's "Czar Bomba" detonated during a 1961 test with a yield of 57,000 kilotons.
Scott Kemp, a professor of nuclear engineering at MIT, described the 2-million kiloton figure as "impossibly high" for any nuclear weapon.
"The numbers don't add up to realistic weapons values; the axes do not agree with the caption," he told AP. "At best ... this is an oversimplified theoretical simulation, not a serious model of an actual weapon."
Dalnoki-Veress, one of those who originally questioned the accuracy of the diagram, told AP that the diagram does not make sense unless its creators "are thinking of an impossibly massive bomb." Initial bomb tests of countries developing such weapons are usually in the one or two kiloton range.
Reached by telephone Friday, one of the two officials who give the diagram to the AP acknowledged that the data on the left-hand vertical side were manipulated but said that was the work of the Iranian scientist who created it. He asserted that did not change the thrust of his country's claim — and the IAEA's fear — that Iranian scientists were working on bomb yield calculations.
The 155-nation IAEA deals with a wide range of nuclear issues but has focused attention on Iran's nuclear program. The agency has pushed its monitoring and investigation of that country's known and suspected nuclear activities to its top priority. It depends on its own research, open source information, satellite imagery and intelligence from IAEA member nations for that task.
While it briefs the agency's 35 board members in quarterly reports, the IAEA remains secretive with even those countries, sharing information on a selective basis. IAEA officials say releasing too much information can jeopardize further investigations.
They also explain delays in conveying what they know to the board by saying that the agency must be sure of its information before sharing it. For the media, access to such information is difficult and usually gained only through leaks. Its claims of being neutral are accepted by Western nations but challenged by Iran and its supporters. The Iranians complain the IAEA promotes a Western agenda.
The drawing seen by the AP is not the only diagram obtained by the IAEA that raised agency suspicion.
The U.N. agency reported on Nov. 8, 2011, that it had obtained diagrams it suspects shows Iran doing studies in nuclear yields, adding: "The application of such studies to anything other than a nuclear explosive is unclear to the agency." And the senior diplomat on Tuesday confirmed that the graph seen by the AP was indeed one of those cited by the IAEA.
Shahriari — the scientist suspected by the IAEA to have made and modified the diagram and provided the spread-sheet with the right information — was assassinated two years ago. Iran said it believes Israeli agents were responsible for the killing.
"Nobody would have understood the original, so he modified it into an artificial unit to make his case," the first diplomat said.
Albright said Friday that the explanation "made sense."
"The IAEA has always said that they had additional information to add to their information calculating the (nuclear) explosive yield," he told the AP.
The first diplomat agreed with critics who said similar graphs can be found in textbooks, the internet and other public sources. But he said the agency found that irrelevant because of the totality of its information about Iran's alleged interest in plotting the force of a nuclear explosion, which included the drawing seen by the AP. In that context, the drawing and other similar to it heightened agency concerns, he said.
Olli Heinonen, who was in charge of the IAEA's Iran file until 2010, said criticism that the drawing was in English was invalid because Iranian researchers use English in much of their work.