Fake AP Graph Exposes Israeli Fraud and IAEA Credulity

 Israeli Fraud Fake AP GraphThat Associated Press story displaying a graph alleged to be part of an Iranian computer simulation of a nuclear explosion — likely leaked by Israel with the intention of reinforcing the media narrative of covert Iranian work on nuclear weapons – raises serious questions about the International Atomic Energy Association’s (IAEA) claim that it has credible evidence of such modeling work by Iran.

The graph of the relationship between energy and power shown in the AP story has now been revealed to contain absurdly large errors indicating its fraudulence.

Those revelations indicate, in turn, that the IAEA based its publication of detailed allegations of nuclear weapons-related Iranian computer modeling on evidence that should have been rejected as having no credibility.

Former senior IAEA inspector Robert Kelley, who has challenged the accuracy of IAEA reporting on Iran, told Lobe Log in an e-mail that “It’s clear the graph has nothing to do with a nuclear bomb.”

“The pretty, symmetrical bell shaped curve at the bottom is not typical of a nuclear explosion but of some more idealized natural phenomena or mathematical equation,” he said. “Clearly it is a student example of how to perform integrals to which someone has attached some meaningless numbers.”


Nuclear physicists Yousaf Butt and Ferenc Dalnoki-Veress also pointed out that the graph depicted by AP is not only so rudimentary and crude that it could have been done by an undergraduate student, but is based on a fundamental error of mind-numbing proportions.

The graph shown in the AP story plots two curves, one of energy versus time, the other of power output versus time. But Butt and Dalnoki-Veress noted that the two curves are inconsistent. The peak level of power shown in the graph, they said, is nearly a million times too high.

After a quick look at the graph, the head of the Department of Physics and Astronomy at Cal State Sacramento, Dr. Hossein Partovi, observed, “[T]he total energy is more than four orders of magnitude (forty thousand times) smaller than the total integrated power that it must equal!” Essentially, the mismatch between the level of total energy and total power on the graph is “more than four orders of magnitude”, which Partovi explained means that the level of energy is 40,000 times too small in relation to the level of power.

One alert reader of the account of the debunking of the graph at the Mondoweiss blog cited further evidence supporting Kelley’s observation that the graph shown by AP was based on an another graph that had nothing to do with nuclear explosions.

The reader noted that the notation “kT” shown after “energy” on the right hand scale of the graph does not stand for “kilotons” as Jahn suggested, but “Boltzmann constant” (k) multiplied by temperature (T). The unit of tons, on the other hand, is always abbreviated with a lower case “t”, he pointed out, so kilotons would be denoted as “kt”.

The reader also stated that the “kT” product is used in physics as a scaling factor for energy values in molecular-scale systems, such as a microsecond laser pulse.

The evidence thus suggests that someone took a graph related to an entirely different problem and made changes to show a computer simulation of a 50 kiloton explosion. The dotted line on the graph leads the eye directly to the number 50 on the right-hand energy scale, which would lead most viewers to believe that it is the result of modeling a 50 kiloton nuclear explosion.

The graph was obviously not done by a real Iranian scientist — much less someone working in a top secret nuclear weapons research program — but by an amateur trying to simulate a graph that would be viewed, at least by non-specialists, as something a scientist might have drawn.

Although AP reporter George Jahn wrote that officials who provided the diagram did so “only on condition that they and their country not be named”, the country behind the graph is not much of a mystery.

Blogger Richard Silverstein has also provided a key piece of evidence indicating that the source of the graph was Israel’s intelligence agency, the Mossad. He reports that a “highly-placed Israeli source” told him the diagram “was stolen by the Mossad from an Iranian computer” using one of the various malware programs deployed against Iran.

That information links the graph in question to a longer stream of documents, supposedly obtained by Israeli intelligence from inside Iran’s nuclear program and then given to the IAEA over the past few years.  Former IAEA Secretary General Mohammed ElBaradei refers in his memoirs to documents provided by Israel in 2009 “purportedly showing that Iran had continued with nuclear weapons studies until at least 2007.” ElBaradei adds that the Agency’s “technical experts” had “raised numerous questions about the documents’ authenticity”, and suggested that US intelligence “did not buy the “evidence” put forward by Israel” in its 2007 National Intelligence Estimate.

Jahn’s story indicates that this and similar graphs were the basis for the IAEA’s publishing charges by two unnamed states that Iran had done computer modeling that the agency said could only have been about nuclear weapons.

Jahn cites a “senior diplomat who is considered neutral on the issue” as confirming that the graph accompanying his story was one of “a series of Iranian computer-generated models provided to the IAEA by the intelligences services of member nations.”

Those “computer generated models” were discussed in the November 2011 report, which referred to “[i]nformation provided to the Agency by two Member States relating to modelling [sic] studies alleged to have been conducted in 2008 and 2009 by Iran….”  The unnamed member states were alleging that the Iranian studies “involved the modelling [sic] of spherical geometries, consisting of components of the core of an HEU nuclear device subjected to shock compression, for their neutronic behaviour at high density, and a determination of the subsequent nuclear explosive yield.”

Nothing in that description of the alleged modeling is documented by the type of graph shown by the AP story.

The IAEA report concludes by saying, “The information also identifies models said to have been used in those studies and the results of these calculations, which the Agency has seen.”

gareth porterIn other words, the only evidence that the IAEA had actually seen was the graphs of the alleged computer modeling, of which the graph shown in the AP story is alleged to be an example. But the fact that data on that graph has been credibly shown to be off by four orders of magnitude suggests that the Israeli claim of Iranian computer modeling of “components of the core of an HEU nuclear device subjected to shock compression” was completely fabricated.

Former IAEA Inspector Kelley also told Lobe Log that “We can only hope that the claim that the IAEA has relied on this crude hoax is false. Otherwise their credibility has been shattered.”

Gareth Porther
Lobelog Foreign Policy

Saturday, 1 Decmeber 2012


  1. says

    First, let me explain my background. I was a physics major at Caltech before turning to computer software and computer component design and becoming a Chief Technology Officer.

    I am also a progressive. As such, I want to see the end of nuclear weapons. Any additional country obtaining nuclear weapons is a very bad thing, and I am saddened to see so many in the progressive movement, which used to favor disarmament, being so relaxed about the possibility of nuclear proliferation.

    As a physicist (I still think of myself as such), I have to say that the science presented in this article by Gareth Porter is complete bunk. Porter’s argument that the graph is too inaccurate to be real just doesn’t hold water, as I’ll explain below.

    That does not mean that the graph is necessarily real; but it does mean that there is not the slightest remaining evidence of fraud, and we need to be very worried about the possibility that this threat is real.

    We are progressives. We don’t like war. If we want disarmament (we used to), we should see this as a very bad sign. We should see it as a call to action to reduce the nuclear threat we face.

    Now for the details.

    The Boltzmann-constant argument is wrong. It is true that “kT” is often used by physicists to refer to this constant times temperature. But that gives a value of energy in unspecified units. It makes no sense to combine this with a specific unit to say “kT/sec.” Physicists do sometimes do some strange things in writing units, but this is not normal. I have searched, and I can find no use of “kT/sec” where “T” stands for temperature.

    Also wrong is that kiloton must be abbreviated with lower case letters. It took me seconds to do a Google search and find plenty of documents written by scientists and governments that abbreviate it as “kT.”

    Of greater concern about this document is the discrepancy between the left and right axes. I can only speculate. One possible speculation is that this is a forgery. Another is that it is not the result of a computer simulation but a crude example of what the output of that simulation might look like once run, and the author just crudely put together the example.

    But what might be the best speculation is that it is a simple typo. If you are trying to describe the total output, certainly the correct units on the right side of the graph are kilotons. On the other hand, you are going to measure power in a real experiment using calorimetric techniques. That is, you measure how much heat is generated. You might end up with a measurement in something like calories per second. Or BTUs per second. Or therms per second. In fact, if the units on the left were therms/sec rather than kilotons/sec, the graph is no longer off by a factor of 40,000. Instead it is within a few percent of being exactly right (based on pixel counts). That few percent error is as likely to be my measurement error as it is an error in the graph.

    • says

      I’ll add that a therm/sec is just a few percent different than another unit that is commonly used by nuclear physicists, a Joule/shake. I get an even smaller resulting error when I use Joule/shake, but I don’t think my measurements are accurate enough to say which of those is actually right.

  2. JoeWeinstein says

    Speaking of credulity! In this case apparently either the author’s or the editors’ credulity knows no bounds!

    Namely: the abbreviated title of the piece – ‘Israeli fraud’ – is not what Porter’s text actually demonstrates.

    As Porter’s text says, the graph actually shown (even if its units were corrected) does not document the alleged specific type of modeling. It does however – if you buy the claim that the graph was hacked from an Iranian computer – directly document that someone in Iran – likely a student working at a low level of cyber-security – was getting a background in energy-release physics. As the article notes, kT is indeed a standard unit of molecular energy – namely in a gas at given temperature. (k=Boltzmann’s constant, T = absolute (i.e. Kelvin) temperature.

    Note that – although perhaps keen to hint at ‘Israeli fraud’ – Porter does NOT in fact actually demonstrate (or even claim) that, concerning this particular graph, that the Israelis (or whoever else might have hacked to obtain and furnish the graph) were making a fraudulent claim. Insofar as we can read in Porter’s text, the graph providers did NOT claim that this particular graph was itself an example of the specific alleged type of modeling (i.e. of nuke-making explosions).

    The Iranians are not stupid and Mossad (or whoever from whatever nation actually hacked in) is not infallible. Likely for starters the best the hackers could do – and remember this is some years ago – was to demonstrate an ability to hack into some lower-security-level computers that were mainly used by students in one or another Iranian university or institute nuclear studies program. Porter’s prior articles have mentioned existence of several such programs.

    The article is concerned about the IAEA’s maybe being at times overly credulous. However, one should bear in mind why suspicions exist concerning possible military use of Iran’s extensive network of nuke programs. They exist because of the special nature and history of the regime. The regime leads the world in support for terrorist organizations and in its leaders’ outrageous and genocide-inciting rants.

    That’s why Israel and the USA and various neighbors of Iran have strong PRECAUTIONARY reason to suspect the worst – especially when Iran’s processed amounts of nuclear material immensely transcend those needed for medicine and the like, and the country’s need for nuke power is unusually low: Iran is particularly blessed with not only fabulous oil and gas reserves but also ample suitable area for massive future solar power.

Leave a Reply

Your email address will not be published. Required fields are marked *