An Israeli fighter; the Natanz nuclear site. (Collage Tablet Magazine; Natanz satellite photo: Space Imaging Middle East; IAF F-16 photo: Israeli Air Force)

Can Israel successfully bomb Iran’s nuclear facilities? That was the question that my colleague Whitney Raas and I set out to answer in our 2007 article “Osirak Redux? Assessing Israeli Capabilities to Destroy Iranian Nuclear Facilities,” published in the journal International Security. Relying on extensive mathematical calculations, we argued that an Israeli Air Force strike had at least as good a chance of succeeding as Israel’s 1981 raid on the Osirak nuclear reactor in Iraq. Success wasn’t a guarantee—but it was definitely within the realm of the achievable.

Things have changed since then. For starters, we learned in 2009 that Iran has a second uranium-enrichment facility buried deep in a mountain just north of Qom. Indeed, an International Atomic Energy Agency report released this month indicates that Iran’s nuclear program continues to make steady progress. That report has only intensified talk of a possible Israeli strike. The question is whether such a strike still has a good chance of succeeding. I believe it does.

First, a brief review of where things stood four years ago.

While the Israeli Air Force lacked the capability to target all Iranian nuclear facilities, it could, we believed, attack the most critical subset: the uranium-conversion facility at Esfahan, the fuel-enrichment plant at Natanz, and the heavy-water production plant and heavy-water reactor under construction at Arak. (We did not include the light-water reactor at Bushehr in our calculations as it is not a major proliferation threat and because, unlike some of the other facilities, it is near the Persian Gulf coast and thus it can be targeted by Israeli submarine assets.) Of those three critical targets, only the Natanz facility would pose a real challenge for the Israelis to destroy because it is underground and very large. Esfahan and Arak, although they are also large, are above ground and thus relatively easy to target. The probability of success against the facilities Esfahan and Arak—using 25 F-151 and 25 F-161 aircraft carrying so-called bunker-buster bombs, refueled by KC-707 aerial tankers—was over 90 percent. But the probability of successfully targeting the massive Natanz facility was only about 45 percent, assuming the Israeli Air Force used 25 5,000-pound bunker busters.

In terms of range, we calculated that Israeli aircraft could arrive at and return from a target 1,700 kilometers away with a full weapons load without refueling. (The shortest distance from the Negev to Natanz, for example, is 1,750 kilometers, which means planes would have to refuel in the air for any of the routes.) We determined three basic routes to the targets: fly north and east along the Turkey-Iraq border (or over Syria part of that distance); fly almost due east across Iraq; or fly south and east across Saudi Arabia and then north across the Persian Gulf. The Iraq route would be shortest, but highly sensitive to the United States, creating potential interference from U.S. planes. The Saudi route would be difficult given the long distance, unless the Saudis tacitly allowed Israeli tankers in their airspace to refuel. The Turkey or Syria-Turkey route flew close to Turkish and Syrian surface-to-air missiles and fighter bases but seemed plausible because Israel could refuel in international air space over the Mediterranean.

Iranian air defense is weak, and it relies on either advanced but short-range Russian systems such as the Tor-M1 surface-to-air missile, or on longer-range but much older systems supplied by the United States during the time of the Shah, such as Improved HAWK missiles. Iranian fighter pilots do not receive substantial flight training and were not very effective during the war with Iraq in the 1980s. Therefore, we believed that it was unlikely that Israel would lose more than 10 percent of its aircraft, with a high probability of losing none, though some aircraft might have to abort the mission if engaged by Iranian missiles or fighters.

Since we published our paper, the Israeli Air Force has demonstrated its ability to conduct extremely precise counterproliferation strikes. On Sept. 6, 2007, Israel targeted an under-construction nuclear reactor near Dayr az Zawr, in the eastern part of Syria. The operation was a complete success: The Israeli Air Force incurred no losses, and it proved its strike capability and its ability to penetrate Syrian air defense—and possibly Turkish, as the flight path may have crossed Turkey—with impunity. This gives substantial confidence that Israel will be capable of addressing the Iranian air-defense challenge effectively.

The withdrawal of U.S. forces from Iraq means that the Iraq route is now much more feasible. Without U.S. aircraft and radars, Iraq will have minimal ability to control its airspace. A route crossing northern Syria (which may still be in turmoil next year) and Iraq would likely encounter no significant opposition if well-planned.

In addition, the Israeli Air Force has modernized and expanded its aerial tanker fleet. Several sources have confirmed that the United States has been transferring bunker-buster bombs to the Jewish state. Finally, the Israeli Air Force has conducted training missions with simulated operations as far as Gibraltar at the western edge of the Mediterranean, which indicates it could effectively organize a very large long-range strike.

Set against Israel’s demonstrations and acquisitions is the 2009 revelation of a second Iranian uranium-enrichment site: the Fordow facility near Qom. Fordow is substantially tougher to target than Natanz, since it is located in a mountain under a reported 90 meters of rock and concrete. However, it is also smaller than Natanz, intended to hold only 3,000 centrifuges compared to Natanz’s 50,000. If a single large bomb detonated in the centrifuge hall at Fordow, those 3,000 centrifuges would be rendered unusable.

In thinking about targeting a facility buried so deep, I make a few assumptions. First, I assume the Iranians have ensured that neither the door to the facility nor the ventilation shafts lead directly into the centrifuge hall. If they have not, this presents a vulnerability obvious to anyone who has seen the movie Star Wars: A single penetrating weapon accurately targeted could reach the centrifuge hall by traveling down one of these openings. Even minimal construction efforts, such as right angle bends in a ventilation shaft, could eliminate or at least reduce this possibility.

Therefore, I assume that the Israeli Air Force will have to do things the hard way, targeting many weapons on a single aimpoint in order to burrow through the rock. To my knowledge this has never been attempted against such a deeply buried target, and there are two principal challenges, both of unknown difficulty. First, the weapons themselves, dropped from miles away and thousands of feet in the air, must arrive at very close to the same angle in order to create a pathway each subsequent weapon can follow. Otherwise much of the penetrating power of the bombs will be wasted. The second unknown is the spoil problem, where the sides of the pathway, destroyed by previous explosions, clog the pathway for subsequent bombs.

Using the same methodology Raas and I used in our International Security paper, and making some assumptions about the two problems noted above, I calculated the probability of the Israeli Air Force successfully targeting the Fordow facility. I assume 25 F-15I aircraft, each carrying one 5,000-pound bunker buster and two 2,000-pound bunker busters, would expend all 75 of those bombs on a single aimpoint at Fordow. If angle-of-arrival control is good, the Air Force could have between a 35 percent and 90 percent chance of at least 36 of those weapons arriving on the same aimpoint (this calculation is very sensitive to assumptions about individual weapons). If the spoil problem compounds the depth by only 30 percent or less (in other words, pulverized and collapsing rock adds the equivalent of no more than 27 meters of solid rock) this would likely be sufficient to have at least one weapon penetrate the facility. It thus seems plausible that Fordow can be targeted successfully, if my assumptions are correct.

But make no mistake: It would be a very complex operation. Each weapon release would have to be delayed so that the explosion of the one before it did not damage the following one as it was about to impact the target. This operation would require at least 40 minutes from the time of first bomb release to the last, assuming only 30 seconds of separation between them (which may be insufficient, as the plume from each strike will interfere with laser designation). This would give even the inferior Iranian air defense substantial opportunity to interfere with the strike, and the F-15Is might be forced to ditch their heavy bomb load in order to maneuver.

Additionally, as all of the F-15Is would be dedicated to Fordow, this option would require F-16Is carrying the smaller 2,000-pound bunker buster to be used against the Natanz facility, which would make that strike more difficult. However, IAEA inspections note that as of this month centrifuges have been installed in only one of the two halls at Natanz, thus cutting in half the size of the target and making it more manageable.

I believe that the Israeli Air Force, thanks in part to its efforts since 2007, retains the ability to successfully strike Iranian nuclear facilities. Still, the Fordow facility is substantially more challenging to target than Natanz, and it requires an unprecedented level of precision. This would make the operation more difficult than the original Osirak strike and many times more difficult than the strike against Dayr az Zawr.