Since World War II air power has been key to the American way of warfare–and civilian casualties have been a constant result, from Japan and Korea to Southeast Asia and now Afghanistan and Iraq. This year a seeming surge in airstrikes has led to a corresponding spike in civilian casualties. For example, in a two-day span in May, an airstrike in southern Afghanistan killed at least twenty-one civilians, while a US helicopter attack north of Baghdad killed five civilians, including two children. Yet very little is known about the air war. Due to an apparent disregard by the mainstream media, with a few notable exceptions, the full story remains one of the best-kept secrets of the Iraq War.

What we do know is that since the major combat phase of the war ended in April 2003, the United States has dropped at least 59,787 pounds of cluster bombs in Iraq–the very type of weapon that Marc Garlasco, the senior military analyst at Human Rights Watch (HRW), calls “the single greatest risk civilians face with regard to a current weapon that is in use.” And expert opinion argues that rocket and cannon fire from US aircraft may account for most coalition-attributed Iraqi civilian deaths. The Pentagon has restocked hundreds of millions of dollars’ worth of these weapons in recent years.

The question of cluster-bomb use is especially shadowy. This is hardly surprising. After all, at a time when many nations are moving toward banning cluster munitions–at a February 2007 conference in Oslo, forty-six of forty-eight governments represented supported an international ban by 2008–the United States stands with China, Israel, Pakistan and Russia in opposing new limits of any kind.

A cluster-bomb unit (CBU) bursts above the ground, releasing hundreds of smaller submunitions, or “bomblets,” that increase the weapon’s kill radius. It’s a weapon, Garlasco notes, that “cannot distinguish between a civilian and a soldier when employed because of its wide coverage area. If you’re dropping the weapon and you blow your target up, you’re also hitting everything within a football field. So to use it in proximity to civilians is inviting a violation of the laws of armed conflict.” Worse yet, US cluster munitions have a high failure rate. A sizable number of dud bomblets fall to the ground and become de facto landmines, which, Garlasco points out, are “already banned by most nations.” He adds, “I don’t see how any use of the current US cluster-bomb arsenal in proximity to civilian objects can be defended in any way as being legal or legitimate.”

In the opening phase of the Iraq War (March-April 2003), coalition forces dropped almost 2 million cluster submunitions, according to HRW. In response to a Freedom of Information Act request last year from the Mennonite Central Committee, which has studied the use of cluster munitions for more than thirty years, the Air Force said sixty-three CBU-87s were dropped in Iraq between May 1, 2003, and August 1, 2006 (since each CBU-87 contains 202 bomblets, that would be a total of 12,726 submunitions). The US Central Command Air Forces, or CENTAF, told me earlier this year that “there were no instances” of CBU use in Iraq in 2006, yet when I attempted to clarify the possible dating discrepancy, CENTAF refused to confirm that no CBUs were dropped between January 1 and August 1 of last year.

Evidence of the danger to civilians from these weapons can be found in US military documents. A June 2005 internal memorandum from the Army’s 42nd Infantry Division describes how a 15-year-old Iraqi shepherd “was leading the sheep through north Tikrit, near an ammo storage site, when he picked up a UXO [unexploded ordnance] from a cluster bomb. The UXO detonated and he was killed.” Asked to pay $3,000 to his mother in compensation for the boy’s life, the Army granted that his death was “a horrible loss for the claimant” but concluded there was “insufficient evidence to indicate that U.S. Forces caused the death.”

Iraqi documents also chronicle the effects of cluster munitions. A September 2006 report by the Conservation Center of Environment & Reserves, an Iraqi NGO, examined alleged violations of the laws of war by US forces during the April 2004 siege of Falluja. According to its partial list of civilian deaths, at least fifty-three people were killed by air-launched cluster bombs. An analysis of data collected by another NGO, the Iraqi Health and Social Care Organization, showed that between March and June 2006, of 193 war-injured casualties analyzed, 148 were the result of cluster munitions.

Aside from cluster bombs, Air Force officials acknowledge that coalition aircraft dropped at least 111,000 pounds of other types of bombs in Iraq in 2006, in a total of 10,519 “close air support missions.” This figure does not include guided missiles, unguided rockets or cannon rounds; nor, according to a CENTAF spokesman, does it take into account the munitions used by some Marine Corps and other fixed-wing coalition aircraft, or any Army or Marine helicopter gunships. Nor does it include munitions used by the armed helicopters of the many private security contractors flying their own missions in Iraq. If the military released the figures for rockets, which can be outfitted with various warheads and fired from either fixed-wing aircraft or military helicopters, they might prove impressive: In 2005 the Army signed a five-year, $900 million contract with General Dynamics for the Hydra-70 rocket; by this April, $502 million in orders had been placed.

The number of cannon rounds–essentially large-caliber “bullets”–fired by CENTAF aircraft is also a closely guarded secret. The official reason given is that “special forces often use aircraft such as the AC-130″ gunship, which fires cannon rounds, and “their missions and operations are classified.” However, an idea of the number of cannon rounds expended can be gleaned from a description of a single operation on January 28, when, according to CENTAF, F-16s and A-10 Thunderbolts not only “dropped more than 3.5 tons of precision munitions” but also fired “1,200 rounds of 20mm and 1,100 rounds of 30mm cannon fire” in a five-square-mile area near the southern city of Najaf.

According to Les Roberts, co-author of two surveys of mortality in Iraq published in the British medical journal The Lancet, “Rocket and cannon fire could account for most coalition-attributed civilian deaths.” He added, “I find it disturbing that they will not release this [figure], but even more disturbing that they have not released such information to Congressmen who have requested it.” Roberts himself witnessed the destruction caused by cannon fire in Baghdad’s vast Shiite slum, Sadr City, in 2004. He recalls passing through 100- to 200-meter-wide areas of neighborhoods that had been raked by such fire. “It wasn’t one house that was beat up,” he recalled. “It would be five, six, seven buildings in a row.” This portrait of devastation is echoed by independent Iraqi journalist Ali al-Fadhily, who told me he had witnessed helicopter gunships in action, noting: “The destruction they caused was always immense and casualties so many. They simply destroy the target with every living soul inside. The smell of death comes with those machines.”

While the destructive capacity of helicopter gunships has been well documented, the actual scale of use is hard to pin down. Flight hours do give some indication. According to James Glanz of the New York Times, Army helicopters logged 240,000 flight hours in 2005, 334,000 in 2006 and projections for 2007 reach 400,000. These numbers don’t include Marine squadrons that operate in Anbar Province, heliborne missions by private security contractors or those of the nascent Iraqi Air Force.

Given the official secrecy and lack of media interest, the human fallout of this one-sided air war is difficult to determine. One valuable source is the national cross-sectional cluster sample survey of mortality since the 2003 invasion, published in The Lancet last October. It used well-established survey methods that have proven accurate in conflict zones from Kosovo to Congo. Carried out by epidemiologists at Johns Hopkins University and Iraqi physicians organized through Baghdad’s Mustansiriya University, the Lancet study estimated 655,000 “excess Iraqi deaths as a consequence of the war.” The study also found that from March 2003 through June 2006, 13 percent of violent deaths were caused by coalition airstrikes. If the 655,000 figure, which includes more than 601,000 violent deaths, is accurate, this would mean that about 78,000 Iraqis had been killed by bomb, missile, rocket or cannon up to last June. There are indications that the air war has taken an especially grievous toll on Iraqi children. Figures provided by the Lancet study suggest that 50 percent of all violent deaths of Iraqi children under 15 in that same period were due to coalition airstrikes.

That an occupying power regularly conducts airstrikes in and near dense population centers should have raised serious questions in the mainstream media; unfortunately, reports on the air war are sparse and mostly confined to regurgitations of military announcements. HRW’s Garlasco notes, “Because of the lack of security, we’ve had no one on the ground for three years now, and so we have no way of knowing what’s going on there.” Les Roberts points out that in the first years of the war, more bombs were dropped by American planes than there were attacks by insurgents using roadside explosives and car bombs, yet we never saw evidence of this in the press. With the military unwilling to tell the truth and unable to provide the stability necessary for NGOs to operate, it falls to the media to begin ferreting out substantive information. It seems, however, that until reporters begin bypassing official US military pronouncements and locating Iraqi sources, we will remain largely in the dark regarding the secret and deadly US air war in Iraq.