By Rick Atkinson
Washington Post Staff Writer
Monday, October 1, 2007
As Gen. John P. Abizaid began his second year at U.S. Central Command in July 2004, the simple solutions he had hoped would defeat improvised explosive devices in Iraq seemed further away than ever. More than 100 American soldiers had been killed by bombs in the first half of the year, and IED attacks were spiraling toward an average of 15 per day.
Eager for creative ideas, Abizaid told Centcom subordinates in August that he would accept what became known as "the 51 percent solution": If a new counter-IED gadget or technique had a better than even chance of success, it would be welcome in the theater. "Listen, if you have something that's greater than 50 percent, then get it forward," he also told Brig. Gen. Joseph L. Votel, director of the Pentagon's Joint IED Task Force. "I've got the greatest testing ground in the world in Iraq."
That testing ground was soon put to use in IED Blitz, an elaborate experiment concocted in late summer by the Pentagon's joint staff.
Spotting bombs had proved fiendishly difficult. Some soldiers fastened rally lights on their Humvees for better visibility at dusk, or mounted leaf blowers on truck bumpers to clear the debris often used to camouflage IEDs. Aerostats -- small blimps with closed-circuit cameras tethered 300 to 1,000 feet up -- watched the landscape for insurgent activity. But soldiers staring at television monitors often found their concentration waning after just 30 minutes.
IED Blitz would focus as many forms of surveillance as possible in a "persistent stare" at a bomb-infested 20-kilometer stretch of Route Tampa, just south of Balad on the road to Baghdad. The blitz would enlist satellites, U-2 spy planes, 14 Mako unmanned aerial vehicles, a pair of larger I-Gnat drones, and the Horned Owl, a Beechcraft turboprop airplane equipped with ground-penetrating radar used to assess whether road shoulders had been disturbed by digging.
Attacks had grown increasingly extravagant, with "daisy-chained" munitions that included as many as 22 artillery shells wired together to explode simultaneously in a 300-yard "kill zone." Intelligence analysts assumed that such ambush sites took hours or even days to prepare. On the basis of past attack patterns, they predicted that 60 IEDs would be planted in 75 days on this short segment of Route Tampa.
Hundreds of thousands of photographs would be snapped as part of a technique called "coherent change detection." Two images of the same scene taken at different times would be compared, pixel by pixel, to spot changes in the landscape -- such as the anomalies caused by an insurgent planting a bomb. Ground convoys could be warned, and, if the reconnaissance was nimble, hunter-killer teams could flush emplacers or triggermen.
The operation, estimated to cost at least $3 million, would be directed from Defense Department offices leased in Fairfax County. "The only unacceptable outcome is to spend all this time and money and not have a definitive outcome because we didn't dedicate enough assets," a senior Pentagon official warned in one planning session. "Don't do it half-assed."
Blitz began on Tuesday, Aug. 24, 2004. So brilliant were the digital color images that analysts could read the brand names on plastic water bottles littering the roadside. They could distinguish an apple from a pomegranate at a fruit stand.
What they could not see was a bomb or a bomber. Technical problems soon emerged. A U-2 making two passes above the road seven hours apart had to fly routes within 5 degrees of each other for those pixels to align; the success rate was "way less than 30 percent," according to an Air Force officer. The I-Gnat routes needed to overlap within about 1 percent of each other, a comparably difficult task. Even blowing trash, a feature of every Iraqi landscape, sometimes made coherent change detection incoherent.
Sandstorms and high winds also plagued Blitz. Insurgents apparently monitored drone takeoffs from Balad airfield to calculate when they could safely move about; when two days of foul weather grounded the aircraft, six IED attacks occurred on Route Tampa. Numerous images showed Iraqis in pickup trucks staring into the sky and making obscene gestures at the drones, which were as noisy as lawn mowers.
When automated photo analysis did not work properly, analysts in Fairfax County were forced to manually review the pictures. "It was an eyeball event," the Air Force officer later recalled. "You have 5,000 images taken in the morning and 5,000 at night. The day shift would take 5,000, the night shift 5,000." The Blitz roadway was divided into five-kilometer segments with teams assigned to scrutinize each one. "If a paper cup was on that route and wasn't there yesterday," the officer added, "you knew it."
Blitz organizers had intended for the operation to last four weeks, but Centcom pressed for better results. A month passed, then two. Keeping a round-the-clock persistent stare grew more difficult, particularly as demands intensified to shift resources elsewhere in Iraq. "Slowly but surely," an Army colonel said, "it unraveled."
Eight of the 14 Makos crashed. Horned Owl also was a disappointment, as Abizaid's deputy, Air Force Lt. Gen. Lance L. Smith, found when he flew a mission in Iraq. The radar was bedeviled by false positives, including oil barrels and car hulks. The Iraqis, Smith observed, were "wonderful buriers." The aircraft would be sent home for further tweaking.
The most disheartening day came on Thursday, Nov. 4. By chance, virtually all surveillance assets -- satellites, U-2s, drones -- happened to be focused simultaneously on one small swatch of Route Tampa. Traffic appeared normal. Two hours later, another sequence of images revealed a scorched crater where a bag of artillery shells triggered by a detonation wire had just killed one American soldier in a truck and severed the leg of another. Dozens of photos showed the burning vehicle veer across the median, and rescue vehicles convene at the site. No images revealed the IED being placed, or the triggerman.
Analysts soon surmised that bomber cells around Balad in late summer had shifted "to a just-in-time device-placement method," as a Defense Intelligence Agency analyst put it. Instead of requiring hours or days to survey an ambush site and bury a device, "hasty emplacement" took two hours or less.
Blitz ended on Nov. 15. In 10 weeks, 44 IEDs had detonated or were discovered by ground clearance teams. Asked how many had been detected by aerial surveillance, the Air Force officer said, "To be honest with you, I can't say any of them.
"We had only a 20-kilometer stretch," the officer added. "There are thousands of kilometers in Iraq." By the end of 2004, about 5,600 IED attacks had occurred across the country, according to Centcom.
If "the results were less than expected" in Blitz, as the DIA analyst concluded, intelligence experts also learned a great deal: The surveillance sensors might not find bombs, but they could conceivably be used to hunt bombers by watching where vehicles spotted near an ambush site were subsequently driven, a technique later called backtracking. "It's a vehicle-borne insurgency," the Air Force officer said.
That realization would take many months to bear fruit. The immediate lesson of Blitz was summarized by an Army lieutenant colonel. "You can do the unblinking eye," he said, "but you can't do it everywhere, and you can't do it very long."
Blitz underscored that in the theater and at home, U.S. efforts remained fixated on "the fight at the roadside": finding or neutralizing planted bombs, and attacking those who buried them. But killing an emplacer rarely disrupted a bomber cell for more than several days, and few insurgents captured by U.S. forces proved to be ringleaders, financiers or IED builders.
A classified U.S. government proposal called the Cerberus Project specifically targeted bombmakers as "disproportionately valuable to the terrorist operation chain" because of their technical skills. By collecting forensic signatures of individual bombers, such as fingerprints, handcrafted circuits and soldering marks, analysts could identify patterns and better understand the IED networks spreading through Iraq.
Britain and Israel had similarly targeted bombmakers in Northern Ireland and Lebanon, respectively, and the Joint Chiefs of Staff declared "the enemy bombmaker as the operational center of gravity" for U.S. warfighters, according to a senior officer involved in the operation. Yet such painstaking efforts would not pay off for months or years. A Cerberus Project briefing paper in August 2004 warned that "the window of opportunity to prevent the terrorist and insurgent bombmaking knowledge base from being transferred, or evolving . . . is closing."
Lt. Gen. Smith also advocated a "left of boom" strategy that would attack those IED networks long before a device detonated. He was appalled that many field commanders seemed to accept bomb casualties as an inevitable cost of fighting the insurgency. "We have got to go after the system," Smith urged. "We have got to find out where they're made."
But "the system" was amorphous, shadowy. Battling the IED itself, particularly through heavier armor, electronic jammers and better patrol tactics was tangible and urgent, even if it remained "right of boom, or right at boom," as a senior Army officer put it.
"It's so easy to just defend against the device, because that's what's cleaning your clock," a top Defense Department scientist said later. "We needed to move to the left of boom, but we didn't know how to do it." One briefing chart included a picture of a light bulb with the caption, "Does anybody have a bright idea?"
Abizaid had asked the Pentagon for "a Manhattan Project-like" effort to fight IEDs, which he called "my number one threat in Iraq." Like Smith, he worried that too many soldiers looked at roadside bombs as they did snipers or other vexing battlefield threats, without realizing that rising bomb casualties eroded American support for the war. Paradoxically, IEDs were "a strategic issue," Abizaid observed, "but not necessarily a tactical issue."
Whether the nation could conjure an IED solution, as the Manhattan Project had delivered the atomic bomb in 1945, also remained uncertain, given how little of the country seemed mobilized for war. "We asked for the Manhattan Project," Abizaid would later joke, "and we got the Peoria project."
Through the winter of 2004-05, bombs grew more lethal and bombers more ingenious. Explosives often were "boosted" with propane or gasoline canisters. Triggers now included pressure plates, walkie-talkies and long-range cordless phones with a range of three to five miles -- common in Iraq, where no Federal Communications Commission restricted their usage. In December 2004, insurgents in Baghdad lured police into a house, then detonated an estimated 1,700 pounds of explosives, killing at least 28 people, according to a U.S. Army War College study. Such death traps soon were known, inelegantly, as "house-borne" IEDs.
When Marines fought to reclaim Fallujah in November 2004, explosives experts examining a Euphrates River bridge discovered artillery shells hidden behind the metal base plates of light stanchions, with a triggering wire leading to a palm grove. Other rigged shells were found in drainpipes, flowerpots and phony, hollow curb stones made from plaster molds. A Marine regimental commander later recalled wondering, "What are we getting into?"
Car bombs, which had averaged one per week in Iraq in January 2004, kept doubling and redoubling. From September to December, 247 "vehicle-borne" IEDs targeted coalition forces, who learned to watch for the sagging automobile suspensions that might indicate a trunk packed with artillery shells. Mayhem often brought political consequences: After an explosion in Iraq killed eight Ukrainian soldiers in early January 2005, Kiev announced it would withdraw its 1,600 troops from the war by midyear.
A counter-IED field manual compiled by the Army and Marines warned of devices placed in "fake bodies or scarecrows in coalition uniforms." Known to bomb squads as "come-ons," such lures soon used real bodies, including IEDs tucked into the chest cavities of dead Iraqis, whose corpses also were booby-trapped with anti-tampering triggers, according to a Navy explosive ordnance disposal officer.
Insurgent bombmaking prowess derived largely from three former Baathist organizations, which had slipped underground after the 2003 invasion, according to a senior Defense Department official: the Special Republican Guard, the Special Security Organization and the so-called M-21 directorate of the Mukhabarat intelligence service. In the fall of 2004, Centcom drafted a list of 38 suspected insurgency leaders, most of whom were living in Syria, a former top Pentagon official said.
As the Cerberus Project had warned, the bombmaking "knowledge base" soon expanded, with pockets of expertise throughout Iraq. "The number of networks has been placed at 169, but that may be just 169 that we know about," the DOD official said last spring. "Some of them are Sunni rejectionists, some are Sunni Baathists, some are al-Qaeda, some are Shiite. It's a crazy quilt."
A recent intelligence study described fluid, decentralized cells, typically with five to 10 members, including a financier, a bombmaker, an emplacer, a triggerman and often a video cameraman. Some freelance bombmakers "have posted their contract for hire services through the Internet with video footage of past acts serving as promotion and bona fides," the study noted.
If attacks showed complex ingenuity, devices tended to be simple, usually suggesting technical skills equivalent to those of a ham radio operator or a vocational school graduate, according to a DOD scientist. Simplicity made it easier to employ unlettered emplacers, who by late 2004 were generally recognized as being mercenaries rather than ideologues.
Perhaps reflecting the triumph of supply over demand, emplacer fees continued to decline, typically ranging from the equivalent of $300 to as little as $25. Killing a coalition soldier might earn a $700 bonus. In Afghanistan, a recent coalition price list showed that the families of suicide bombers usually were paid $500 to $2,000, with bounties as high as $10,000 for assassinating a NATO soldier.
Joe Votel, the Joint IED Task Force director, had come to regret Abizaid's Manhattan Project allusion. The metaphor implied a facile, scientific solution to IEDs, a technological silver bullet. "That was easy," Votel quipped about the atomic bomb built in the New Mexico desert. "You were in a sanctuary, you developed a bomb, you dropped a couple of them and it was done."
Fighting IEDs had proved far more frustrating than Votel had anticipated when he took the job in October 2003. His staff consisted largely of contractors or military officers lent to the task force for a few months. Expertise was hard to come by; "leveraging academia," as he called it, required a greater knowledge of the scientific community than the task force possessed. Simply exploiting the British know-how accrued in decades of battling roadside bombs in Northern Ireland was annoyingly difficult because of Pentagon "SECRET/NOFORN" rules that barred even close foreign allies from access to secret information.
Battling an IED network was akin to dismembering a cocaine cartel before drugs flooded the market. It required exceptional intelligence, agility and great patience. But with hundreds of bombs detonating every month, the pressure was intense to send as many jammers and other "deliverables" as possible to the field. "You felt an obligation to the warfighter," a young officer said.
Congress encouraged the task force to become "extremely risk tolerant," as one scientist put it, and to finance duplicative efforts such as multiple jammers. According to a senior Senate staff aide, Votel was told repeatedly: "We're going to have lots of failures. That's okay. We want you to push the envelope. Take risks."
Risks were taken. To safeguard vulnerable gunners in Humvee turrets, the task force spent $9.4 million to buy 728 armored suits called the Cupola Protective Ensemble. Shipped to Kuwait for field testing, the outfit proved to be hot and constricting. Engineers scavenged cooling units used in Kiowa helicopters, installed them in the suits and plugged the contraption into the Humvees, which "fried the vehicle alternators," according to an Army colonel, who added, "We hadn't thought this through."
About 300 government programs were scrutinized for "low-hanging fruit" -- mature technologies that could be easily plucked and sent to the war zone: sniper rifles, little Raven drones, aerostats and laser "dazzlers" that temporarily blinded the driver of an oncoming vehicle, including a potential suicide car bomber, and caused him to veer off course. "There was a mad rush to get equipment over as we searched for the magic gizmo," a retired admiral recalled.
The task force identified 17 "capability gaps" -- such as counter-IED training methods and surveillance techniques -- and solicited suggestions from industry, which responded with 851 proposals. Congress also tried to help. When Rep. Duncan Hunter (R-Calif.), chairman of the House Armed Services Committee, touted a Kevlar infantry poncho of dubious practicality, a mischievous staff officer fabricated a photograph of the congressman wearing the outfit, which he labeled "SpongeBob HunterPants" after the Nickelodeon cartoon character. Votel was horrified. "Destroy that," he ordered. "If this gets out, we're dead meat."
Centcom's requirement for "uparmored" Humvee kits jumped to 14,000, according to a committee document, and, in January 2005, Abizaid agreed that no unarmored vehicles should operate outside secure bases in Iraq. But an uparmored Humvee weighed a ton more than its soft-skinned predecessor, with a consequent strain on engines, suspensions, transmissions and tires.
As always in war, frictions accumulated. The simple grew difficult. In late summer 2004, the task force had undertaken what seemed a relatively easy task: buy more explosive-sniffing dogs for the theater. Help was solicited from the British and Israelis, both of whom had extensive experience with "off-leash" dogs that gave a handler about 200 yards of standoff distance from possible bombs.
Protracted discussions ensued at the Pentagon over breeds, training methods and cost. (A trained search dog typically cost $4,000 to buy and $35,000 to deploy.) The Marines admired an Israeli method that used collar-mounted radio receivers to control their dogs, but the animals' linguistic limitations proved problematic: Most understood only Hebrew. "We spent an unbelievable amount of time on this," a former DOD official said.
A Defense Intelligence Agency analyst noted in February 2005 that "off days will occur" with dogs, and that the "find rate" for buried explosives may be only 75 percent. Canine specialists disputed the figure but agreed that a working dog grew easily distracted after 30 minutes, not unlike a soldier watching an aerostat monitor. Ultimately, the task force agreed to finance 48 dog teams in 2005 and another 48 in 2006.
Sometimes the most popular deliverables were decidedly low-tech. A pocket-size pamphlet called the "Visual Language Translator for IED Detection" was shipped overseas by the tens of thousands. Cartoon sketches showed various concealment options, which sympathetic Iraqis could point to in identifying bomb sites: a guardrail, a dumpster, a dead sheep. Cartoon greenbacks implied a handsome reward for anyone who turned in a bombmaker.
And a phonetic pronunciation key gave every soldier the ability to ask, in Arabic, the question that might save his life: Weh-nil kun-boo-leh? Where is the bomb?
Eighty percent or more of the roadside bombs planted in some areas of Iraq used RC -- radio-controlled -- triggers. Warlock Green jammers, which Army engineers had designed from a counter-artillery system, continued to arrive from Thousand Oaks, Calif., where they were largely hand-built by EDO Corp. and cost about $100,000 each.
Because most RC bombs were simple low-power devices, using key fobs or wireless doorbells, more jammers specifically designed to handle that threat also began arriving in the summer of 2004, including EDO's Warlock Red and a jammer popular with Special Forces troops called the Mobile Multi Band Jammer.
Col. Bruce Jette, who earned a PhD in solid state physics from the Massachusetts Institute of Technology and commanded the Army's Rapid Equipping Force, persuaded a firm in Las Cruces, N.M., to build 13 prototype Self-Screening Vehicle Jammers in six weeks for $350,000. Hundreds more would be built and shipped to Iraq. A similar jammer called the IED Counter Electronic Device (ICE) found favor with the Marines, along with a later modification dubbed MICE. (A disdainful competitor insisted it should be called LICE.) Jammers proliferated, in number and in variety.
Votel and his task force had long planned to underwrite a single powerful jammer that would cover as much of the RC spectrum as possible and simplify military logistics. On Dec. 1, 2004, the Army notified industry that it intended to phase out Warlock production in anticipation of buying a common jammer, which would be called Duke. Within days, EDO signaled that it would begin laying off workers, a development so alarming to Duncan Hunter that he blocked an Army request to reprogram $2 billion.
In a stormy, profane confrontation, Votel and Robert Simmons, staff director of the Armed Services Committee, argued over jammer policy, according to two participants in the session. "We have a strategy here," Votel pleaded. "Let us get started on this."
But initial tests on a Duke prototype seemed inconclusive to Simmons. "This is unproven," he told Votel. "We've already seen testing problems. We've got something that's proven to work within a limited spectrum, and you're going to stop that?"
Votel called Gen. Richard A. Cody, the Army vice chief of staff. Hunter and his staff were adamant about shipping more Warlocks into the theater quickly, he reported. Cody called Deputy Defense Secretary Paul D. Wolfowitz, who agreed to keep the Warlock line open. Hunter lifted his hold on the Army reprogramming request, and on Jan. 13, 2005, the Army placed a $56 million order with EDO for 1,440 more Warlocks, evenly divided between Greens and Reds.
Votel liked Simmons, and he admired Hunter's passionate interest in combat procurement. But he felt deep frustration about his inability to convey the complexity of the electromagnetic challenge in Iraq. "I fully admit that I'm a C student on this," he often said, and joked that "I should have paid more attention in electrical engineering class" at West Point. But when another committee staffer insisted, "You know, this isn't rocket science," Votel snapped, "Actually, this is rocket science. This is a hard freaking problem." Lives and billions of dollars were at stake.
The hard problem grew harder by the week. An XVIII Airborne Corps analysis showed that U.S. troops in Iraq were struggling to manage 82,000 radio frequencies, including jammers and communication channels. They also tried to juggle more than 300 databases, such as IED spreadsheets, many of which were not interoperable.
The proliferation of low-power jammers such as Warlock Red and ICE had the salutary effect of nearly eliminating low-power, radio-controlled IEDs. From a substantial majority of RC devices in June 2004, low-power would decline to about 6 percent by the summer of 2005, at a time when the Pentagon reported 4,200 portable electronic jammers in Iraq.
But, as always, insurgents adapted quickly. High-power RC devices proliferated, against which many U.S. jammers were inadequate. By increasing the power of the radio transmitters and switching to a higher frequency -- a simple spectrum analyzer could show a bombmaker the range in which the jammers operated -- the insurgents soon countered the countermeasure, again.
It was painfully obvious that many in the Army had forgotten that the electromagnetic spectrum could be a battle space. Except for artillery direction-finding and signals intelligence such as eavesdropping, Army electronic warfare expertise and equipment had atrophied as the Soviet threat dissolved after the Cold War. "This was fine for the environment of 1990, and it was fine for the environment of 2002," said Rear Adm. Arch Macy, commander of the Naval Surface Warfare Center in Washington. "Then we got to 2003, and the beggars changed the rules of the game."
By early 2005, an Army scientist said, "when it came to dealing with the electromagnetic spectrum, the Army didn't have the competency." The Marines were hardly better. A former Marine commander said that when jammers arrived in his regiment, "all we knew was that a box would show up that made the vehicles hotter."
If only two jammers were available for a four-Humvee convoy, should they go on the first and last vehicles? The first two? The two in the middle? How far apart should the vehicles travel? How fast?
"These things were not known," the retired admiral said. No system existed to train the force. "The tasks were pretty basic," a Navy captain added. "How do you make sure the jammer power is on? Check for proper air flow through the system? How do you know it's working? . . . Tactical employment was largely run by rumors about what worked and what didn't."
As more jammers arrived through the winter, soldiers complained that the boxes sometimes blocked radio transmissions, other jammers and Blue Force Tracker, a satellite-based system that showed commanders the precise location of their troops. "There was no coordination," Macy said. Jammers "interfered with each other, they interfered with other gear, VHF radios, Blue Force Tracker. . . . This was happening more than acceptable. And acceptable is almost zero."
For patrol leaders, "it was either jam or talk," said an electrical engineer who worked as a defense contractor in Iraq and Afghanistan. Some soldiers distrusted the jammers and left them in their shipping containers. "That's a terrible position to put troops in," the retired admiral said. "And it's technically stupid."
Some veterans would later look back with regret. "The enemy can change in hours, or days. We have to be able to do it at least in weeks," said a retired lieutenant colonel who worked for the task force.
More than 250 American soldiers would be killed by another type of IED that was spreading across the battlefield and against which even the best jammers proved useless.
The first confirmed EFP -- explosively formed penetrator -- had appeared in Basra on May 15, 2004, and Votel had briefed Vice President Cheney in late June on the phenomenon, using a model to demonstrate how it worked. The weapon, which fired a heavy copper disc with devastating impact, typically used a passive infrared trigger that detonated the bomb when a sensor detected radiation from a warm passing object, such as a Humvee. Because no radio waves were involved, jammers had no effect.
A Defense Intelligence Agency weapons team had noted in the late 1990s that EFPs with infrared triggers were used by Iranian-backed Hezbollah forces against the Israelis in southern Lebanon at least as early as 1997. The few EFPs that were in Iraq during the early summer of 2004 invariably appeared in Shiite-controlled areas near the Iranian border, such as Basra and southeast Baghdad. That suggested "international linkages" to Iran, Votel told Cheney.
A colonel at the Israeli Embassy had repeatedly warned the task force about infrared-triggered EFPs. "He and other Israelis were pounding on the desk, saying, 'Listen, we've already been through this historically. This is what's going to happen next,' " a task force officer later recalled.
"We should never have been surprised by that," an Army colonel with long experience in Iraq and Afghanistan added. "But until it became a reality, people really paid no attention."
A senior Centcom officer concurred. "We honestly did not believe that these guys were capable of doing this kind of stuff. . . . We underestimated them." But as EFP numbers began to inch up, intelligence reports indicated that CD-ROMs on how to construct an EFP were circulating among insurgent cells; one bombmaker reputedly held a doctorate in electrical engineering from Baghdad University. After a particularly sober briefing, Smith responded simply, "Holy cow."
By early 2005, what one officer had described as "an ominous thing on the horizon" was moving to the foreground in Iraq. Most EFPs were built with several pounds of pure copper, either milled or punched with a 20-ton hydraulic press into a concave disc with a 140-degree angle, two to 11 inches in diameter. Triggered by the infrared sensor, a blasting cap in turn set off explosives packed behind the copper disc -- known as a liner -- inside a steel or plastic pipe. The detonation wave, moving at 8,000 meters per second, struck the liner, which inverted into a tadpole-shaped slug.
An EFP eight inches in diameter threw a seven-pound copper slug at Mach 6, or 2,000 meters per second. (A .50-caliber bullet, among the most devastating projectiles on the battlefield, weighs less than two ounces and has a muzzle velocity of 900 meters per second.) Unlike an armor-killing shaped charge, the EFP warhead did not turn into a plasma jet, but remained semi-molten. Copper was preferred because it is ductile and malleable, and does not shatter like steel. Typically fired at ranges from five to 10 meters, the slug could punch through several inches of armor, spraying metal shards across the crew compartment.
Beginning in early June 2004, insurgents often placed EFPs together for a shotgun effect. Many EFPs were hidden in foam blocks, carefully shaped to look like roadside rocks or curbstones; often the only telltale sign was a half-dollar-sized hole bored through the foam for the passive infrared lens, which became known among U.S. troops as the "Eye of Allah."
Debate intensified within the U.S. government over Iran's role in distributing EFPs. Abizaid was skeptical until British troops reportedly captured a cache of copper discs along Iraq's southeastern border. Other evidence accumulated. For example, according to a former DIA analyst, the C-4 plastic explosive found in some EFPs chemically matched that sold by Tehran's Defense Industries Organization and identified by specific lot numbers. Intelligence also indicated the Quds Force of Iran's Islamic Revolutionary Guard Corps was training and giving explosives to certain Iraqi Shiite groups, a senior DOD official said.
Still, Abizaid advocated a measured response. Iran after all could have supplied anti-ship or anti-tank missiles that would be far deadlier to U.S. forces. EFPs were a tiny percentage of all IEDs -- more than an annoyance but less than a casus belli. "You know they're doing it, but you don't know that you want to go to war over it," he said. "The vast majority of problems in Iraq are generated in Iraq by the body politic."
Votel nonetheless felt pressure to find a remedy. Month by month, EFPs were better camouflaged and more effective. In February 2005, he sent a lieutenant colonel to Tel Aviv. The Israeli solution to IEDs often included using armored bulldozers to scrape away the top 18 inches of earth where bombs might be hidden, according to an Israeli engineer colonel. That tactic had limited utility in Iraqi cities, but Israel also had made technological strides.
Votel's emissary examined six promising counter-IED systems. Under an agreement with the defense ministry, for about $1 million Israeli engineers would bring four of them to Yuma Proving Ground in Arizona for further testing. System names would be changed to conceal the Israeli connection, so that a microwave gadget called VOW became Dragon Spike I, and MACE -- Microwave Against Concealed Electronics -- became Dragon Spike II. The testing was scheduled for June 2005.
By the end of May, IED attacks in Iraq exceeded 1,000 a month. President Bush declared Memorial Day "a day of prayer for permanent peace." Without mentioning Iraq or Afghanistan, he asked Americans to observe a "national moment of remembrance."
The vice president was more upbeat. Appearing as a Memorial Day guest on "Larry King Live," Cheney predicted, "I think they're in the last throes, if you will, of the insurgency."
Staff researcher Madonna Lebling contributed to this report.
Tuesday: A blast from below