Flying 33,000 feet over Iraq, the two US Navy pilots saw a pair of Patriot missiles streaking up at them, trailing orange flames.
The pilots turned sharply to evade but it wasn't enough. ''They're tracking m. . ." called out the pilot of one of the planes, Lieutenant Nathan White of Abilene, Texas, just before he was cut off. The pilot of the other jet saw his wingman's plane disintegrate in ''a pie shaped outline of sparks."
Helicopters found White's body 10 days later.
Just why White was shot down by a Patriot unit near Karbala on April 2, 2003, is at the center of a friendly fire investigation that recently wrapped up -- and which may once again bring scrutiny to the Army antimissile system largely made by Raytheon Co. of Waltham.
In a summary of the friendly fire investigation report released Dec. 10 by US Central Command, one finding was that Patriot units mistook White's plane for a hostile missile, apparently one that didn't exist. It also states that Patriot's proper launch procedures were violated.
But a redacted version of the command's friendly-fire report provided to The Boston Globe paints a much fuller picture of the Army's difficulties using Patriot during the invasion of Iraq, including gaps in crew training and the frequent appearance of what defense specialists call ''false tracks," or symbols of potential targets that were actually thin air. The issues show the unintended dangers that computerized weapons systems can pose, and the need for better human oversight.
All radar system displays can show false or ''spurious" tracks, which can be caused by sandstorms, equipment problems or electro-magnetic interference from nearby radars. How officers position Patriot can also affect performance.
The Army and Raytheon have been working to address the problem of false tracks since the Gulf War of 1991, when they caused 24 of the 158 Patriot missiles fired during the conflict to be aimed at empty sky. Those engagements were one reason the Army backed off its early claims about Patriot's performance.
Despite the improvements, Patriot crews faced ''numerous spurious" missile tracks during the invasion of Iraq in 2003 as well, the redacted friendly-fire report states. Exactly why so many false tracks appeared isn't made clear. Also, soldiers weren't trained to expect the false tracks, and ultimately didn't react correctly when White's plane was misidentified, apparently after it was mixed up with a false track.
During the invasion Raytheon tried to prevent the false tracks with new software, with little success, according to one soldier's statement.
A spokesman for Raytheon said the company agrees with the report's conclusions but said executives wouldn't comment beyond a statement that ''We are confident that the Patriot system performed as designed."
The report also states that data recorders that were supposed to show Patriot's exact performance often didn't operate. The lack of data hampered the friendly-fire investigation and leaves some uncertainty to its findings, according to the report.
Neither the Army nor Central Command public summaries of the investigation mention the frequency of false tracks, training issues, or the data problems. A memo accompanying the report dated Dec. 2, 2004 from John P. Abizaid, the general who leads Central Command, accepts that ''systems and training deficiencies with the Patriot were the primary cause of the incident." Degraded communications also contributed, he wrote.
A Navy commander, T. J. Keating, added that military doctrine did not ''prepare Patriot to be fully integrated into a complex joint battle space."
Army officials in charge of Patriot units at Fort Bliss, Texas, referred questions about the report to officials at Central Command at MacDill Air Force Base in Florida, which oversaw the friendly-fire investigation. A spokesman there said officials wouldn't comment on the report, at least until they have finished considering a request for the report under the Freedom of Information Act.
In addition to White's death, another Patriot unit in the same battalion shot down a British Tornado jet earlier in the war, killing both fliers. The plane was also misidentified as a missile. According to Abizaid, both incidents show the need for better tactics and procedures to prevent Patriot and systems like it from automatically engaging friendly targets.
In a Dec. 10 press release the Army said it is developing new Patriot hardware and software based on lessons from the 2003 campaign. Elsewhere, it has described training improvements.
Despite the friendly-fire incidents, the Army says Patriot performed well in the conflict, shooting down all nine Iraqi ballistic missiles engaged. The Army declined to say exactly how many Patriot missiles were fired or if any missed.
Patriot surface-to-air missiles were originally meant to shoot down planes. They were configured to shoot down missiles during the Gulf War. On the ground they are organized in ''batteries," including a command vehicle where soldiers control the system, mobile radar sets, and missile launchers mounted on trailers. In all fifty-four Patriot batteries were deployed for the invasion of Iraq.
Victoria Samson, a research analyst at the Center for Defense Information, a Washington think tank, who has been skeptical of some missile-defense work, said the false tracks might have arisen as Patriot units moved north through Iraq to defend advancing Army formations. They weren't designed for such rapid movement, she said, which might have caused their radar fields to overlap, creating electro-magnetic interference.
Theodore Postol, a physicist at the Massachusetts Institute of Technology who has been a chief critic of Patriot in the past, called it ''scandalous" that the problems weren't addressed prior to the invasion, or publicly described afterward. He doubts much will change in the future.
''If the summary report doesn't acknowledge these problems, then why should we expect the problems to get solved?" Postol said.
The friendly-fire report was provided to the Globe by the pilot's father, Dennis White of Abilene, Texas. He received it from one of its authors, Army Colonel Robert L. Jassey, who manages Patriot units at Fort Bliss.
The Army would not permit Jassey to be interviewed. But his report faults the soldier acting as the Patriot's tactical director for setting the system to fire without doing enough checks to confirm the target.
For instance, the soldier should have called up more information on the computer screen that would have shown the target was flying like a plane and not a missile. Even after Patriot's missiles were launched, the soldier still had at least 25 seconds to correctly identify White and order the missiles to self-destruct.
The tactical director and other Patriot crew members also ''had a limited understanding" of how ballistic missiles appear on radar, the report states. Investigators found ''systemic gaps" in training throughout the Patriot battalion involved in the two aircraft shoot downs, the report states. Soldiers hadn't mastered the system's launch modes, classification methods, or how to operate with poor communications.
At bottom, crews must be more skeptical, the report states. ''The root cause of the majority of these training shortfalls is a blind faith in 'the system' . . . it is based on a false assumption that the Patriot system is infallible."
The report also says that Patriot crews often saw false missile tracks on their radar displays, symbols of hostile missiles that, in hindsight, were ghosts. One soldier in the unit that shot down White's plane saw seven false tracks in four hours, according to a sworn statement.
''I have never seen anything like this until I got here into the Middle East" another Patriot crew member said.
According to a statement from a Patriot operator two days after White was shot down, ''Since the beginning of my deployment we have experienced [many] 'spurious' tracks. My unit was one of the first to find this out. Raytheon then developed patches for the software that was supposed to fix the problem. To my knowledge they did not."
Dennis White, himself a former Air Force pilot, says he's frustrated by the lack of specific discussion in the report about the cause of the false tracks and just how his son's plane was misidentified. The redacted report puts too much blame on individual soldiers, he said. He hopes Congress will investigate Patriot's weaknesses and whether the Army and Raytheon took steps to bring them to the attention of other armed services.
''It does not take a rocket scientist to make the conclusion that something was very wrong with the Patriot radar," White wrote in a letter he recently posted to a website honoring his son.
The site also contains an e-mail the 30-year-old sent from the aircraft carrier USS Kitty Hawk.
Describing his combat missions, Nathan White wrote: ''get catapulted from standstill to 140 miles an hour in less than 2 seconds, navigate through a maze of airborne highways that try to deconflict aircraft and of course steer you clear of the army's patriot batteries . . . Make it night time and throw in some thunderstorms and then it really gets exciting."
The friendly-fire report picks it up from there on the night of the accident. After bombing a military target near Baghdad, the two single-seat planes turned south toward Karbala.
Without recorded data, what exactly happened next is murky. According to a briefing that Jassey gave to Dennis White, up to four Patriot units were tracking White's F/A-18 ''as an unidentified aircraft." One of them misclassified the plane as a ballistic missile ''due to some type of electromagnetic anomaly." An accompanying diagram shows overlapping Patriot radar fields, suggesting they may have interfered with each other.
The bad information was relayed to another battery in the same battalion, where the symbol for the unknown aircraft changed to the symbol for a ballistic missile, for reasons the report doesn't specify. Then, the briefing slide states, ''The operators began to react to the perceived threat."
Ross Kerber can be reached at kerber@globe.com.![]()