When then-Maj. Gen. Franklin L. “Buster” Hagenbeck of the Army published critical remarks about the Air Force in fall 2002, the subject of those words—a battle called Operation Anaconda—became a major sore point between airmen and soldiers.
The public part of the controversy quickly faded into the background. Hagenbeck soon was promoted to a three-star post. Top Army and Air Force leaders worked hard to effect improvements in air support of ground forces, which paid tremendous benefits in Operation Iraqi Freedom in spring 2003.
Even so, Anaconda—a March 2002 battle against Taliban and al Qaeda fighters in Afghanistan’s soaring eastern mountain ridges—remained a source of interservice friction.
An official Air Force report, declassified and released in February, makes clear that the argument was not superficial. This two-week battle became the impetus for the services to revise everything from equipment and training for ground controllers to component command relationships at the highest levels.
The Air Force report, “Operation Anaconda: An Airpower Perspective,” is the source of most of the facts offered in this article. The rest is based on publicly available reports and published material.
Into Public View
The interservice quarrel flared into public view in September 2002. In that month’s issue, Field Artillery magazine, an Army journal, published a lengthy interview with Hagenbeck. He had served as commander of Coalition Joint Task Force Mountain in Afghanistan and was, as such, commander of Anaconda.
The Hagenbeck interview marked a low point in the Anaconda affair. In it, the Army officer endeavored to defend decisions he had made in the operation. However, he also took the opportunity to make ill-informed comments about Air Force operations.
Hagenbeck heaped praise on Navy and Marine Corps pilots, pointing out that they routinely flew low in support of the land forces. “They were terrific,” he said. When it came to the Air Force, though, Hagenbeck took a very different view. Among his complaints:
There was more. When asked about close air support (CAS) operations, Hagenbeck said: “The most effective close air support asset we had was the Apache, hands down.” The Army’s AH-64 Apache helicopter is the service’s principal combat aviation system.
However, the eight Apaches that he had sent on a sweep of the objective area at the start of Anaconda were hit so many times by enemy fire that only three came through in flyable condition. Hagenbeck had to send out an emergency call for help to Marine Corps AH-1 Cobra helicopters at sea on USS Bonhomme Richard. The Cobras made the flight up to Bagram Air Base, near Kabul, to help out. At the same time, Air Force C-17s began moving replacement Apaches from the US to Afghanistan.
It would be fair to say that Hagenbeck’s after-action interview showed a commander with a light grasp of the workings of the air component and even key aspects of the battle.
In Hagenbeck’s defense, it should be noted that he had only a few weeks to get to know the operation he was about to lead.
US Central Command had been tracking enemy activity in the Khowst-Gardez region since early January 2002. It was a collection point for al Qaeda and Taliban fighters fleeing south after the fall of Kabul. Special operations forces teams developed a plan to surround the Shah-i-Kot Valley with SOF teams and Afghan irregular forces and then bring in 1,400 Army soldiers by helicopter to take blocking points on eastern ridges above the valley.
This was an encirclement operation. In theory, it would bag all of the al Qaeda and Taliban remnants and be over in 72 hours.
Anaconda was to be the biggest use of conventional ground forces to that point in Afghanistan. For that reason, the original SOF planners, in a series of early February meetings, decided to ask Task Force Mountain, led by Hagenbeck, to take tactical control.
Hagenbeck and his boss, Lt. Gen. Paul T. Mikolashek, the combined forces land component commander, or CFLCC, were briefed on Feb. 17, 2002. By then, the execute date—Feb. 28—was 11 days away. (Bad weather pushed the execute date to March 2.)
Army Gen. Tommy R. Franks, head of CENTCOM and overall coalition military commander, knew there was an operation in the offing; in his memoirs, he said that he discussed it at a National Security Council meeting on Feb. 7. The final plan for Anaconda was briefed to him via video teleconference on Feb. 26.
“Love it,” he told his commanders.
However, the plan had three major flaws.
First, Hagenbeck’s final operations order, issued Feb. 20, underestimated the enemy’s numbers. Why remains unclear. CENTCOM estimated al Qaeda and Taliban forces in the Khowst-Gardez region at several hundred or even 1,000. These estimates went into detail on the organization of the al Qaeda and Taliban units and offered a spread range, giving the numbers greater weight. Hagenbeck’s order placed the number of enemy at 125 to 200 in the Shah-i-Kot Valley.
TF Mountain had for some reason narrowed its focus to counting enemy fighters in a small geographic area. In remarks to a Washington Post reporter one year later, Hagenbeck would admit, “We only probably had about 50 percent of the intelligence right.” Instead of a 7-to-1 US advantage in numbers, the US and enemy forces would be about equal in size. Among them were well-equipped Arab and Chechen al Qaeda with crew-served weapons, sniper rifles, and mortars. Moreover, Taliban and al Qaeda fighters held the best terrain.
Second, the plan had logistical weaknesses. “The operation itself was right on the edge of the logistical envelope,” claimed Rear Adm. James A. Robb, CENTCOM’s planning head at the time. All troops, equipment, and supplies were coming by air. The director of mobility forces, USAF Brig. Gen. Winfield W. Scott III, recalled that there was “essentially no land support” for operations launched out of Bagram.
The Army needed to move up to 1,000 soldiers from Kandahar to Bagram to stage the operation. They also needed to augment the fuel bladders at Bagram to provide the gas to keep helicopters flying. Yet the commander of TF Mountain was slow to generate even rudimentary airlift requirements and pass them along to the air component.
Just moving fuel was a major challenge. C-17s would “wet-wing” fuel by landing at Bagram, off-loading internal fuel into bladders, and then taking off to hit a tanker overhead. Then, they would repeat the process. Bagram’s short, crumbling runway was too treacherous for other aircraft.
“We gathered up every available flying resource that we could in that part of the world,” said then-Maj. Gen. John D.W. Corley, who directed the combined air operations center in Saudi Arabia. Those pressed into service included C-17s earmarked to support Vice President Dick Cheney’s visit to the region. “We moved 193,000 gallons of gas between the [February] 23rd and 28th, of which zero was moved by ground,” Scott said.
Third, no one had notified the air component commanders that a major operation was in the works.
Exactly when the air component learned of Anaconda is the subject of much argument. CAOC records, as detailed in the Air Force’s report, indicate there were working-level discussions of Anaconda by Feb. 21, when the battlefield coordination detachment (BCD) briefed CAOC intelligence staffers. Anaconda was mentioned in a Master Air Attack Plan briefing on Feb. 22. Over the next three days, CAOC staff also discussed the role of AC-130s, status of intelligence-surveillance-reconnaissance (ISR) collection, and potential for the availability of only a single carrier, USS John C. Stennis, after USS Theodore Roosevelt left the area and before USS John F. Kennedy arrived.
Corley said he was first briefed on Feb. 22—six days before the execute date—and became “a little pessimistic” when he learned none of his senior staff had heard of the operation, either. The BCD’s job was to sift through dozens of pending potential CFLCC operations and tag those that needed CAOC attention. Mikolashek also asked about air component coordination on Feb. 17. While there was working-level discussion, Anaconda was not elevated as it should have been by either the BCD, TF Mountain, or CFLCC headquarters.
Air Force Gen. T. Michael Moseley, the combined forces air component commander (the “air boss”), was not briefed on the plan until he returned from a trip on Feb. 25.
Not a Clue
Carrier battle group commanders—who would supply the bulk of the fighter sorties—were in the dark, too.
“We didn’t have a clue what they were going to do,” said then-Rear Adm. Mark P. Fitzgerald, in command of Roosevelt’s battle group. The Navy forces knew Anaconda was coming, he said, but added, “We had no clear idea of specifically where. We knew it was up in the Tora Bora area, but that was about all we knew. We didn’t know what their scheme of maneuver was. We didn’t know how many people they were putting in there.”
He went on to say that there was no definition of how it was going to work. The communcations plan was to talk to a forward air controller on a normal frequency. It was not an integrated operation by any means.
One F-15E crew member who flew Anaconda sorties observed afterward, “All this planning for a 1,500-man operation, and the Army couldn’t pick up the phone and make a call?”
Hagenbeck’s failure personally to ensure component coordination was glaring. With over 30 FACs in a 64-mile area, deconflicting strikes would be tricky. As it turned out, the CAOC could easily muster 60 on-call CAS sorties per day, but simply generating sorties was not the whole issue.
“The bigger issue,” Moseley later said in the official Air Force report, “is there was never an opportunity to orchestrate and figure out what was needed.”
As Moseley recalled the situation: “Had we known this was going to go on, we would have stood up a full ASOC [air support operations center] and moved [the right people] to Bagram a week or two weeks ahead of this and then conducted a set of rehearsals with carriers, with the bombers, with the whole thing. And I would have forward-deployed the A-10s … [for] indigenous quick reactions.”
The lack of advance high-level discussion prevented the coalition air component from imparting to Anaconda planners full knowledge of its capabilities. Moseley later said there was no dedicated ISR plan—a significant lack because infrared sweeps of the terrain might have revealed more information about enemy strength and positions.
There was also little chance to improve air support request management even after a few ASOC personnel headed to Bagram. Communications out of the mountain-ringed “Bagram bowl” were limited. Hagenbeck talked about how hard it was for him to talk to his subordinate commanders. The air coordinators with TF Mountain also had very limited communications.
When one looks back and examines the pell-mell Anaconda planning, one conclusion stands out. By assuming that the combat operation would face fewer than 200 enemy fighters, and that it would be over quickly, the forward land component at Task Force Mountain—led by Hagenbeck—thought it had matters in hand and the role of the air component would be small.
In hindsight, it is clear the planners were viewing the Shah-i-Kot Valley operation through rose-colored glasses.
At CENTCOM, No Worries
At CENTCOM headquarters, no one was on edge, either. In a vivid demonstration of the lack of air-land coordination, CENTCOM raised no objection when Stennis unknowingly opted out of action in the first day of Anaconda.
The carrier’s leaders wanted to help their crews switch back from a nighttime to daytime schedule. On March 2, they shut down the deck for a “steel beach picnic” of barbecue and games. In addition, sailors were treated to movies throughout the ship.
Then came word about Anaconda. “There was a lot of action going on, guys are getting shot at, and there was a lot of frustration on the carrier, ‘Well, why are we having a picnic today?’?” said Navy Lt. Cmdr. Todd Marzano, an F/A-18 pilot. “There was another carrier out there [Roosevelt]. There were also obviously ground-based bombers that could help. So it’s not like we’re the only show in town, but, when you’re up there, you want to help out and participate as much as you can,” he said. Instead, for Stennis pilots, the first day was a no-go.
Anaconda had no sooner begun than US forces started paying the price for poor component-level coordination. To begin with, there was minimal preparatory bombing of the objective area of caves and trails. Hagenbeck later said he did not want to run a preparatory bombing campaign “arbitrarily” without knowing what was in the al Qaeda cave complex. This was a clear product of underestimating the enemy situation and course of action.
Afghan forces starting the attack on March 2 came under inadvertent fire from an AC-130 and broke off their attack. Other air strikes were called off in the confusion.
Then came the unexpectedly costly Apache helicopter sweep. In the wake of that setback, soldiers hitting the landing zone fell under concentrated enemy attack right away. Mortar and small-arms fire was so intense that Hagenbeck ultimately asked bombers to level two Shah-i-Kot Valley villages. Poor communications links, the lack of a true ASOC at Bagram, and failure to schedule joint planning and rehearsals with the air component now created a tough situation for the ground controllers requesting support—and for the airmen trying to provide it.
It was typical on that day to have three or four “troops in contact” situations running simultaneously. Air support flexed to meet the demand. According to the Air Force report, one B-1 on the afternoon of March 2 spent more than two hours releasing 19 Joint Direct Attack Munitions (JDAMs) on 10 different targets in response to requests raised by multiple ground controllers. Another B-1 dropped 15 JDAMs on six different targets in six bomb runs. Yet another B-1 hit a ZSU-23 anti-aircraft gun as Taliban fighters tried to wheel it out of a cave.
A pair of Marine Corps F/A-18Cs embarked on Roosevelt were the first aircraft to strafe enemy firing positions. Scarface 73 and Scarface 74 made three passes, each delivering 400 rounds from 20 mm cannons just as darkness fell. That night, Air Force AC-130 gunships attacked with 40 mm and 105 mm guns and passed coordinates for targets to other strikers.
Day 1 of Anaconda saw the air component deliver 177 precision weapons (JDAM and GBU-12s), of which 162 were a result of immediate CAS. That averaged out to more than six precision bombs per hour, or one every 10 minutes. Actual drops ebbed and flowed with the ground situation, but continued day and night. The peak came in the afternoon, when bombers and fighters dropped 64 precision weapons between 1 p.m. and 6 p.m., local time.
At the time, at least, TF Mountain certainly seemed to appreciate these airpower operations. “Numerous bombing strikes were made against dug-in enemy forces, … resulting in moderate to heavy enemy casualties,” said TF Mountain’s nightly report on March 2. It also cited “near-continuous use of CAS assets,” mainly for troops-in-contact situations.
A battlefield airman on the ground that day described his calling for an attack on a threatening enemy mortar. “An aircraft was diverted to our position almost instantaneously,” said this staff sergeant. Six bombs knocked out the mortar. Then, he went on, “we did the exact same,” putting three bombs on another gun. Finally, they called for a B-52, which dropped six more bombs on another mortar position on the ridge.
Despite the high responsiveness of the air support, Moseley was not content with the battle’s first few days. He talked to the CFLCC, Mikolashek, dispatched key officers from the CAOC to Bagram, and brokered a deal to bring A-10 attack aircraft closer to the fight.
Statistics in the declassified Air Force report paint a picture of air operations far different from the image created in Hagenbeck’s after-action comments. Drawn from a database compiled by the deputy fire support coordinator—a man Hagenbeck called “my go-to guy”—the catalog of dropped bombs shows constant air strikes, day and night, with a mix of precision and nonprecision munitions.
Immediate CAS supplied consistent support. Total numbers of weapons dropped did not tail off at night. In fact, on some days, nighttime totals were higher than those in daytime.
Within several days, the CAOC also worked out with the land component an arrangement for striking “preplanned” targets found in areas chosen by CFLCC planners. Having preplanned targets available improved the volume of bombs dropped. A total of 751 bombs were dropped into the Anaconda battle area in the first three days.
Tales From the Battle
Accounts of individual missions illustrate the dangers, frustrations, and successes, from the airman’s perspective.
The Air Force report recounts the work of Predator unmanned aerial vehicles, which tracked a large body of al Qaeda reinforcements moving down a road. Forward air controllers vectored several two-ships of A-10s and sections of F/A-18s onto the target. A coalition SOF controller surveyed the grisly scene the next day and found the entire al Qaeda group had been wiped out.
Bombers were mainstays of the operation because of their long times on station and heavy payloads. They, too, experienced the push and pull of air control. They could release individual weapons as necessary, and most made multiple target passes.
Navy pilots provided the majority of fighter sorties each day. “There was a lot of chaos down there those first few days of Anaconda,” said Marzano, the F/A-18 pilot. “The initial plan that they had constructed for the grid system overhead the target area for organizing the flow of aircraft in and out was somewhat disorganized, and it was hard to work the target area and deconflict with other aircraft out there.”
Within days of the start of Anaconda, the increase in preplanned targets and improvements in air control brought the air component to peak efficiency. On March 10, Marzano and his wingman launched their aircraft, with a mix of two GBU-12s on each aircraft and two Mk 82s fused for airbursts. Those Mk 82s were ideal for air-bursting the al Qaeda into the next life, as Moseley put it.
Nightly TF Mountain reports to the land component commander were widely cited in the official Air Force report. They gave a clear sense of the pace of the battle. Smart tactical decisions by soldiers made an enormous difference. So did air support.
Nevertheless, Hagenbeck, when interviewed by Field Artillery, dismissed all such reckonings. “A ground force commander does not care about the number of sorties being flown or the number and types of bombs being dropped and their tonnage,” he opined. “Those statistics mean nothing to ground forces in combat. All that matters is whether or not the munitions are time-on-target and provide the right effects.”
Examples from the battle showed Hagenbeck was not well-attuned to the actual role played by air support.
In the battle of Takur Gar (sometimes called “Roberts Ridge” after Petty Officer 1st Class Neil C. Roberts, a Navy SEAL who died there), seven Americans lost their lives. Nearly 40 Americans—Rangers and SOF—spent the day pinned down under al Qaeda fire after a helicopter insertion went badly wrong. Two rescue helicopters were shot down near the ridge. With them was a ground forward air controller known as Slick 01. He later said he controlled about 30 CAS strikes that day, which kept al Qaeda forces pinned down. These strikes ranged from the F-15E’s first-ever strafing mission to assists from F/A-18 pilots to bombers dropping precision weapons on enemy clusters. Persistent airpower kept the teams covered until the 38 survivors could be extracted around 8 p.m. local time.
Technically, the action at Takur Gar involved troops not under Hagenbeck’s command. A week later, though, the situation was different, and, once again, the contribution of airpower was underrated. Hagenbeck mentioned a period “several days into the fight,” when “the ceilings dropped [and] we had limited air coverage.” This was the bad weather period when the land forces had to delay their final assault to take their main objective, called Objective Ginger. Still, TF Mountain organized more air strikes for a final March 9-10 push that would seal off the Shah-i-Kot Valley. The days of March 9 and 10 were also the single heaviest days for air strikes, with more than 300 bombs dropped during each day.
Good Came From Bad
Harrowing though it was, Anaconda ultimately was a success. The combined efforts of land and air forces succeeded in clearing the enemy from the valley and surrounding hills.
As the Air Force report notes, a rather large amount of future good came from Anaconda. Its problems and mishaps spurred a drive to improve component coordination and fix procedures wherever possible.
Though the report does not say so explicitly, it is nevertheless true that this drive was pushed largely by Air Force officers. Moseley remained as air component commander for Iraq and took the opportunity to forge much closer relations with the new CFLCC, Army Lt. Gen. David D. McKiernan.
Air Force and Army three- and four-stars twice discussed Anaconda in private sessions. Gen. John P. Jumper, Air Force Chief of Staff, made it a priority to review and standardize ground controller training. Air commanders such as Moseley and his deputy, then-Rear Adm. David C. Nichols Jr., vouched for the change in component relationships that took place between Anaconda in March 2002 and Operation Iraqi Freedom in March 2003.
CAS systems in Baghdad in 2003 or in Najaf and Fallujah in 2004 were masterpieces of precision mapping, deconfliction, and component coordination, from working level to three-star level. Those operations made Anaconda seem like ancient history.
In Afghanistan, new procedures led Lt. Gen. David W. Barno, the Army’s commander there, to offer this praise in late 2004: “Airpower from all the services—intelligence, surveillance and intelligence assets, mobility aircraft, close air support, and space systems—have given ground forces in Afghanistan the ability to operate in smaller units and respond quicker with more accurate weaponry than at any other point in history.”
What lingered, unfortunately, was the ill will engendered by Hagenbeck’s intemperate utterances.
Field Artillery later published an article acknowledging that “ground controllers identified targets, and, more often than not, attack aircraft hit those targets.” Yet even this article said Anaconda’s problems stemmed from a general “lack of adherence to or even understanding of joint doctrine.”
The corporate Army was largely silent on the matter.
What the Air Force report makes only too evident is this fact: Responsibility for Anaconda’s poor planning rests not with lack of fealty to joint doctrine or the inherent complexities of CAS or the new territory of the war on terror. Responsibility rests with the commander. Better decisions would have produced a better result.
Rebecca Grant was a participant in the USAF Anaconda study. She is president of IRIS Independent Research in Washington, D.C., and has worked for Rand, the Secretary of the Air Force, and the Chief of Staff of the Air Force. Grant is a fellow of the Eaker Institute for Aerospace Concepts, the public policy and research arm of the Air Force Association’s Aerospace Education Foundation. Her most recent article, “Expeditionary Fighter,” appeared in the March issue.