Problems Encountered

All in all, CENTCOM committed remarkably few major errors during its campaign to topple the Ba’athist regime of Saddam Hussein. To be sure, as a subsequent review of postcampaign detainee operations in Iraq led by two former secretaries of defense, James Schlesinger and Harold Brown, harshly concluded, the Bush administration’s overall strategy wrongly “presupposed that relatively benign stability and security operations would precede a handover to Iraqi authorities.” The administration also failed utterly to “plan for a major insurgency” and “to quickly and adequately adapt to the insurgency that followed after major combat operations.”¹ To infer from that stern judgment, as Washington Post military reporter Thomas Ricks later did, that the initial invasion and takedown of the Ba’athist regime was based on “perhaps the worst war plan in American history,” however, was a considerable over-reach.² Such a blanket dismissal of CENTCOM’s chosen approach for defeating Hussein’s forces was tantamount to throwing out the baby with the bathwater. A more discerning perspective was that offered by former secretary of state Colin Powell, who candidly remarked that what really went wrong with Iraqi Freedom was not the three-week major combat phase of the campaign, which was “brilliantly fought,” but rather the transition afterward to “nation-building.” In his well-informed opinion, there had been “enough troops for war but not enough for peace, for establishing order.”³

Regardless of the wisdom of the Bush administration’s decision to invade Iraq in the first place, there seems to be no question that the invasion was badly underresourced for the many needs of postwar stabilization and transition to democratic rule, about which senior administration leaders had been more than amply warned before the invasion by a multitude of credible advisers both in and out of government. This book, however, focuses on the strategic and operational use of air power in pursuit of CENTCOM’s more immediate combat objectives during the three-week regime-removal phase of the campaign. By that more narrow measure, it is fair to conclude that those who conducted the campaign at all levels turned in a performance that was both exemplary and rich with useful lessons for the next major conflict the United States may face. Vice Adm. Joseph Dyer, the commander of Naval Air Systems Command, observed afterward in that respect, “Almost all of our thinking is proactive and looking toward future developments and amazingly little of it in terms of correcting deficiencies that were in evidence in this conflict.”4 CENTCOM’s deputy air component commander, Rear Admiral Nichols, made the same point even more directly in stating that instead of thinking in terms of revealed problems in need of fixing, U.S. and allied airmen should be reflecting on “what we validated.”⁵ During the course of a two-week Iraqi Freedom “lessons learned” symposium conducted by CENTAF at Nellis AFB in July 2003, numerous air component representatives in attendance who had been personally involved in the planning and conduct of the campaign at all levels suggested that unlike previous air wars from Desert Storm through Enduring Freedom, this one celebrated air power’s successes and would offer relatively few suggested equipment and procedural improvements for future campaigns. The collective sense of the participants was that few findings from the symposium would challenge prevailing service trends that had been established and validated in previous wars.⁶

Nevertheless, although the campaign to topple Hussein set new records for air power achievement and broader joint force effectiveness, its execution was not without certain areas of combat performance that, in the words of JFCOM’s commander at the time, Admiral Giambastiani, “fell short of expectations.”⁷ General Moseley, also in attendance at the after-action review symposium, denoted seven areas of activity that, in his view, “worked less well” during the campaign: (1) insufficiently timely BDA, especially with respect to strategic targets and KI/CAS; (2) inadequate information transfer between the CAOC and shooter aircraft; (3) inadequate battlespace deconfliction; (4) totally unsatisfactory Army Patriot missile deconfliction and firing logic; (5) inadequate prioritization of theaterwide information operations; (6) a relatively poor performance by the ASOC supporting V Corps in comparison with the ASOC’s more professionally manned and more efficient Marine Corps DASC counterpart; (7) and speed, range, maximum service ceiling, and defensive system inadequacies with respect to CSAR platforms.⁸ The most notable air-related problems identified during the review were a succession of friendly fire incidents, continuing friction points in the air-ground interface, insufficiently timely provision of battle damage assessment, inefficiencies in the management of in-flight refueling support, and inadequate arrangements for the timely sharing of sensitive information with the nation’s coalition partners.⁹

Incidents of Fratricide

The coalition’s combat fratricide rate was lower during the three-week campaign to topple Hussein than it was during Operation Desert Storm, when 35 of the 148 U.S. fatalities (24 percent) and an even higher proportion of the British casualties incurred resulted from friendly fire. All the same, a postcampaign report by the Center for Army Lessons Learned identified 17 friendly-fire incidents during the major combat phase of Iraqi Freedom.10

The three episodes of Blue-on-Blue surface-to-air fire that occurred involved the Army’s Patriot SAM system and allied combat aircraft. In the first, a Patriot PAC-2 missile struck an RAF Tornado GR4 on March 23 as the aircraft was returning to its base in Kuwait, instantly killing the crew, Flight Lieutenant Kevin Barry Main and Flight Lieutenant David Rhys Williams. The next day, a U.S. Air Force F-16 pilot whose aircraft had been locked up by a Patriot acquisition and tracking radar about thirty miles south of An Najaf fired an AGM-88 HARM at the offending radar in presumed self-defense, destroying the radar but remarkably causing no friendly injuries or loss of life. And on April 2, near Karbala, a Patriot PAC-3 missile struck a U.S. Navy F/A-18C operating from Kitty Hawk and killed the pilot, Lt. Nathan White.¹¹ Neither of the two downed allied fighters in these Blue-on-Blue incidents appeared on the Patriot’s threat display as an aircraft.

Investigators who reviewed the details of the three incidents considered the possibility that the Patriot system’s tracking radars had generated false targets and the Patriot operators had believed they were engaging enemy missiles rather than aircraft. Because aircraft and missiles have such different radar profiles, the investigators further considered the possibility of electromagnetic interference caused by the close proximity of the offending Patriot batteries to such other friendly radio frequency emitters as ground-based artillery radars, airborne surveillance sensors, and electronic jammers. In this regard the chief of Army air defenses, Brig. General Howard B. Bromberg, commented, “This is the densest battlefield we’ve seen. I believe there could be something there.” The Patriot batteries were clustered together looking northward toward Iraq from Kuwait, with PAC-2 and PAC-3 radars reflecting back on one another. General Bromberg added: “You have three incidents like that—they’ve all got to have some interrelationship. I’m convinced they did.”¹²

Other early speculation pointed to a possible failure by the downed Tornado crew to transmit the proper IFF code, as well as to a possible use of improper settings by the offending Patriot crew to identify tracked targets accurately by correlating their speed, altitude, trajectory, and other flight characteristics. A board of inquiry convened by CENTCOM ultimately exonerated the Patriot battery crew who downed the GR4 during the latter’s approach to landing at Ali Al Salem Air Base in Kuwait, reportedly concluding that the crew “mistook the aircraft for an antiradiation missile based on its high-speed descent and lack of a functioning IFF.”13 The British MoD refused at first to comment on that finding, although RAF sources insisted that IFF is a go/no-go checklist item. Informed sources initially deemed it unlikely that the aircraft’s IFF could have failed in flight without the pilot noticing the failure and promptly alerting ground controllers. AWACS operators monitoring the airspace over southern Iraq likewise should have noticed immediately had the aircraft’s IFF suite not been functioning.

A subsequent investigation revealed that although the Tornado crew believed that they were properly squawking in all assigned IFF modes, the aircraft’s IFF system had indeed failed completely. The investigation further determined that the Patriot crew had been undertrained and should not have assumed that the Tornado was hostile simply because it was not squawking IFF Mode 4—or, for that matter, any other transponder mode.¹⁴ An early assessment by a long-established specialist in Patriot performance at the Massachusetts Institute of Technology, based on careful examination of a 32nd AAMDC briefing released shortly after the campaign ended, determined that a large number of Patriot radars were operating in close proximity to one another in the immediate war zone; that those multiple, independently operating radars were often in line of sight and tracking the same aircraft; that their target contacts could be displayed as spurious ballistic missile tracks; that the Patriot unit would automatically engage the false target; and that Patriot operators could have intervened to stop the engagement, but only if they acted within a time window of less than a minute. He further reported that Patriot operators “were not trained to deal with this scenario, and this scenario was not incorporated in engagement training software embedded in deployed Patriot units. . . . Thus the combination of lack of timely information from other air defense surveillance assets, timelines of tens of seconds or less to fire on the believed target, and no software support or training to recognize and deal with such situations put the Patriot crews in an impossible situation.”¹⁵

In early 2004 a television reporter who had been embedded with a Patriot battery during the campaign uncovered evidence that friendly aircraft had been showing up on Patriot radars as incoming enemy missiles and that this problem had first surfaced with a forward-deployed Patriot unit a month before the campaign’s start. The reporter discovered that the Army had not alerted other Patriot operators or coalition aircrews to the problem. The deployed Patriot batteries had been operating in “weapons free” mode at the time of the two fratricide incidents, meaning that the launch crews could fire the missiles without first seeking higher approval. The account also reported that the battery that downed the Tornado GR4 was showing the incoming fighter as an enemy cruise missile, and that the one that subsequently downed the Navy F/A-18 had displayed its target “as a highly credible enemy ballistic missile symbol that did not resemble any false missile track that had been seen before.”16

Whatever the explanation, allied aircrews soon developed a deep distrust of the Patriot system and made every effort to remain a healthy distance from its threat envelopes in their mission planning. One F-16 flight leader recalled what that effort entailed for his four-ship flight launching with no notice against an emerging target complex in the heart of Baghdad on March 30: “We . . . got airborne heading straight for the targets. Normally, we would have taken a westerly course in order to stay behind the northwest/southeast running line of our ground advance. The main reason for this was to stay behind the two Patriot batteries that had already been involved in two friendly fire incidents. To put it bluntly, the Patriots scared the hell out of us.”¹⁷

The assessed threat to friendly aircraft presented by the Patriot SAM system had at least two undesirable consequences. First, it obliged allied aircrews to squawk secondary and unencrypted civilian Mode 2 IFF codes with their transponders, which could be tracked by Patriot but also by some Iraqi SAM systems. Second, allied aircrews experienced a notable spike in their stress levels whenever they operated within a Patriot threat envelope. More than a few who were locked up by a Patriot radar had been forced to drop chaff, engage in aggressive countermaneuvering, and make frantic radio calls to the AWACS to get the offending Patriot to break its lock.¹⁸

Starting in August 2003 and continuing into June of the following year, a Defense Science Board (DSB) task force on Patriot system performance examined the Patriot’s operating experience during the campaign looking for lessons that might be incorporated into the continued development of Patriot and its planned follow-on system, the medium extended air defense system (MEADS). One of the key issue areas the DSB explored was the blend of causal factors that most likely accounted for the two Patriot-related fratricide incidents. The DSB’s findings, issued in January 2005, noted that the Patriot’s role in the campaign had been defense against enemy tactical ballistic missiles, with a secondary self-defense role against enemy antiradiation missiles, and that the system had no assigned air defense role. The board adjudged the Patriot’s service in its primary combat role a “substantial success.” The system had engaged all nine Iraqi TBMs that were fired against coalition forces; eight of those intercepts were confirmed kills, and the ninth engagement was deemed a probable kill. Given the reality of the Iraqi TBM threat and these manifest accomplishments against it, there was no question that the substantial deployment of Patriots, entailing up to forty U.S. Army fire units and twenty-two more from four coalition countries, was an essential ingredient of CENTCOM’s overall force mix.19

With respect to the two surface-to-air fratricide incidents, the DSB cited “a complex chain of events and failures,” noting that there was “insufficient data to pin down the exact causes of failure” that occasioned the two inadvertent downings of coalition aircraft. Among the known shortfalls in CENTCOM’s capabilities and operating modes that might have been contributing factors, however, were consistently poor performance of the combat identification capability embodied in the Mode 4 IFF systems carried on allied aircraft; an absence of adequate situational awareness in CENTCOM’s combined air defense system; and the Patriot system’s “operating philosophy, protocols, displays, and software, which seemed to be a poor match to the conditions of [Operation Iraqi Freedom].” The report added that the Patriot’s operating protocol “was largely automatic, and the operators were trained to trust the system’s software,” which was designed to accommodate the possibility of “heavy missile attacks.” The task force further noted that with the “enormous” number of coalition aircraft sorties flown during the campaign (some 41,000) and the large deployment of Patriot fire units in-theater (62 in all), “the possible Patriot friendly aircraft observations were in the millions, and even very low-probability failures could result in regrettable fratricide incidents.”²⁰

With respect to minimizing the likelihood of future surface-to-air fratricide incidents, the DSB’s report concluded that the Department of Defense needed to “find and fix the Mode 4 IFF problem” and “improve the situational awareness of [the U.S. defense establishment’s] air defense systems.” As for the Patriot system in particular, the DSB further spotlighted a need “to shift its operation and control philosophy to deal with the complex environments of today’s and future conflicts,” especially conflicts that “will likely be more stressing than [Iraqi Freedom] and [will] involve Patriot in simultaneous missile and air defense engagements.” Key to such a system improvement, the DSB concluded, will be “a protocol that allows more operator oversight and control of major system actions.”²¹

Despite such efforts to get to the bottom of the fratricide incidents and correct their causes, however, General Moseley concluded more than three years after the campaign ended that the Patriot problem “is still not fixed.”22 For its part, the Army would say for public consumption only that “application of lessons learned . . . has already improved upon Patriot’s performance and the system will be continuously refined. . . . Some changes include the integration of satellite radio technology at the battalion information coordination center which provides improved situational awareness through voice and data connectivity with higher headquarters identification and engagement authority, as well as enhanced command and control and software improvements that enable better identification, classification, and correlation of airborne objects.”²³

The Army’s Patriot operators were not the only ones culpable in inadvertent friendly fire occurrences. Air-to-ground fratricide was also significant during the three weeks of major combat. An especially notable incident took place on March 23 when a two-ship element of Air Force A-10s was targeted against a group of what turned out to be friendlies by a Marine Corps ground FAC and subsequently fired on those troops near An Nasiriyah, disabling tanks, APCs, and Humvees and killing an undetermined number of Marines. In what the official history of I MEF’s contribution to the campaign later characterized as “arguably . . . the most notorious friendly fire incident of the war,” that A-10 formation (call sign Gyrate 73) responded to a call for immediate CAS from the commander of Bravo Company of the 2nd Marine Regiment’s 1st Battalion, which had been assigned the mission of securing two bridges on Highway 8 over the Euphrates River (the southern bridge) and Saddam Canal (the northern bridge) in An Nasiriyah. The account noted that the A-10 was “usually a welcome sight on the battlefield and had already done good work on March 23 against other targets. But now it was bearing down on friendlies.”²⁴

A postmortem report on the incident issued by a CENTAF investigating board two months later noted that “witness statements and testimony indicate that the majority of [friendly] casualties were most likely caused by friendly fire.” The report went on to say, however: “Considering information made available upon reconvening, this is no longer the board’s opinion. . . . Of those 18 Marines [killed during the encounter], it is the opinion of the board that enemy fire killed 8 Marines. Due to the mixture of intense enemy fire, combined with friendly fire from Gyrate 73 flight, the board is unable to determine, by clear and convincing evidence, which type of fire killed the remaining 10 Marines.” In a cover memorandum forwarding the report to General Franks, General Moseley noted that “the investigating board concluded that the primary cause of the incident was a lack of coordination regarding the location of friendly forces due to a number of contributing factors.”²⁵

A subsequent CENTCOM assessment of the incident issued on March 6, 2004, reported that the battalion’s Charlie Company “began taking heavy enemy fire from artillery, rocket-propelled grenades (RPGs), mortar and small-arms fire. At approximately the same time, the air officer, located with the forward command post, called the Bravo Company forward air controller (FAC) . . . requesting CAS to combat enemy forces attacking their location. . . . The A-10s targeted what turned out to be Charlie Company assets, making multiple passes against them. Eventually, the A-10s were told to cease fire, which they did.”26

The CENTCOM assessment reconstructed the incident as follows:

           The Bravo Company commander, collocated with the FAC, directed the FAC to engage the targets north of the canal. The A-10s spotted a burning vehicle (thought to be an enemy vehicle, but turned out to be a damaged Charlie Company vehicle) north of the bridge and reported it to the FAC, who could see the smoke and verified that it was in the target area. . . . The FAC was not able to see the A-10s or a specific target. Therefore, he confirmed the target location with the A-10s and attempted to verify the location of the lead element with the Bravo Company commander. . . . Based on the information he possessed concerning the [unit’s] scheme of maneuver [and] believing that only enemy forces were ahead, [the Bravo Company commander] cleared the target for fire. No additional authorization was sought. The FAC informed the A-10s that there were no friendly forces north of the bridge and they were cleared to engage.²⁷

In his cover memorandum forwarding the assessment, CENTCOM’s commander at the time, Gen. John Abizaid, recommended “a reexamination of joint doctrine as it relates to Type III CAS control.” He also recommended that the Marine Corps consider “appropriate administrative or disciplinary action against the Bravo Company FAC” who called in the A-10s.²⁸ With respect to the latter recommendation, a press account that appeared several weeks after the incident reported that the A-10 pilots had been repeatedly cleared to engage what were thought to be paramilitary Fedayeen Saddam fighters by a “disoriented” Marine Corps ground FAC who was located behind the Marine company’s position.²⁹ Essentially bearing out that early report, a Marine Corps publication issued in September 2009 confirmed that CENTCOM’s incident investigation “logically concluded that the cause of the incident was the [FAC’s] violation of a standing order not to use Type III [the least stringent form of CAS] without approval from higher headquarters. If he had contacted the battalion commander, he would have known that friendly forces were north of the Saddam Canal. Even if he failed to make contact with the battalion commander but still adhered to the standing order, the incident would not have occurred.”30 (The incident, it bears noting, erupted after a group of Iraqi troops had begun firing on the Marines after having first pretended to surrender.)³¹

On March 27 an A-10 mistakenly strafed another group of Marines, a regimental combat team that had become caught up in a firefight and had requested immediate fire support. The A-10 pilot was given an incorrect grid zone designator that had the Marine combat team right in the middle of it. Thus improperly cued, the A-10 pilot misidentified the friendly unit as hostile and was cleared to engage it with his 30-mm cannon. This time, remarkably, no one was hurt.³²

A week after the fratricidal Tornado GR4 downing, an A-10 attacked two British Army Scimitar armored reconnaissance vehicles, killing one British soldier. In another fratricide incident involving British forces, an A-10 on March 28, 2003, mistook reconnaissance elements from D Squadron of the British Army’s Household Cavalry for Iraqi forces and in two consecutive engagements with them left one British trooper dead and three wounded.³³ In a similar incident, an American SOF team aiding Kurdish forces called in two fighters to attack an Iraqi tank that was firing at them from a mile away. Instead, one of the fighters mistakenly dropped a bomb on the Kurdish convoy, killing eighteen Kurds and injuring three American soldiers.

Finally, three U.S. Army soldiers were killed and six were wounded near Baghdad International Airport on April 2 when an F-15E inadvertently struck their position in a quick-reaction attack against what its crew thought was an active Iraqi SAM site. The 3rd ID’s after-action assessment assigned responsibility to an uncoordinated and improper attempt by the CAOC to engage targets inside the FSCL: “Much to their credit, [V] Corps was able to stop a total of 14 of these attempts. However, an F-15E under CFACC control was successful in one attempt. The F-15E misidentified an MLRS [multiple-launch rocket system] as a [SAM] launcher approximately 15 miles from his position. The pilot found 14 vehicles in that area and asked permission from [AWACS] to engage those targets. . . . The end state was fratricide—3 killed in action, 6 wounded in action, and 3 vehicles destroyed. This is unacceptable.”³⁴

The Army report cited combat identification errors; fatal navigation errors; a loss of fire control; errors in reporting, battle tracking, and clearance to fire; ineffective maneuver control; and either weapons errors or failures in troop discipline as the primary causes of the fratricide incidents. The report concluded that the likelihood of future air-to-ground fratricide incidents could best be reduced through such means as standardized tactical air controller training and equipment across all U.S. forces. The use of standardized geographic designation systems, cross-checking battlespace information, and ensuring that all friendly aircraft remain within assigned boundaries through proper preparation and detailed intelligence was also crucial. Proper maintenance of situation awareness and understanding at the lowest level along with the marking and positive identification of targets, both day and night, was also essential. As for ways of minimizing the likelihood of surface-to-air incidents, the report emphasized proper IFF procedures, the development and standardization of battlefield identification systems, and positive identification of targets.35

In his assessment of the Tornado episode and other instances of fratricide over the three-week course of the campaign, Air Marshal Torpy conceded that occasional fratricide in warfare is simply “one of the facts of life. It is our job to make sure that those tragic incidents are reduced to the absolute minimum.”³⁶ Although the GR4’s downing temporarily strained the trust relationship between the British and American contingents, General Franks later echoed the ultimate British sentiment on the matter when he commented: “When there are friendly fire incidents across coalition boundaries, it brings allies closer together.”³⁷ After the incident, the British contingent made appropriate adjustments to the rules of engagement and the manner in which the U.S. Army’s Patriot was employed.

The Failed Apache Deep-Attack Attempt

Although air-land integration within CENTCOM had significantly improved prior to Iraqi Freedom, the long-standing discontinuity between Air Force and Army cultures with respect to how best to draw down and neutralize an enemy’s ground forces nonetheless resurfaced during the campaign’s initial days. Air Force proponents in the CAOC argued in favor of using fixed-wing air power to degrade enemy force capability to the greatest extent possible before allied ground units moved to direct contact, while their Army counterparts insisted that early close ground force engagement with the enemy was the ideal mode of operations. Such thinking clearly underlay the Army’s abortive attempt to conduct an independent deep attack against a concentration of Republican Guard forces by a formation of AH-64 Apaches without prior preparation of the battlespace by fixed-wing air power.

In a move that appears to have been completely uncoordinated with CENTCOM’s air component, the V Corps commander, Lieutenant General Wallace, approved a staff request to launch a deep attack mission that would send AH-64 Apache attack helicopters from the 11th Attack Helicopter Regiment (AHR) to engage three brigades and the organic artillery of the Republican Guard’s Medina Division deployed northeast of Karbala and south of Baghdad. The Apache community had been itching to get a piece of the action from the campaign’s very start. A postcampaign review of Army operations noted that “some pilots had compared this attack to the 101st Aviation Brigade’s legendary deep attack operation in Operation Desert Storm; they too were going to be heroes. Their frustration continued to build, adding to the 11th [AHR’s] collective desire to get into the fight.”38

The attack was plagued by a host of miscalculations from the very outset of mission planning. First, because the corridor near An Najaf through which the Apache crews would penetrate to their assigned targets was lightly populated, planners elected to forgo the normal precursor suppressive fire, principally by artillery. Second, the regiment lacked adequate situation awareness of the target area because the shamal had grounded the UAVs that would have provided the Apache crews with refined target coordinates. Because of spotty and incomplete intelligence reporting on the disposition of Iraqi forces in the area where the attack was to be concentrated, the regiment’s pilots lacked specific grid locations for their targets and had only rough approximations, within a kilometer or so, of Iraqi company positions. Essentially, they would have to grope about in a calculated way to find the enemy forces.³⁹

In addition, the speed of the 3rd ID’s northward advance during the preceding two days made it necessary to move the Apache attack up twenty-four hours. Although nine potential ingress routes had been initially considered and proposed by the 11th AHR, only three southern routes were ultimately approved by V Corps due to concern at headquarters that use of the western approach options would encroach on airspace that had been allocated to the 101st Airborne Division. To make matters worse, friendly convoys bearing fuel and ammunition had been delayed by traffic jams, and inadequate fuel supplies at the planned point of departure necessitated dropping one of the three squadrons of Apaches from the attack plan. (Sixty AH-64s had been flown into a staging area at Objective Rams to marshal for the planned attack.)⁴⁰

The lack of close coordination with the air component resulted in several problems. First, no prior dedicated fixed-wing defense suppression was requested or provided. On top of that, allied aircraft overhead in CAS stacks and supporting friendly artillery units were not notified of a two-hour delay in the planned launch of the attack, and thirty-one ATACMS missiles were fired on a preplanned schedule into the intended target area before the Apaches had even left their marshalling area at Objective Rams, a miscue that warned the enemy of an impending attack. By the time the Apaches reached the target area, allied fixed-wing strike aircraft that had been holding in CAS stacks overhead had departed as a result of low fuel states, leaving the Apache crews without available CAS in case it should be needed.41

In the end, at 0115 local time on March 24, only thirty of the initial planned contingent of sixty Apaches got safely airborne, one of the AH-64s having crashed during takeoff after its crew became disoriented by swirling dust. Once aloft and in their prebriefed ingress formation, the thirty Apaches proceeded along the fifty-mile route to the designated target area, flying one hundred feet above the ground and maintaining a nose-to-tail separation of fifteen rotor lengths. Few of the Apaches ever got close enough for their crews to engage the enemy. As their final attack run was under way, the power grid in the An Najaf area went black for a few seconds, most likely as a signal to Iraqi gunners that the Apaches were en route to their objective. When the lead element of the Apache formation neared the target area, the defending Iraqi forces opened fire. Aware that the Apaches would have to ascend to two hundred feet to clear the power lines, Iraqi gunners aimed their streams of fire above the power lines, and the helicopters flew into a fusillade of enemy fire. One Apache was downed and its two-man crew captured.⁴²

Enemy gunfire damaged all but one of the Apaches. On average, each sustained sixteen to twenty bullet holes. Sixteen helicopters suffered damage to main rotor blades, six to tail blades, six to engines, and five to engine driveshafts. During their hasty withdrawal from the target area, two of the Apaches barely avoided a midair collision. For their efforts, the 11th AHR successfully attacked a dozen Iraqi vehicles. A month would go by before the regiment was fully ready for combat again. The land component staff subsequently came to believe that the Apaches’ assembly areas in the Iraqi desert had been under surveillance. General Wallace also subsequently told reporters that an Iraqi major general in An Najaf had used a cellular telephone to warn the Iraqi defenders that a wave of AH-64s was heading toward their position near Karbala.⁴³ Wallace freely granted that the attempted operation “did not meet the objectives that [he] had set for that attack” and concluded that “deep operations with Apaches, unless there’s a very, very, very clear need to do it, are probably not a good idea.”⁴⁴

In the immediate wake of the failed attack, the 101st Airborne Division’s attack aviation units shifted their mission focus from nighttime deep attack operations to armed reconnaissance, and the planning for a second Apache foray against Karbala on March 28 clearly incorporated lessons learned from the many mistakes made during the lead-up to the first.45 For a time, a later report noted, V Corps leadership “debated whether to attempt the mission at all,” but eventually they gave the green light.⁴⁶ Precursor defense suppression attacks and on-call CAS from allied fixed-wing air power were better coordinated for this attack, and the Apache crews now had their flanks protected by F/A-18s and other fixed-wing aircraft. This time, the 101st Aviation Brigade “relied heavily on its air liaison officer, who in turn requested an airborne forward air controller . . . on the mission to ensure [that] CAS would be coordinated directly between fixed- and rotary-wing aircraft.”⁴⁷ Also, the raid was preceded by four minutes of preparatory artillery fire to pin down any Iraqi gunners who might have been lying in wait for the attackers, and the Apache crews expressly avoided any built-up areas from which concealed man-portable infrared SAMs and small arms fire might emanate.

Moreover, the Apaches were led this time by Kiowa Warrior scout helicopters that were sent in to validate targets first. The Apaches destroyed seven Iraqi AAA positions, three artillery positions, five radars, and twenty-five vehicles, with no Apache losses, but this second performance still left much to be desired in terms of the ultimate payoff achieved for the effort and risk that went into it. After that, V Corps abandoned any further pursuit of deep attack air operations, and its use of Apaches was generally limited to providing organic armed reconnaissance and close-in CAS for Army ground units.

As for the good-news part of this story, the unsuccessful deep attack attempt on March 24 was the only major Army setback during the entire three-week campaign. And it resulted in significantly improved air-ground coordination. After that failed effort, the Army’s vice chief of staff, Gen. John Keane, pointedly asked: “Does our doctrine still make sense?” General Keane admitted that the Apache formation “ran into an organization that was much more spread out” than had been expected. As a result, he said, “we are taking a look at aviation doctrine and how to use Apaches at long distances.” Subsequently, the land component routinely requested Air Force A-10s to suppress enemy ground fire before planned Apache operations. “In other words,” said General Keane, “we had air power with them as well.”⁴⁸ General Wallace later reported that once this arrangement was in place, “the U.S. Air Force had a heyday against those repositioning Iraqi forces.”⁴⁹

The Army’s postcampaign assessment candidly admitted that the Apache experience “will affect how the Army trains and equips units for years to come.”⁵⁰ The assessment noted that the incident “placed in question the efficacy and utility of attack helicopters in Army doctrine” and attributed its shortcomings to the centrality of “the human ego in war” and “the indomitable warrior spirit to get into the fight.”51 In a remarkably unflinching display of frank institutional introspection, the assessment concluded that “the Army will need to consider under what conditions flying attack helicopters deep will produce the kind of benefits that warrant the potential risk.”⁵²

Most Air Force airmen would readily agree with this judgment. Without question, the flawed and ultimately abortive deep-attack attempt by the 11th AHR showed yet again the limited ability of even the best attack helicopters to cover extended areas of terrain and to conduct precision attacks in heavily defended airspace when compared with far more versatile fixed-wing fighter aircraft. At more than $30 million a copy in today’s dollars, an Army AH-64 costs roughly the same as an Air Force F-16, yet it lacks the F-16’s range, persistence, striking power, and survivability. Mindful of that comparative performance limitation, former Air Force chief of staff Gen. Merrill McPeak suggested in a postcampaign comment that the Army should limit its use of the Apache to CAS or, “if it must go deep, hand it over [to the air component] for joint tasking.”⁵³

Inefficiencies in Controlling Joint Battlespace by V Corps

Doctrinally imposed limitations on the joint delivery of fire support, especially with respect to the placement and use of the FSCL, were a continuing bone of contention between the air and land components that many airmen felt needlessly inhibited the most effective application of joint fires in support of V Corps. David Johnson called differences regarding the management of battlespace “the single greatest issue between the Army and the Air Force” and put “Army deep attack concepts and the placement of the FSCL . . . at the heart of the matter.”⁵⁴ Michael Knights expressed the same line of thought in his apt observation that “battlespace is a jealously hoarded commodity in modern warfare.”⁵⁵

The FSCL was the primary fire support mechanism for dividing CENTCOM’s battlespace between the land and air components. Any enemy terrain on the far side of the FSCL was essentially a free-fire zone for the air component because there could be no possibility of friendly troops coming into contact with enemy ground forces in that portion of the battlespace. All kill boxes on that side of the FSCL were open to attacks from the air. The terrain on the near side of the FSCL was the land component’s battlespace. Kill boxes that lay within that terrain were closed to attack from the air unless the land component commander expressly opened them for a finite window of time to permit air attacks under the control of properly trained and certified FAC-As or ground-based JTACs who were tasked by the ASOC supporting V Corps.

During the initial allied ground advance into Iraq, General McKiernan, with General Franks’ concurrence, extended the FSCL to eighty-four miles ahead of the line of advancing coalition ground forces. (The one exception was the transitory opening of kill boxes by V Corps during the “operational pause” prompted by the three-day shamal.) The extended FSCL put an additional strain on the tankers supporting CENTAF’s strike fighters because the latter had to fly farther north in order to provide effective interdiction and CAS. General McKiernan similarly moved the FSCL forward dozens of miles in front of coalition forces to facilitate the Apache assault planned for March 24. That decision, observed General Leaf, “cost us [the air component] . . . a full night of fixed-target strikes inside the FSCL. We—the entire coalition team—had not hit our stride in achieving the command and control required to operate in volume effectively inside the fire support coordination line.”56

This problem could have been easily anticipated and headed off had V Corps’ most senior planners harked back to very similar situations that had arisen during Operation Desert Storm. In his presentation to the Commission on Roles and Missions in September 1994, then Air Force chief of staff General McPeak noted Army commanders’ decision, on the third day of the ground offensive, to fix the FSCL well beyond their ability to affect the close battle with their own organic artillery and attack helicopters. In doing so, the XVIII Airborne Corps commander prevented the air component from interdicting the main resupply line connecting Baghdad and Kuwait for seventeen hours. General McPeak mentioned a similar extension of the FSCL inside the Kuwaiti theater of operations by the VII Corps commander that had essentially created a sanctuary for Republican Guard units, whose commanders took advantage of the opportunity to escape to Basra. Although the overall joint force commander, Gen. H. Norman Schwarzkopf, had expressly directed the air component commander, Lt. Gen. Charles Horner, to engage and destroy those forces, Schwarzkopf’s ground commanders, according to McPeak, “unilaterally placed boundaries that effectively contradicted [Schwarzkopf’s] theater priorities.” That, said McPeak, showed clearly “what can happen when boundaries are set by people who do not have full authority over military operations on both sides of the seam.”⁵⁷ An Air Force pilot who served as an ALO with the 101st Airborne Division during Desert Storm pointed out that a big part of the problem was an outdated conception of “close air support.” As this pilot bluntly put it, “what we really have is either air power applied in close proximity to troops or air power applied not in close proximity to troops. Definitions and lines on maps that don’t allow the flexibility required by nonlinear battle plans should be scrapped.”⁵⁸

In the case of air component KI/CAS planning on behalf of V Corps in Iraqi Freedom, the coordination in ATO preparation was conducted through the Army’s BCD in the CAOC. V Corps positioned FSCLs farther ahead of the forward line of friendly troops than usual because of an anticipated rapid rate of advance. An assistant to General Leaf recalled on this point: “Every day, General Leaf would arrange for the FSCL to be pulled a little back, but every night the Army majors would throw it far out again.”59 The land component would then request ground support sorties from the CAOC through the normal process. Within the V Corps area of operations, General McKiernan opened and closed kill boxes for interdiction attacks as he deemed appropriate, and the Air Force ASOC attached to V Corps provided terminal attack control for CAS strikes.

There was on occasion a pronounced disconnect between the CAOC and the Air Force’s battlefield airmen who staffed the ASOC supporting V Corps with respect to the proper roles and responsibilities of an ASOC, with each entity harboring a different understanding of how land combat operations between a division’s line of advance and the FSCL should be conducted.⁶⁰ Indeed, as one CAOC planner later recalled, “the single greatest issue that arose during Iraqi Freedom concerned the prosecution of counterland targets in the V Corps area.”⁶¹ Much of the difficulty alluded to here emerged from contrasting perspectives within the air component as to how best to employ fixed-wing air power in support of counterland operations. The CAOC planners who had developed CENTAF’s KI/CAS concept of operations insisted that unless friendly forces were present in a way that absolutely required detailed air-land integration, the kill boxes in that particular area should be opened for unrestricted air operations as targets of opportunity presented themselves. For its part, the ASOC sought instead to extract the most from the air component assets that were available to it by means of a procedure called “corps CAS.”

V Corps used the term “corps CAS” to refer to its intended use of CAOC-supplied air power to prosecute or shape targets in its assigned area of operations. Yet from the CAOC’s perspective, certain aspects of that concept needlessly hampered the shaping effort. For example, V Corps and its supporting ASOC considered the aerial engagement of any target on their side of the FSCL to be CAS, which, by definition, requires rigorous and time-consuming communications and command and control procedures. Yet in the view of those in the CAOC, those complicated CAS procedures were not required because the “corps shaping” in question was not conducted in even remote proximity to friendly ground troops.

Indeed, from the CAOC’s perspective, “corps CAS” was not CAS at all. It was de facto interdiction, which, according to accepted joint doctrine, lies beyond the roles and responsibilities of an ASOC. CAOC planners recalled that “the V Corps area of operations began at the division’s forward boundary, approximately fifteen nautical miles ahead of the forward line of own troops (FLOT), and extended to the FSCL. . . . Unfortunately, the ASOC was unable to conduct corps CAS at a rate commensurate with the number of air assets that were allocated to it.”62 As one might expect, friction ensued when the ASOC requested air support and then used it to conduct what were essentially battlefield interdiction operations. The ASOC also encountered problems in exercising the required degree of command and control throughout its assigned battlespace because the rapid forward movement of coalition ground troops typically exceeded the ASOC’s ability to communicate with CAS aircraft. As a result, there were often times when a large percentage of the aircraft sent by the CAOC to an airborne CAS stack in V Corps’ area of operations at the request of the ASOC were not used and had to return to their bases with unexpended ordnance. This inefficient force management exasperated aircrews. One F-16 pilot complained:

           We’d take off without a target, fly for eight hours trying to get someone on the ground to give us something to drop on, and then we’d come home with our bombs on board. It was frustrating for guys who wanted to feel like they’d contributed to the mission. [So a friend of mine who was a CAOC planner] sent us an e-mail saying, “Okay, I hear you guys. We’re going to start Operation Home Depot, and we’re going to go out in the desert and build you some targets so you can drop your bombs!” That made us laugh, but what he was really saying to us was that we’d still done our jobs, even if we came home with all our bombs.⁶³

To be sure, such recurrent “no-drop” incidents were by no means invariably the fault of the ASOC’s close control of open kill boxes in its battlespace at the insistence of V Corps. Some instances involving aircraft that had been holding overhead in CAS stacks to support V Corps rightly suggested a basic force utilization deficiency within the system.⁶⁴ Yet in other cases, weapons bring-back did not necessarily indicate a misutilization problem so much as the natural inefficiency that is part and parcel of CAS operations. Factors that affected the weapons bring-back rate ranged from an absence of assigned targets to communications problems, a lack of available tankers, and a less than ideal allocation of CAS assets by the ASOC.

The CAOC did indeed use bring-back rate as one measure of the efficiency of air support to the land component. Without more specific facts that might account for the number of prospective targets that went unserviced, however, weapons bring-back numbers did not, in and of themselves, provide a reliable measure of mission effectiveness. Simple communications deficiencies occasioned an appreciable number of combat aircraft returning to their bases without their pilots’ having even established radio contact with the ASOC. Added to that, the initial shortage of available tankers and the considerable distances that orbiting strike aircraft had to travel to return to their bases sometimes meant that those aircraft could not deliver their munitions within the scant on-station time they were allowed.65

But the main problem lay in the fact that the common geographic reference system based on kill boxes and keypads had never been formally ratified in joint doctrine. Accordingly, Army operators did not naturally accept that system in their own tactical contingency planning. Although the kill-box approach had been amply validated as an effective alternative to the FSCL in Desert Storm and had since been refined in a succession of combat-related SPINs, the FSCL for its Army benefactors continued to be a more familiar and comfortable artifact signifying Army ownership and control of joint battlespace.

The senior CAOC director later reported that the Army’s principals in the land component had agreed before the start of Iraqi Freedom to accede to the kill-box method but had reverted to the classic FSCL approach once the fight was under way.⁶⁶ CENTAF had offered training on kill-box operations to V Corps prior to the campaign, but the latter “chose instead to concentrate on other phases of their spin-up training. As a result, when the war started, their staff was not familiar with kill-box operations or the KI/CAS concept of operations, having never practiced it. So they fell back on more traditional methods that were less appropriate for the rapidly advancing ground situation.”⁶⁷

Viewed in hindsight, CENTAF was itself largely to blame for the situation. The commander of the USAF Air Ground Operations School at Nellis AFB at the time the Army’s spin-up training for the looming campaign was in full swing recalled that “the big lesson I learned—and have continued to learn since 2003—is that the ‘good idea cutoff point’ for a major Army event is the mission rehearsal exercise (MRX). The final [CENTAF] KI/CAS planning occurred while V Corps was doing its MRX in Poland. Like it or not, the Air Force missed the opportunity to get kill-box operations integrated into the Army’s planning for the event. When we went in after the fact and attempted to change peoples’ minds or add a new way to do business, we failed.”⁶⁸

The Air Force ASOC that was supporting V Corps was organized, trained, and equipped to prosecute CAS missions with an FSCL positioned only some eighteen to twenty-two miles forward of the FLOT. Extending the FSCL outward to eighty-four miles allowed General McKiernan and his subordinate commanders to control a larger portion of the battlespace in front of their advancing units, an entirely appropriate fires management measure for a linear engagement against massed enemy ground forces in open terrain. Yet Operation Iraqi Freedom was envisaged and planned to include nonlinear engagements in or near congested urban areas.69 By pushing the FSCL some sixty miles farther out than normal, V Corps’ commander extended the battlespace beyond his force’s ability to deliver usable combat effects. As one airman pithily put it, “Although the Army has some organic assets that can reach out to targets that far away from the FLOT (i.e., ATACMS), for the most part, most organic fires for the Army can only affect targets at a range of 32 kilometers, or about 20 miles. From 20 miles to the 84-mile FSCL is essentially a sanctuary for the enemy.”⁷⁰

Indeed, some Air Force airmen viewed V Corps’ unusual extension of the FSCL as directly contravening the guidance promulgated in Joint Publication 3–0, Doctrine for Joint Operations, which stipulates that “placement of the FSCL should strike a balance so as not to unduly inhibit operational tempo while maximizing the effectiveness of organic and joint force interdiction assets.”⁷¹ In a similar spirit, Joint Publication 3–09, Doctrine for Joint Fire Support, stipulates that “the decision on where to place or even whether to use an FSCL requires careful consideration. If used, its location is based on estimates of the situation and [concepts of operations]. Location of enemy forces, anticipated rates of movement, the concept and tempo of the operation, organic weapon capabilities, and other factors are all considered by the commander. . . . Placing the FSCL at greater depths will typically require support from higher organic [headquarters] and other supporting commanders.”⁷² For its part, Air Force Doctrine Document 2–1.3, Counterland Operations, specifies that “the optimum placement of the FSCL varies with specific battlefield circumstances, but typically it should be placed where the preponderance of effects on the battlefield shifts from the ground component to the air component. In this way, the FSCL placement maximizes the overall effectiveness of the joint force, and each component will suffer only a small reduction in efficiency.”⁷³ During at least portions of the three-week campaign, V Corps clearly failed to honor these accepted stipulations of joint doctrine.

In fairness to General Wallace and his preferred approach to managing the battlespace on his side of the FSCL, however, a subsequent assessment of issues in command and control in air warfare rightly noted that the air component’s inability to track the effects of its operations figured prominently in justifying the V Corps commander’s determination to maintain close control over deep operations in his immediate area of regard. According to this account,

           Army officers who worked for the BCD in the CAOC . . . pointed out that the inability of the air component to determine and communicate the effects of air power was the biggest source of friction between the air and land components. As the ground troops made their way through the sandstorms, they needed to know how big an effect the [air] attacks were having on the Iraqi Republican Guard units. When the storms were finished, even General McKiernan, the CFLCC [combined force land component commander, pronounced “see-flick”], was unable to pinpoint weaknesses in the enemy toward which he could have directed offensive actions to fracture them. . . . The air component’s practice of satisfying many of the land component’s air support requests (ASRs) with kill-box interdiction did not give land commanders visibility into the results. . . . These were missions sent to a patch of airspace, not to a target, so the land component was unable to tell whether its requests were being serviced by the air component.74

The assessment concluded on this point that “the components had good relationships at the top, and they had worked out a joint strategy. Nonetheless, they had different and somewhat incompatible local systems for developing and tracking target data.”⁷⁵ In explaining this disconnect, the assessment noted that Wallace viewed corps shaping as more effective than KI: “His data showed that enemy strength did not decrease appreciably after KI, but did after his Corps shaping.”⁷⁶ In a postcampaign interview that further helped explain his preference for corps shaping over kill-box interdiction, Wallace recalled that every advance by V Corps troops at the platoon through brigade levels during the three-week land offensive ultimately became a “movement to contact” because his command had been consistently unable to acquire a clear picture of enemy force dispositions.⁷⁷

More telling yet, Wallace observed that from his vantage point, the corps-shaping sorties that had serviced targets beyond the 3rd ID’s forward boundary “were 270 percent more effective [in terms of targets destroyed per sortie] than kill box interdiction in V Corps’ [area of operations.]” He freely conceded that with respect to the contentious issue of the FSCL versus KI as the preferred battlespace management tool, “both control measures are commonly used,” the use of both entails “a continuous debate across the services,” and “doctrinal definitions and [the] use of both are not consistent across the services.”78

After acknowledging the need for the Army and the Air Force to reconcile these unsettled issues, Wallace addressed complaints that V Corps had “pushed the FSCL too deep during [the campaign] and made attack short of the FSCL difficult.” He countered by noting that it was CENTCOM (with inputs from the CFLCC), not V Corps or its ASOC, that determined FSCL placement; that the FSCL was usually no more than thirty nautical miles from the FLOT during the three-week campaign; that the FSCL was more than fifty nautical miles deep for only eight hours (less than 2 percent) during the campaign, and was more than twenty nautical miles deep for only three days (less than 15 percent of the campaign).⁷⁹

Wallace’s recollections provide an important backdrop against which to examine the practical import of the assessed inefficiencies in joint battlespace management. Clearly, the CAOC, on the one hand, and the CAOC’s subordinate ASOC that was channeling air support to V Corps, on the other, had contrasting views with respect to the relative effectiveness of corps shaping versus kill-box interdiction. In this regard, the lead author of the most thorough account of Army operations during the three-week campaign was on solid ground in noting that “the CAOC versus ASOC differences [reflected] an internecine argument within the Air Force and not between the Air Force and Army.”⁸⁰ From the perspective of V Corps, however, the ASOC clearly met the command’s tactical- and operational-level needs throughout the land offensive. In all, 2,117 CAS missions were assigned to the ASOC, with 625 of those missions retasked to other agencies. The ASOC itself thus controlled 1,492 missions, 886 of which were so-called corps-shaping missions. Of all the corps-shaping missions, 525 (or 62 percent) ultimately dropped ordnance. In the case of division CAS, 307 of 606 missions (51 percent) dropped ordnance against valid targets. As one authoritative assessment of this performance later reported, “combining those two mission areas [corps shaping and division CAS], 55 percent of the missions executed by the ASOC and its subordinate TACPs dropped munitions on valid targets. That percentage is by far the highest ever seen in major combat.”⁸¹

General Wallace carefully avoided taking sides in the intra–Air Force and cross-service disagreements over the relative merits of ASOC-controlled corps shaping versus CAOC-controlled KI/CAS. Instead, he offered an equitable solution by concluding that achieving desired effects against enemy force dispositions as expeditiously as possible is what matters most in the air-ground interface. Toward that end, he added, “the FSCL, if retained, must be modified to facilitate rapid joint attack short of the FSCL” and “kill boxes, if used, must be more precise” because “joint effects in complex terrain demands it.”82 Of greatest importance by far with respect to the proclaimed inefficiencies in air apportionment discussed above is that the V Corps commander himself, the ultimate Army consumer of the air component’s contribution to the land battle, was, by all indications, more than content with the CAS that he and his troops received throughout the three weeks of major fighting.⁸³

A Marine Corps Approach That Arguably Worked Better

In marked contrast to the issues that divided the CAOC and ASOC with respect to battlespace management in V Corps’ area of operations, General Moseley so trusted the MAGTF approach to air-land combat, and the Marines who were implementing it, that he delegated authority to I MEF to control the airspace above its immediate area of operations. As a result, I MEF created its own direct support ATO for execution by the 3rd Marine Aircraft Wing, and the Marine tactical air operations center and tactical air control center (TAOC and TACC respectively) opened and closed kill boxes for deep interdiction operations. For close interdiction inside the FSCL, the Marine direct air support center (DASC) provided terminal attack control.

Key to the greater efficiency of I MEF’s unique arrangement for integrating CAS assets into the land battle was its willingness to open kill boxes inside the FSCL when its battle staff were confident that no friendly forces were present, something V Corps would generally not do. Opening kill boxes inside the FSCL allowed coalition aircraft to attack Iraqi forces without detailed coordination with the DASC and the resulting bottleneck that tended to create. The principal differences between the practices of the DASC and the ASOC supporting V Corps were that the ASOC was aligned at the corps level, was responsible for integrating air power for a multidivision fight, and accepted decentralized execution of allocated air assets, whereas the DASC was aligned at the division level, was responsible for integrating all organic Marine Corps firepower, and envisaged centralized combined arms execution in support of the MEF commander.

The DASC, assigned the radio call sign Blacklist, was configured much like the TACC but on a smaller scale. It coordinated with the fire support coordination center that was organic to the 1st Marine Division. Its operations staff, the “Dascateers,” also handled emerging target requests that came in from outside normal Marine Corps channels. When Marine air operations ranged more deeply into Iraq beyond the normal radio range of the ground-based DASC, it was augmented by an airborne DASC (the DASC-A), with the assigned call sign Sky Chief, on board a Marine Corps KC-130 to provide better communications connectivity across the entire span of the battlespace.84

By almost all accounts, the DASC supporting I MEF was able to integrate more fires into the Marines’ battlespace for CAS and interdiction than was the ASOC assigned to V Corps. Particularly during the initial days of combined air-land operations, the ASOC was able to integrate an average of six combat sorties into the fight per hour, whereas the DASC was able to integrate twice that amount of air support in a smaller area in the same amount of time. An informed Air Force airman later noted that the well-trained DASC controllers’ ability to make effective use of excess air component strike sorties in their zone “got to the point where Air Force KI/CAS pilots all wanted to go to the Marine sector where they knew they’d have a better chance of putting bombs on a meaningful target.”⁸⁵ (Eventually, as joint air-ground operations hit their stride, the ASOC too became able to integrate CAS sorties more efficiently in closer keeping with I MEF’s well-honed approach. Once that occurred, General Leaf noted, FSCL placement became less an issue as the air and land components succeeded in improving the coordination of their operations within kill boxes.)⁸⁶

Even so, I MEF’s fire support coordination measures in combination with its command and control arrangement generally allowed CENTAF aircraft supplying on-call CAS to provide more responsive fire support than did V Corps, whose less permissive arrangement created inefficiencies. In V Corps’ area of operations, the ASOC was frequently overburdened by the land component’s requirement for terminal attack control between the forward line of friendly troops and the FSCL for any open kill boxes. As a result, the ASOC was forced to turn away sorties that could have been effectively used to shape the battlefield. I MEF avoided this problem by dividing the area between the forward line of friendly troops and the FSCL with a battlefield coordination line (BCL) and by using two agencies—the DASC for terminal attack control between the forward line of friendly troops and the TAOC for tactical control between the BCL and FSCL.

As the campaign unfolded, short-range aircraft such as the Marine Corps’ AV-8B Harriers and AH-1W Cobra attack helicopters and the Air Force’s A-10s were generally used in the CAS role, while the longer-range Marine Corps F/A-18s supported I MEF kill-box interdiction requirements and the air component’s deeper–strike mission needs. As the FSCL moved ever closer to Baghdad, the Marine Corps used a combination of forward basing, organic refueling, and nonorganic refueling from Air Force and RAF tankers to support Marine air operations.⁸⁷ With respect to the efficiency of this arrangement, Air Chief Marshal Burridge testified before the British Parliament: “The U.S. Marine Corps are configured as relatively light forces, and they do not have indigenous deep fires, that is, a lot of artillery. They have very little artillery. Their equivalent of artillery is the Marine Air Wing F/A-18s. They live together very intimately, and their ability to do close air support, both the ground forces’ ability to control it and the air’s ability to integrate with it, is very impressive, very impressive indeed.”88

A RAND assessment of persistent air-ground integration issues aptly attested to the contrast in operational styles between the two services in noting that “the Army and Marine Corps commanders largely fought independent campaigns [in Iraqi Freedom], with air power employed as each of these components deemed appropriate.”⁸⁹ A well-informed survey of command issues ranging from Operation Desert Storm to the major combat phase of Operation Iraqi Freedom summed up the stylistic contrast as follows:

           The Army handled all missions short of the FSCL as CAS and required them to be controlled by the ASOC or a TACP designated by the ASOC. The Marines, on the other hand, chose to create another line . . . [the BCL] . . . which was closer to the ground troops than the FSCL. The Marine DASC opened up kill boxes further out than this BCL—so as a result, air support going into the Marines’ sector had a greater chance of being sent to an open kill box than that going to the Army sector. To the aircrew, the difference in flexibility was so stark that pilots regularly requested to be sent to work with the DASC rather than with the ASOC. . . . The difference [was] in the degree of coupling the two services saw in these operations. To the Marines, beyond the BCL, the efforts of the ground and air forces were not tightly coupled and did not need to be closely managed. . . . To the Army, the efforts of the ground and air forces were tightly coupled all the way out to the FSCL. . . . Aircrews were more closely managed, so [V Corps commander General] Wallace had more visibility into the targeting, but the aircrews related it took longer to perform a mission in the Army’s sector than in the Marines’ sector.⁹⁰

The 3rd ID’s after-action assessment of the campaign took note of the superior efficiency offered by the Marine Corps’ BCL in managing fixed-wing air assets in air-land warfare and expressly acknowledged that it “facilitates the expeditious attack of surface targets of opportunity between [the BCL and the FSCL].” Calling the BCL “clearly the most permissive measure, but also one that requires a thorough understanding of our doctrine and its use,” the assessment concluded that “this measure would give Corps the battlespace they desire for air assets at their disposal” and recommended outright that “U.S. Army doctrine must be changed to incorporate the BCL as [a fire support coordination measure].”91

When the land component’s northward movement all but ground to a halt during the three-day shamal and General McKiernan pulled in the FSCL to just beyond the Euphrates River, the Marine Corps’ approach to managing air and ground fire support in common battlespace seemed validated. The shortened FSCL opened up numerous previously forbidden kill boxes for the air component to work. A true air-ground joint concept of operations emerged at that point, producing, as Michael Knights commented, “something akin to the arrangement that had been in place in the MEF sector throughout the war. The MAGTF concept underpinned the difference, allowing U.S. Marines (and their subordinate British division) to fight as a true air-land partnership rather than as a ground and air component trying to get out of each other’s way.”⁹²

The Continuing Need for Closer Joint Force Integration

Among the cross-service problem areas spotlighted by the Iraqi Freedom experience were the acknowledged deficiencies in the Air Force’s approach to organizing and manning its ASOCs. The Air Force staffs its ASOCs with officers from assorted operational flying backgrounds and with enlisted personnel from similarly diverse service career fields rather than assigning personnel specifically trained for ASOC operations. This means that there are very few expert Air Force ASOC operators, a situation that naturally perpetuates a general absence of ASOC proficiency and continuity. Further, the service had until recently given its ASOCs less than top billing in its rank ordering of attention and priorities. A senior staffer in the ASOC attached to V Corps during the campaign recalled with brutal frankness in this regard: “The CAOC is a significant piece of the CFACC’s planning, command, and control capability, but it is not the entire structure. The CFACC’s [ASOC] and subordinate [TACPs] are another significant piece” of that capability. “Unfortunately, prior to [Operation Iraqi Freedom], these [latter] organizations did not enjoy the same USAF focus and resources . . . [and] the USAF had no published [concept] for ASOC operations and [for the ASOC’s] subordinate TACPs assigned to the supported land maneuver forces.” To make matters worse from a training perspective, until the very start of the campaign, “the ASOC in the 4th ASOG had never exercised with the CAOC. . . . The first time the [V Corps] ASOC actually controlled a strike against a target in the ground commander’s [area of operations and] the first time the CAS cell at the CAOC had ever directly worked with ASOC operations was the first day of combat operations.” Identified shortfalls in the ASOC at that time “included radio limitations, a lack of a target coordinate generation system, and a lack of interconnectivity with Army targeting systems.”93

In marked contrast, the Marine Corps has long followed a substantially different approach in its MAGTF organization, which, from top to bottom, is a well-trained and appropriately manned command and control system for closely integrated air-land combat operations. Each MAGTF, wherever it may be deployed worldwide, brings together a TAOC, a DASC, a DASC-A, and a DASC-forward into a coherent arrangement for conducting air-to-ground mission sets in support of a ground combat element commander. Furthermore, MAGTFs are routinely exercised in realistic peacetime training evolutions, and their capabilities and limitations are well determined and understood before the time arrives for them to have to visit fire and steel on enemy targets. CENTAF planners concluded after the major combat phase of Iraqi Freedom that the Marine Corps’ approach to command and control in counterland operations would be well worth considering for future contingencies that might entail an ASOC as a subsidiary component of the CAOC.⁹⁴

For their part, Army assessors eventually came to realize the opportunity cost of their traditional practice in situations in which the land component lacked the needed situational awareness and weapons reach to conduct deep attacks inside the FSCL with its own organic assets. The 3rd ID’s after-action report freely admitted that “the U.S. Army must redefine the battlespace based on our ability to influence it.”⁹⁵ The report went on to note that

           the FSCL was 100 kilometers beyond the range of standard munitions from our M109A6s and M270s. This created a dead space between the area that the Army could influence and the area shaped by the [air component]. The placement of the FSCL was so far in front of the forward edge of the battle area that neither divisional nor corps assets could effectively manage the battlespace. . . . The argument seems to be that [the air component] would not adequately address V Corps targeting requirements; 3 ID (M) [3rd Infantry Division, Mechanized] violently disagrees. [The CAOC is] manned and equipped to effectively manage this battlespace forward of the FSCL; V Corps is not and has demonstrated their inability to manage said battlespace. 3 ID (M) believes that [the air component] is better prepared to engage targets to effectively shape the battlefield versus V Corps’ use of CAS.⁹⁶

To be sure, cross-service misunderstandings of important service-related specifics did occur. For example, General Franks once asked General Moseley for eight Global Hawks—more than the Air Force had in its entire RQ-4 inventory. Misuse of terminology also sometimes led to misunderstandings between elements of different components in resource allocation, options planning, and execution. And there were at least three reported instances in which U.S. ground units failed to coordinate with the CAOC with respect to long-range ATACMS missiles that were fired into CAOC-controlled airspace. Fortunately, no fratricide incidents resulted from these three communications failures, all of which were attributed to V Corps.97 An Air Force airman who had commanded an A-10 squadron that was mainly employed in support of SOF operations in Iraqi Freedom summed up the situation: “The air-ground coordination was . . . far from perfect and in many cases didn’t exist. Just like Desert Storm, we made it work to support the guys on the ground.”⁹⁸

Despite recurrent problems, however, the mutually supporting application of air and land power in common battlespace in Iraqi Freedom represented a major breakthrough in the conduct of joint warfare. General Leaf observed that the campaign experience represented the first time that the American armed forces had conducted a large-scale combat operation with the air and land components working side by side “as equals.” The air and land component commanders “achieved conceptual interoperability. . . . This was not [only] communications and software. We really had concepts linked. The real key was the collaborative planning at a senior level.” Leaf added: “We used methodology [that was] most appropriate. Sometimes ground preponderance, other times the air, other times in the middle. . . . My staff worked to ensure [that] timing and methodology worked together. For example, airmen usually use latitude and longitude to mark the FSCL, and ground commanders use geographical features. At times when the advance was rapid, airmen used latitude and longitude. At other times, it was necessary to use geographical features. Sometimes a combination was used to set the FSCL. The level of collaboration gave us flexibility, we did it collaboratively.”⁹⁹ In resounding testimony to the ultimate success of the arrangement, General Wallace declared proudly after the campaign was over that he did not lose a single soldier to an enemy weapon that could have been struck by coalition air power.¹⁰⁰ Throughout the three weeks of major combat in Iraqi Freedom, on-call CAS was overhead and available whenever the land component needed it.

The postcampaign recollections of the participating land and air combatants differ substantially, however, regarding the efficacy of the ASOC arrangement that had been put in place with V Corps in the wake of the near-disaster during Operation Anaconda in Afghanistan. General Wallace later said that he “never heard any of my commanders complain about the availability, responsiveness, or effectiveness of CAS—it was unprecedented.”101 General Moseley, however, in his capacity as the air component commander, described the ASOC arrangement as one of several areas of joint operations that had “worked less well” in meeting General Franks’ theater campaign needs and one that definitely warranted a closer “joint look.”¹⁰²

General Jumper expressed similar sentiments when he said: “There is a whole lot more we can do better.” He added that the services needed to keep talking to one another with respect to better air-ground integration to avoid continued misunderstandings in future conflicts. To cite one telling example, throughout the three weeks of major combat, the land component frequently wanted the air component’s CAS providers to fly low, still not realizing that more effective CAS delivery can be provided from as high as 39,000 feet. “Close air support to many can [still] only be defined as airplanes 50 feet above the ground that are releasing something like napalm and creating a lot of heat and a lot of smoke and a lot of noise,” General Jumper said. “What you really want,” General Jumper argued, “is those bad guys dead.”¹⁰³

While both services recognize the need for more realistic joint peacetime training, however, the Air Force and Army have yet to develop and implement approaches to it that would prepare both tactical forces and their operational staffs truly adequately for major counterland warfare. In particular, the tactical-level air-ground exercises that are typically conducted at the Army’s National Training Center at Fort Irwin, California, and in the Air Force’s associated Air Warrior program conducted out of Nellis AFB, Nevada, have characteristically suffered because the respective training and evaluation objectives of the two services have been so widely inconsistent. The Army—naturally enough—wants to initiate its tactical training engagements at Fort Irwin with a close fight immediately at hand so that its company-grade officers can learn how to understand and conduct high-intensity tank-on-tank warfare. For its part, the Air Force has unsuccessfully sought to start such exercises instead with the ground troops not in close contact with the enemy to demonstrate the full potential of air power in air-land warfare.¹⁰⁴ An early draft of JFCOM’s postcampaign “lessons learned” assessment was right on target when it insisted that in order to “duplicate the CENTCOM experience” in future major combat operations along the lines of the second Persian Gulf War, “a system that replicates joint operational-level warfare is required.”¹⁰⁵

Finally, in consonance with their U.S. Air Force compatriots, the RAF participants in Iraqi Freedom also found the integration of CAS with land component operations frequently inefficient, especially in the case of integration with British ground units, because of inadequate prior joint training between the RAF and the British Army over the preceding ten years.106 The RAF’s CAS effort was further hampered at times by the inability of airborne aircrews to secure sufficiently refined coordinates for attacking mobile targets. Because ground forces typically plot positions on maps rather than by using GPS, the targeting information provided by those forces often became rapidly outdated, creating a resultant need for aircrews to visually identify mobile targets before attacking them.

The failure of British ground commanders to recognize that CAS aircrews sometimes had go through the appropriate approval channels to clear specific target requests was another point of contention. In general, the MoD’s after-action report found that in the area of air-land integration, “operational and tactical doctrine does not yet fully reflect the demands of high-tempo, time-sensitive or network-enabled operations.”¹⁰⁷ It further noted that a “lack of experience in requesting, coordinating, and delivering CAS missions . . . was apparent.”¹⁰⁸ This was all the more disconcerting given the prevalence of CAS delivery needs in Iraqi Freedom. The report noted that as many as a third of all KI/CAS missions had to be aborted because of problems in the air-land interface (more specifically because V Corps and I MEF were not able to provide tasking for aircraft conducting KI/CAS missions before those aircraft reached bingo fuel state and had to return home).¹⁰⁹

In explaining why this had occurred, the MoD report observed that Operation Desert Storm had featured an extended and highly focused allied air campaign against fielded Iraqi forces followed by a very brief land offensive in which integrated air-land operations were seldom needed. Operation Enduring Freedom had entailed a closer integration between air and land forces, but the second Persian Gulf War, Air Marshal Torpy observed, was the first military operation in many years “where we have seen such close linkage between the air and land components. . . . We have forgotten some of the things that we were quite good at during the Cold War. . . . We have probably neglected the exercising of those over the years.”¹¹⁰ Now that CAS is an integral part of high-tempo warfare for producing immediate strategic effects, Air Chief Marshal Burridge concluded, “the relationship between air and land is now much, much more important.”¹¹¹ The MoD report spotlighted identified shortcomings in joint RAF–British Army training for CAS and suggested reassessing the number of JTACs and FAC-As maintained in service as well as the adequacy of the equipment provided for FAC-As, such as targeting pods with their inherent limits on range and resolution. It also called for using these assets more frequently in the joint training arena.

The many problems with U.S. air-land integration explored in the preceding two sections were expressly recognized in a postwar undertaking in the United Kingdom called Project Coningham-Keyes, a joint initiative between the British Army’s Land Command and RAF Strike Command to identify and address current capability shortfalls, with special emphasis on tactical-level execution. That initiative led to high-level joint doctrine for CAS being improved in an October 2003 document called Joint Air Operations. Coningham-Keyes focused on training for JTACs and TACPs, with a view toward realigning exercises between the two commands to improve training opportunities. It also examined staff organizations within the land and air component headquarters and considered their interactions. The inquiry determined that there were too few people dealing with the air-land interface and identified an insufficiency of ALOs in ground units. Shortfalls identified by Coningham-Keyes were expected to lead to a virtual doubling of the numbers of certified JTACs and TACPs available to British Army units.112

Delays in Battle Damage Assessment

The flaws in CENTCOM’s means for conducting and delivering timely, effects-based battle damage assessment (BDA) had been identified as far back as Desert Storm. In an effort to help meet this important progress-tracking need, General Moseley set up an operational assessment team within the CAOC’s strategy division to provide a running account of the effects that both kinetic and nonkinetic air operations were having on the Iraqi leadership and armed forces. This assessment process entailed a continuing effort, and its results were folded into a readjustment of targets as needed for the next ATO cycle. In connection with this systematic strategy-to-tasks planning process, the assessment team developed measures of effectiveness for the overall air effort. The CAOC’s assessment division likewise contributed to the effort.

Sources of information that went into this process included initial radioed assessments of mission performance by combat aircrews during their return to base, followed by more formal written reports that were then forwarded to the CAOC. Such reporting provided the best immediate insights regarding the results of missions flown against other than preplanned targets, such as on-call CAS attacks. ISR inputs from aircraft targeting pods, cockpit weapons system video, UAV imagery, and satellites went into the reports as well. Obvious indicators of strike effectiveness would be visible signs of destruction due to a successful weapon impact. Second-order indicators might include a sudden absence of electronic emissions or vehicle traffic. Most measures of effectiveness went well beyond simply assessing physical bomb damage.

This process worked reasonably well for fairly slow-paced target attacks such as those conducted throughout Operations Southern Watch and Enduring Freedom. General Moseley’s chief strategist, Lieutenant Colonel Hathaway, frankly conceded, however, that in the frenetic atmosphere of Iraqi Freedom, with two thousand sorties a day or more, “there was no way to keep pace with execution.”113 Fortunately, he noted, the initial phases of the air war went essentially as planned, with few significant last-minute adjustments required. But it soon became apparent to Colonel Hathaway and others at CENTCOM that the overall picture of combat achievements “would be neither timely nor complete.”¹¹⁴ CENTCOM’s production of BDA was simply unable to keep pace with the speed of air and land operations, a deficiency that in turn undermined munitions effectiveness assessment (MEA) and target reattack recommendations. During CENTAF’s lessons learned conference at Nellis AFB in July 2003, General Moseley underscored his own concern regarding this deficiency when he declared emphatically: “Two wars without [a real assessment process] are enough. . . . I never received adequate, timely feedback. I basically had to wait for CENTCOM to produce the official BDA to have any idea of what happened.”¹¹⁵ One account observed that “official BDA” was simply “not timely enough to adjust the ongoing operations. When Moseley told Hathaway to close the loop, Hathaway [and others] . . . worked . . . to figure out what targets had been attacked, based on the mission reports. Then they made some assumptions based on the type of weapons used—precision munitions were given a high probability of hitting the target. It was a ‘Band-Aid’ on a broken process, but it was the best they could do.”¹¹⁶ As David Johnson described it, “effects-based operations in [Operation Iraqi Freedom] remained more art than science.”¹¹⁷

Operational-level assessment of combat performance likewise suffered because it depends on reliable BDA and MEA to provide a foundation for further effects-based assessment of strike operations. Because of the rapid advance of allied ground units toward Baghdad after the line of departure was crossed, the CAOC necessarily focused the bulk of its collection efforts on finding targets and delegated BDA collection to the bottom of the priorities list. A CENTAF staffer later commented in this respect, “Probably the biggest lesson not learned [after the three-week campaign] entailed data collection during mission execution.”118 I MEF’s commanding general likewise complained that “target tracking and assessment was extremely difficult. . . . There was no reliable and responsive process or means to determine whether air interdiction (AI) targets on the PTL [priority target list] were serviced and successfully attacked during and after ATO execution. The impact was that targeting personnel/LNOs [liaison officers] could not consistently and reliably provide the necessary feedback to MSC [major subordinate command] commanders that their AI target nominations were being serviced or not.”¹¹⁹

A major undesirable result of its flawed BDA process was that all of CENTCOM’s warfighting components, not just the air component, often found themselves short of badly needed real-time situation awareness. To cite one case in point, Colonel Hathaway recalled that

           during the sandstorms on or about the ninth day of execution, the [land component commander], General McKiernan, called General Moseley with concern that the CAOC was hoarding ISR and BDA. General Moseley assured General McKiernan that he [General Moseley] was just as much in the dark with respect to BDA. This lack of timely assessment limited the ability of the [land component] to determine the strength and movement of Iraqi ground forces. That uncertainty forced McKiernan to change his strategy to “maneuver to contact.” This was a much less efficient form of offensive maneuver, but it was [nonetheless] necessary due to the unknowns.¹²⁰

Indeed, CENTCOM’s BDA process has long been widely regarded, at least among participating airmen at all levels, as unsatisfactory to the point of being useless. Colonel Hathaway observed on this point that the scale and rapidity of the air war “emphasized the weaknesses in the assessment chain, weaknesses that might have resulted in significant fog and friction had the plan not been so well thought out or the enemy more competent.”¹²¹ Lack of adequate manning and communications confined the BDA team in the CAOC to collecting and conveying in-flight reports, mission reports, and the fully executed ATO to CENTCOM’s forward headquarters so that staffers there could then conduct more detailed assessments of all assigned targets that had been attacked. A CENTAF planner later noted, “Both BDA and combat assessment failed to meet expectations due to the dynamic nature of the ground war, numerous ATO changes, problems with CENTCOM’s federated processes, the lack of realistic training and exercise of the BDA cycle, and the initial lack of a centralized tracking tool in the combat operations division.”¹²²

A big part of the BDA problem in that respect was—and remains—the persistence of reliance on destruction- and attrition-based rather than effects-based feedback. Regarding this important issue area, an Air Combat Command after-action briefing on the campaign flatly concluded: “We perform force application better than we can assess its effects.”123 The satellite-aided JDAMs were so accurate and reliable that MAAP planners in the CAOC were generally content to conclude that a designated aim point had been adequately “serviced” if a valid JDAM release had been assessed.¹²⁴ In their considered view, poststrike target imagery was essential only for particularly high-priority and high-value targets. This perspective inclined planners to conclude that a major change was warranted in the way BDA was conducted, at least with respect to targets attacked by JDAMs. Because any JDAM that had been released within proper parameters could be instantly deemed adequate for most targets, “the lightning speed with which the air component can destroy volumes of aim points demand[ed] a move away from the archaic, analog system that requires an electro-optical picture showing damage in order to categorize a target as dead.”¹²⁵ More to the point, they concluded that senior leaders should cease insisting on visual assessments of attacked targets and instead settle for accurate reporting of the combat effects achieved by aerial strikes. On numerous occasions during the period of major combat, targets located and identified by overhead imagery, particularly IADS-related targets, were repeatedly renominated for inclusion on the ATO, even though those targets had already been struck on multiple occasions and no electronic emissions had since been detected from their radars. In light of that experience, there seems every reason for CENTCOM’s leaders in future engagements to settle for effects-based BDA rather than continuing to rely on “mere observations of physical damage and bean-counting.”¹²⁶

BDA throughout the major combat phase of Iraqi Freedom typically lagged behind strike operations by as much as several days, and CENTCOM staffers made little effort to assess combat effects as a high-priority concern.¹²⁷ The senior CAOC director during the campaign later characterized CENTCOM’s BDA process throughout the three weeks of major combat as “broken.” He cited as just one of many examples the last-minute re-roling of a B-2 that occurred when CAOC operators determined from overhead sensors that a critical target against which the aircraft had been tasked had already been destroyed.¹²⁸

In a conclusion that clearly emerged from the CAOC’s experience with BDA throughout the three-week air war, Colonel Hathaway suggested that planners in future contingencies will need to formalize the assumptions they use in assessing weapons-delivery effectiveness so that the entire warfighting command understands the premises on which BDA can be provided on short notice. Further, high-value targets “must get a higher priority within the ISR collection deck,” he said, “because not all targets are created equal.”129

A widely cited early draft of JFCOM’s after-action study reported that “collecting, analyzing, and assessing information became the bottleneck in the decision-making cycle. Once information was available and commanders made their assessments, coalition forces took rapid advantage, striking with lethality and effect.” However, the study added, “when the speed of execution exceeded the capability to analyze and assess how those actions were changing the Iraqi system, operations reverted from an effects focus to an attrition focus. . . . Thus, while espousing an [effects-based operations] philosophy, ultimately the coalition moved toward an attrition-based assessment approach. . . . On the whole, coalition forces reverted to counting specific numbers of targets destroyed to determine combat progress rather than evaluating the effects created on the enemy.” The study noted that the coalition “did not have a sufficiently sophisticated and mature approach to rapidly assess the enemy from a system-of-systems perspective. Better cultural understanding and ‘red teaming’ are needed to break the trap of ‘mirror-imaging’ that can miss important indicators.” The study further observed that CENTCOM lacked an adequate approach to assessing nonkinetic effects and a methodology for providing “the subtle synthesis of direct and nondirect indicators needed to evaluate effectiveness.”¹³⁰

To offer an illustration of how this problem played out in practice, CENTCOM’s senior commanders were frequently concerned that allied ground forces might encounter stronger than expected enemy forces or be forced to fight from unanticipated positions. General Leaf noted that the problem was not simply BDA per se, but rather “really understanding . . . [whether] we’ve achieved the [desired] effects.”¹³¹ With allied ground forces having already advanced two hundred miles into Iraq within the first five days of the campaign, there was bound to be a time lag in determining what combat effects had already been achieved by allied air attacks. The challenge in that situation was not really getting “damage” assessment so much as a timely and accurate assessment of achieved effects. In previous conflicts, said General Leaf, “there was a fairly stodgy pace to conducting assessment—go bomb a target and then collect information on what has been bombed, and then do some analysis and decide what the effects were.”¹³² This slow and impacted assessment process adversely affected virtually all of CENTCOM’s warfighting components.

While allied interdiction and CAS sorties were shredding Republican Guard units for four straight days and nights, for example, the land component’s headquarters continued to portray the assessed strength of those units as high as 85 percent because empirical evidence to the contrary was lacking. Some commanders actually slowed their rate of advance commensurately until they could be certain what they were up against. General Leaf recalled that the targeting cycle would typically begin with meetings of the coalition’s daily effects board to focus the targeting process on desired results rather than on such instrumentalities as aircraft and munitions. General McKiernan’s list of target-servicing nominations and priorities was then passed along to General Moseley, whose battle staff would build a single target list and seek optimum ways of meeting those priorities by achieving desired combat effects, ultimately culminating in a daily ATO. In the view of many in both the air and land components, the assessment process in support of this effort could have been speeded up significantly had strike assessors had access even to rough, unanalyzed imagery.

CENTCOM also experienced difficulty in promptly determining the collateral damage consequences of land component operations. A briefing by Air Combat Command on the air war clearly pointed out that “in air operations, we have developed, by necessity, a series of tools and operational procedures to accurately assess and mitigate collateral damage. No such tools exist for surface-centric operations, especially small-unit organic fires.”133 Nearly 1,500 cluster bombs had been dropped on Iraq throughout the three-week offensive, and only 26 of them had been targeted within 1,500 feet of civilian neighborhoods. Those munitions reportedly caused only one known Iraqi civilian casualty. The Air Force had learned its lesson well in Afghanistan, where 1,228 cluster munitions had reportedly killed as many as 127 Afghan civilians. Allied ground forces, however, lacking familiarity with the collateral-damage mitigation tools and procedures that are routinely employed by Western air arms, reportedly expended cluster munitions (in artillery shells and ATACMS warheads) with less discipline during the major combat phase of Iraqi Freedom.¹³⁴

One long-term solution for the BDA deficiency described above would be for the services to develop and field weapons data links that are capable of providing auto-BDA by, for example, putting a red “X” over a just-attacked target symbol on a CAOC situation display to help reduce the need for unnecessary reattacks. Strike assessors must also develop and apply a new effects-based combat assessment methodology.¹³⁵

Problems in Meeting Tanker Requirements

General Moseley later described tanker support as “the single-point failure factor” in the air portion of CENTCOM’s joint and combined operations plan. He added that a failure by the air component to marshal and sustain a tanker flow sufficient to meet the in-flight refueling needs of the plan’s daily ATO could easily have been a “showstopper.”136 As it was, the shortage of tankers created serious complications in both the planning and execution of the campaign. The challenge was further exacerbated by the need to distribute the fewer than 200 available U.S. Air Force and RAF tankers over 15 bases, in contrast to the Desert Storm precedent, when 350 tankers had operated out of only 5 regional bases. Three bases in Saudi Arabia—King Khalid Military City, Jeddah, and Dhahran—had the needed ramp space and fuel throughput to support tanker operations, but the Saudi government disallowed their use for supporting strike missions. Operating locations as far away as Moron Air Base in Spain, Akrotiri on Cyprus, Diego Garcia in the Indian Ocean, and Cairo West were used instead.

The tanker problem was further aggravated by Turkey’s refusal to allow the coalition to use its bases, as well as—for a time—its denial of overflight permission. Turkish bases had constituted some 25 percent of the bases that CENTCOM had earmarked as forward operating locations for its 149 KC-135s, 33 KC-10s, 4 RAF Tristars, and 8 RAF VC-10s. Their unavailability required operating tankers for the first time from bases in Crete and Bulgaria. Because of these and related complications, CENTCOM had 51 fewer KC-135s available for tanker support than it had for Desert Storm.¹³⁷ Moreover, in order for the carrier-based fighters operating from the eastern Mediterranean to participate in the initial attack waves, the expeditionary air wing in Egypt and tankers based in Bulgaria had to provide fuel to other tankers that had no alternative but to fly around Israel and over Saudi Arabia to support allied strike aircraft flying into northern Iraq.¹³⁸ The last-minute loss of basing rights in Turkey and the limited tanker basing in Saudi Arabia created aerial force employment distances and sortie durations throughout Iraqi Freedom that, in many cases, matched those routinely flown into land-locked Afghanistan throughout Enduring Freedom.

Only the three bases in Kuwait that had been made available to allied strike aircraft were close enough to Iraq’s southern border to allow strikers to reach deep into Iraq without refueling. Strike aircraft from all other bases within reasonable range of Baghdad required in-flight refueling during both ingress and egress. With hundreds of allied strike aircraft airborne at any given moment, queues of fighters behind tankers sometimes became so long that pilots had to abort their missions because they lacked sufficient fuel to hold for a preplanned tanker rendezvous. A related concern entailed positioning tankers far enough forward and timing scheduled tanker connections in a way best guaranteed to prevent strike aircraft from having to queue up in the first place and become inviting targets for enemy air defenses. Yet another complicating factor was the number of aircraft that were executing sustained holding and combat operations over Baghdad. During Operation Desert Storm, fuel-consuming CAS and SEAD loitering operations over Baghdad and other parts of Iraq in which ground combat operations were under way were uncommon.139

The war’s initial days brought recurrent complaints about missed tanker connections from Gulf-based Navy strike pilots who could not get to the fight without refueling after the battle lines had moved more than two hundred miles north of Kuwait. Tanker apportionment planners in the CAOC countered that the complainers failed to appreciate the complexities involved. A canceled tanker rendezvous, for example, did not necessarily mean that the assigned tanker had not been on station. With the allied ground advance moving northward toward Baghdad with such dispatch, target sets sometimes shifted so rapidly that airborne strike sorties were deleted from the ATO after the aircraft were airborne because they were no longer needed to service targets, not because a needed tanker was not available. Vice Admiral Keating confirmed that during the war’s first days, various tanker management challenges resulted in some carrier-based aircraft not prosecuting their targets and others having to divert into one of the Kuwaiti bases to top off before returning to their ships.¹⁴⁰ As the campaign progressed, tanker management became less of a problem. Capt. David Rogers, Admiral Nichols’ special assistant, observed at the time, “If you ask somebody whether the tanking situation is better or worse than when we kicked things off, you would get a resounding change in opinion from what they would have told you earlier. Basically, we’re getting fuel closer to the target.”¹⁴¹

Inadequate fuel lines at tanker bed-down sites that necessitated trucking in jet fuel from the outside added to the tanker-related problems. This dependence on local fuel trucks increased tanker turnaround times and reduced tanker availability for strike operations into northern Iraq from the two carriers in the eastern Mediterranean. Captain Rogers called the situation “a more complex shell game than in Desert Storm.”¹⁴² Further, the ground offensive’s rapid northward advance—before the CAOC had time to establish the desired margin of air superiority over the battlefield—soon outranged some of CENTAF’s fighters. The difficulty was further compounded when General McKiernan moved the FSCL well forward of his own forces, leaving some allied aircraft assigned to provide on-call CAS with only minutes of available loiter time. More than a few had to recover with unexpended ordnance because the ASOC could not assign targets to them quickly enough. To compensate for this, the CAOC pushed some tanker tracks forward and also assigned shorter-range aircraft like the AV-8B and A-10 to support operations solely in the southernmost portions of the battlespace.

The problem was ameliorated when coalition ground units took Tallil Air Base and A-10s began flying CAS missions directly out of that forward location.143 Col. James Dobbins, the commander of the 392nd Air Expeditionary Group that moved into Tallil, explained that the A-10s now had “more time over target [because they] . . . did not have to fly as far to get to the fight.” In addition to enabling forward-based A-10 operations, Tallil also supported some airlift and CSAR missions. Although lacking such other critical needs as power grids, communications facilities, and water distribution systems, the base offered an adequate fuel capability for supporting initial U.S. flight operations so that A-10s could land, refuel, upload fresh ordnance, and get promptly back in the air.¹⁴⁴

With more in-flight refueling occurring closer to the war zone as the campaign progressed, the distribution and placement of tanker orbits necessarily became more complicated. In view of the continuing enemy IADS threat, it was not feasible for the CAOC to maintain predictable tanker tracks each day with large stacks of aircraft waiting to take on fuel. That required more airborne tanker stations and often occasioned greater difficulty for strike aircrews in finding and rendezvousing with their assigned tanker. Furthermore, a tanker pilot might be more likely to break off a refueling if he felt that his aircraft was in imminent danger of being fired on. Unlike the coalition’s strike aircraft, the tankers were not equipped with radar warning receivers or chaff and flares to provide threat situation awareness and active countermeasures against radar- and infrared-guided SAMs.

After CENTAF’s defense suppression operations had reduced the assessed threat potential of Iraq’s IADS to a sufficiently low level of risk, General Moseley moved tanker orbits and command and control and manned ISR aircraft, such as the E-3 AWACS, E-8 JSTARS, RC-135 Rivet Joint, and U-2, forward into Iraqi airspace in support of the advancing land forces. Indeed, OPLAN 1003V had always presumed that tanker tracks would be moved into Iraq just as soon as the assessed threat environment would permit it, because there was no other way for CENTAF to support a major allied land offensive in the vicinity of Baghdad. That long-standing plan, however, had presumed that G-day would begin three days after A-day. The FLOT had moved so far north from the line of departure in Kuwait after just the first day of combat operations that Navy and Marine Corps strike fighters operating from carriers in the North Arabian Gulf could no longer reach the fight on the ground without refueling in flight at least once.

As soon as CAOC planners opened the first tanker tracks inside Iraqi airspace, the Navy began using its S-3Bs and F/A-18E/F Super Hornets to provide organic tanking for some carrier-based sorties that were tasked to support the land component. These were the first of many measures the CAOC undertook to overcome the ever-widening distances to allied ground units advancing toward Baghdad. The assessed urgency of moving tankers into defended Iraqi airspace as soon as possible left little time to address possible IADS threats, and tanker planners were understandably reluctant at first to conduct in-flight refueling over southern Iraq, particularly at the lower altitudes that the A-10s required.145 Yet another attempted palliative to address the tanker and fuel shortage entailed tasking A-10s to support I MEF in southeastern Iraq while Marine Corps strike fighters from Al Jaber and carrier air wings in the North Arabian Gulf operated farther north to support the parallel advance by V Corps. In this arrangement, General Moseley, in effect, traded a squadron of air component–controlled A-10s to I MEF to meet General Conway’s CAS needs in return for a squadron of Marine Corps F/A-18Cs that the CAOC could use in meeting its deep-attack requirements farther north.

The two carriers that had been on station in the North Arabian Gulf for the longest time, Abraham Lincoln and Constellation, were selected to receive the greatest possible Air Force tanking support. The air wing embarked in the last carrier to arrive on station, Kitty Hawk, was given a predominantly CAS role in support of V Corps and I MEF in southern Iraq and was tasked to fulfill that role autonomously, which burdened the carrier’s organic tanking capability to its limit. For the first time since Operation Desert Storm, strike aircraft in CVW-5 embarked in Kitty Hawk were allowed to return to the carrier with minimum fuel, and carrier cyclic and flex-deck operations ran at a maximum.¹⁴⁶ Marshaling these capabilities for an anticipated five-day surge option, CVW-5 proved itself able both to meet its CAS tasking and to provide strike packages for attacks against fixed targets in Baghdad. The S-3B tankers of CVW-5 were crucial in making that flexible employment possible.

The other two carrier air wings followed similar plans to enhance their capabilities. It turned out that S-3Bs in the Gulf-based air wings were sufficient in number and capability to permit the release of some Air Force tankers allocated to Abraham Lincoln and Constellation to support other strike sorties emerging from the Gulf region, including from Kitty Hawk, as well as Air Force heavy bombers arriving from Diego Garcia and elsewhere. This concept had been tested and validated during the final days of Operation Southern Watch, during which frontside and backside organic tanking (that is, in-flight refueling of fighters both on the outbound leg to their assigned target area and near or over the carrier shortly before their recovery) allowed F/A-18s to fly one-and-a-half-hour cycles with little difficulty, with S-3Bs launching and recovering immediately before and after the refueling evolutions. It also allowed an increase in on-station time for carrier-based strike fighters conducting CAS missions in southern Iraq around Basra and Nasiriya, and it showed that a five-day surge capability could be sustained on an open-ended basis throughout the days to come. By one informed account, the concept worked “flawlessly” throughout the three-week campaign.147 In the end, the five participating carrier air wings transferred about half a million pounds of fuel a day from organic S-3Bs and Super Hornets in addition to the five million pounds a day they received from U.S. Air Force and RAF tankers.

Coalition pilots recalled that some of their most harrowing moments during the three-week air war resulted from their inability to find a tanker at the end of a mission when their fuel supply was critically low. Said one pilot: “We’d call for our tanker, ask AWACS where he was, and be told, ‘he’s gone home already.’”¹⁴⁸ At first, tanker tracks inside Iraqi airspace were too few to provide adequate support to the land component’s KI/CAS needs, so alternative measures had to be undertaken. Defensive counterair operations were reduced by half; Air Force A-10s began launching combat sorties out of Tallil airfield; and F-15Es that had initially been operating out of Al Udeid in Qatar began generating combat sorties out of Al Jaber Air Base in Kuwait, which was closer to the Iraqi border. These measures allowed aircraft assigned to KI/CAS missions longer times on station without the need for in-flight refueling.¹⁴⁹

There were several days during the march toward Baghdad when in-flight refueling requests from airborne strike aircraft conducting CAS operations went unmet due to conflicting interpretations of General Moseley’s guidance regarding tanker track placement inside still-defended Iraqi airspace and risk mitigation. In particular, the period between day 4 and the regime collapse on April 9 saw some fuel requests for A-10 and other coalition CAS aircraft, as well as for A-10 support to the SOF component’s Task Force 20 in Iraq’s western desert, going unsatisfied. Dealing decisively with this issue required General Moseley’s personal adjudication. CAOC planners later explained that “increased tanker risk acceptance by air refueling track placement into Iraq’s southern plain was necessary to support ground operations, but the doctrinally inspired organizational structure caused resolution of such conflicts to occur at the [air component commander] level. Resolution couldn’t always occur in a timely manner because of the multiple layers of leadership that existed between the MAAP chief and [General Moseley].”¹⁵⁰

Planners in all components had tended to assume going into Iraqi Freedom that Southwest Asia’s enormous fuel reserves would make fuel shortages nonexistent. Yet as CAOC operators later pointed out, “walls of [defensive counterair], robust global power, numerous land component target nominations, and 24-hour command and control and ISR coverage all constituted examples of legitimate needs that also happen to be very expensive in terms of fuel. . . . [P]lanners must not assume all the gas will be there to fulfill their wish lists. Tankers mustn’t be an afterthought in wartime planning—they should be the first thought.”151

Obstacles in Coalition Information Sharing

A final source of frustration in the day-to-day conduct of the major combat phase of Iraqi Freedom had to do with the sharing of sensitive information among the American and allied contingents. Although openness and candor between the allies were far better in this war than in any preceding instance because of the few coalition partners that were involved and their exceptional closeness, there was nonetheless difficulty at times with what one RAF officer described as “translating the trust engendered at the highest levels into sensible information sharing at lower levels.”¹⁵² The British and Australian contingents had full access to the American STU III secure telephone system and also could receive the product of TBMCS indirectly by means of read-only compact discs. Beyond that, however, the allies had more than a few problems getting timely access to U.S.-controlled information.

In particular, many of CENTCOM’s most important command and control processes were based on SIPRNet (secure Internet protocol router network), a U.S. eyes-only system that could not be directly accessed by foreign nationals.¹⁵³ Accordingly, allied personnel in the CAOC were not allowed to work directly with such crucial SIPR-based U.S.-only planning mechanisms as TBMCS. As an attempted work-around solution, CENTCOM provided the allied contingents with the wherewithal to create a mirror image of SIPRNet called Xnet, which allowed them to use a separate TBMCS system that was not SIPR-based and accordingly was not linked to TBMCS. That makeshift arrangement allowed coalition planners to build their mission plans in a separate TBMCS “shell” that ATO production staffers would then merge into the overall ATO. Yet the MAAP cell’s principal means for building mission sets used a MAAP toolkit rather than TBMCS. Because the MAAP toolkit was also SIPR-based, coalition staff were not allowed access to it either. Moreover, although the air component’s daily SPINs were, of clear necessity, approved for release to the coalition partners and were routinely laden with specific coalition-partner information and guidance, the allies’ lack of direct access to both TBMCS and SIPRNet forced CAOC staffers to burn a compact disc every day and hand-deliver it to RAF and RAAF mission planners, which created an additional source of self-inflicted friction.

In other ways as well, the coalition representatives’ American counterparts were obliged to pick up the slack created by the stringent security measures, leading to heavier workloads and longer timelines for assembling attack packages. For example, before options planning for OPLAN 1003V began in earnest, standard U.S. intelligence management procedures dictated classifying most of the targeting information generated by CENTCOM and by the broader U.S. national intelligence community at the Secret/NOFORN level. Once actual combat operations neared, however, both CENTCOM and the broader intelligence community found themselves faced with the massive task of sanitizing large amounts of NOFORN targeting information so as to render it releasable to CENTCOM’s coalition partners. The arrangement for sharing vital planning data that resulted was described by an RAF officer as “slow and cumbersome rather than responsive and agile.”154

The CAOC staffers who produced CENTAF’s after-action review of the campaign experience described the allies’ lack of access to key mission-planning systems as “the single greatest frustration for what was otherwise an outstanding relationship” among the coalition partners. One key planner recalled pointedly that “the impact of this issue was felt throughout the CAOC.”¹⁵⁵ Senior RAF officers who took part in Iraqi Freedom at both command and execution levels were likewise all but unanimous in spotlighting the NOFORN caveat that blocked ready allied access to SIPRNet, Intellink, and other U.S.-only information systems as a persistent and systemic impediment to fuller interoperability with U.S. forces.¹⁵⁶ They further singled out this problem as one that senior leaders in all three countries needed urgently to address in order to find a mutually agreeable “better way” that might allow less-fettered dialogue between special partners in the midst of high-intensity coalition warfare.¹⁵⁷

Air Chief Marshal Burridge later remarked that the challenge presented by the access issue had not been prohibitively imposing in the case of Iraqi Freedom because the coalition partners of major note were solely the United Kingdom and Australia, both of which were close allies of the United States accustomed for years to operating “inside a known agreement for sharing intelligence.” Because the process was manual rather than automatic, however, it required the American participants to find time in the midst of continuous high-tempo combat to decide on and carry out the transfer of information to the allied contingents as it was needed. This inherent impediment to smoother planning and execution was generally overcome through exceptionally close interpersonal and trust relations among the three contingents, but not without recurrent friction and inefficiency. Burridge further pointed out that the efficacy of such a system would be limited, perhaps even to the breaking point, in future contingency responses by a larger coalition that included partners “with whom we would not normally consider sharing high-grade intelligence.”158 Many CENTAF staffers recognized the necessity for both CENTCOM and the broader American national intelligence community to adopt a fundamentally different mindset aimed at facilitating allied combat options planning for future contingencies. Simply adding a caveat indicating “releasable to Great Britain, Australia, and Canada” would obviate the last-minute scrambling to sanitize targeting information that plagued Iraqi Freedom.

On the positive side, there was a clear contrast between the air tasking process employed throughout Iraqi Freedom and the one used during Operation Allied Force against Serbia in 1999, when the air component developed two separate ATOs each day as a result of EUCOM’s determination to keep sorties using NATO assets and approved through NATO channels separate and deconflicted from the more sensitive use of U.S.-only assets, such as B-2 and F-117 stealth aircraft.¹⁵⁹ From the very beginning, CENTAF’s planners maintained a “seamless integration” of U.S.-only stealth assets with other American and allied conventional aircraft. Previously, planners had thought that the stealthy F-117 and B-2 demanded special compartmentalization in any air operation involving allied participation to protect their most sensitive operational and performance details. During the buildup for and conduct of Operation Iraqi Freedom, however, stealth platforms were woven into the CAOC’s conventional planning process from the very start, making the integration of stealth aircraft with conventional assets much more efficient and effective.¹⁶⁰

As was the case with participating units from the RAF, the RAAF units and personnel seconded to CENTAF for the duration of the three-week campaign encountered few problems in the realm of force interoperability. A subsequent bilateral review of interoperability issues between the ADF and the U.S. armed forces conducted in 2004 at Victoria Barracks in Sydney, Australia, however, concluded that “not all interoperability achieved was the result of systematic planning and training. Shortfalls arose, and the high degree of operational-level interoperability achieved during the conflicts [both] in Afghanistan and Iraq often reflected the use of workarounds and ad hoc solutions.” One of three broad categories of identified interoperability shortcomings concerned “information exchange issues, including the use of computer networks to plan, execute, and monitor operations and the extent to which national information policy architectures could respond to coalition information needs.”161

An initial problem faced by the ADF that was not shared by the British contingent, owing to the latter’s previous intimate involvement in Operations Northern Watch and Southern Watch, was near-total unfamiliarity with CENTCOM and its mode of operations. CENTCOM likewise had limited knowledge and appreciation of the ADF’s breadth of capabilities. By the time the final countdown for Operation Iraqi Freedom began in early 2003, however, that issue had largely been resolved as a result of the ADF’s contribution to CENTCOM operations in Afghanistan since October 2001 through Operation Slipper, the air portion of which consisted of the provision of four RAAF F/A-18s to provide local air defense of Diego Garcia and two Boeing 707 tankers operating out of Ganci Air Base near Manas, Kyrgyzstan, with two RAAF P-3 long-range maritime patrol aircraft being added in 2003.

Nevertheless, the Australians still faced an initial disadvantage when compared with their RAF compatriots. The air wing commander for the RAAF F/A-18s that deployed to Al Udeid recalled: “I never knew what an air expeditionary wing was, or what they did prior to getting there. Very few people [in the RAAF] had any idea of what a CAOC was and how it conducted operations and how the [United States] did its business. If you had asked me in November prior to being briefed in how we would conduct business with the [United States], I’d be at a loss to tell you.”¹⁶² The chief of intelligence and targeting within the Australian CAOC contingent under Group Captain Brown similarly remarked in his after-action observations: “Working with the U.S. Air Force at Shaw was definitely a highlight. They were extremely professional guys and had been doing Operation Southern Watch for quite some time, so what they didn’t know about Iraq wasn’t worth knowing. [Yet] this was an out-of-area operation for us, so there was a lot to learn and not a lot of time. It wasn’t like being in the comfort zone of the Southwest Pacific or Southeast Asia, and even knowing the lay of the land, you had to learn pretty much everything from scratch.”¹⁶³

With respect to the sharing of sensitive U.S.-controlled information, however, the RAAF contingent embedded in CENTAF’s CAOC reported very much the same experience as did the RAF contingent, namely, a U.S.-imposed firewall that blocked their direct access to needed mission planning information transmitted via SIPRNet and the necessity for jury-rigged workarounds. To be sure, that was not the case during the initial planning for prospective coalition air operations that took place at Shaw AFB in late 2002. The RAAF’s chief of intelligence and targeting experienced outright astonishment at the remarkable openness of CENTAF’s willingness to share sensitive intelligence information: “I walked in and I got a pass that let me in anywhere. It was quite amazing, actually. I didn’t expect that. And I actually always felt a bit embarrassed letting myself into their [intelligence] area without knocking and saying ‘I’m here’ before I took a few steps. I can’t recall any situation where I felt they weren’t telling you something.’”164

Once the war was on, however, that openness changed. A senior air planner attached to the Australian contingent in the CAOC recalled in a postcampaign interview:

           General Moseley . . . and his deputy, Brigadier General [Robert] Elder, were both excellent. [General Moseley] just directed that whatever the Australians want, give it to them. . . . The biggest problem we had, [however], was plugging into their secure information technology systems and their communications, because that was a national sensitivity issue. The Brits had the same problem. . . . The TBMCS ran on SIPRNet, the secure system. That was where all the reporting occurred and where all the air tasking was issued. . . . They couldn’t give us access to it, and there was a huge problem with that. . . . It was eventually resolved, and I’m not clear how it exactly was resolved, but I know General Moseley had a major hand in it, and it went extremely high in their [the American] system. I heard a report . . . [that] it went as high as the president to get approval to give the United Kingdom and Australia access to SIPRNet. Until we got that, we could not operate. We couldn’t access the ATO, and all the command and control structure ran on that.¹⁶⁵

The U.S.-Australian bilateral review committee’s postcampaign assessment of interoperability issues between the two countries concluded that the impromptu informal workarounds that eventually triumphed over the ingrained obstructions to information sharing, albeit successful in the end, involved “departures from established, approved operating procedures”; were “usually based on a personal, trusted relationship”; and had “the net ongoing operational-level effect of . . . [degrading] the ability of Australian and U.S. forces” to interact seamlessly when it came to the most efficient planning and conduct of combat operations in a fast-paced and high-intensity coalition campaign context.¹⁶⁶ The review team singled out as the main culprits the American SIPRNet system at the Secret level and the JWICS at the Top Secret level. Both “by definition and design, [those pivotal planning systems] cannot be used for campaigns involving meaningful contributions from other countries.”167 Fortunately, however, such workarounds as the provision of relevant country-specific sections extracted from the ATO via a SIPRNet terminal either in hard copy or on a compact disc for the British and Australian contingents “were developed [and employed in a way] that allowed allies access to mission-critical information that was essential for their effective and safe participation” in the campaign. Nevertheless, the team added, such workarounds “are a poor solution to problems with information sharing among coalition members.” Furthermore, consistent with the overarching stumbling block with regard to intracoalition interoperability that the British contingent repeatedly spotlighted in its postcampaign assessments, the U.S.-Australian review team also pointedly cited a statement by General Franks in testimony before the Senate Armed Services Committee on July 9, 2003, that identified “coalition information sharing in [Operation Iraqi Freedom] as an area that must be improved at all levels.”¹⁶⁸