We wrote this chapter with a little trepidation, since it addresses some sensitive and potentially controversial topics. Nevertheless, we think it is important to discuss how the air campaign against fielded ground forces was guided and executed at the tactical level so that readers understand and appreciate the essential war-fighter lessons from our OAF experiences. The authors neither examined all aspects of the air campaign nor attempted to analyze and draw conclusions about the instructions that originated at higher levels of command. We only describe how the ROEs affected our operations—without speculation on the decision-making process that developed them. We discuss how ROEs can best serve an air campaign’s objectives, particularly in low- to medium-intensity air operations against ground forces. Our observations and conclusions are from a tactical perspective; from that perspective and at various times, we found the ROEs operationally constraining and war extending.
As the campaign’s first AFACs (and the ones who spent the most time over Kosovo), we detected, identified, selected, and engaged most of the fielded Serb forces that NATO engaged. We discuss target identification and ROEs in the same chapter because they had the greatest in-flight influence on determining which targets to leave alone and which to destroy.
Target identification was the critical process through which AFACs located potential targets and determined whether or not they were valid. We used many methods. First, we received a daily list from the CAOC that contained possible targets and their locations, such as “four tanks at coordinates yyyy North and xxxx East.” The list was developed during the 12 hours prior to its release, using the best available information. Even so, we quickly learned that it was hopelessly outdated and generally useless. While some of the information may have been incorrect from the start, it was more likely that the Serb forces had moved during the 12–24 hours it took for the data to be gathered, analyzed, and disseminated—and for us to launch, get overhead, and maneuver into a position to attack. The US Army’s Hunter UAVs were a much better source of timely information. Their usefulness, however, was still a function of the elapsed time from their observation to our getting overhead.
The crews on the E-8 JSTARS aircraft provided the location of vehicles or convoys that were moving in the KEZ. We found these crews to be very professional and accurate in their assessments of convoy size, makeup, and location. Initially we relied heavily on JSTARS for target information and found ourselves rushing (as fast as a Hog can rush) across Kosovo to identify all convoys, discovering that the vast majority of them were civilian vehicles. The Serbs quickly countered the JSTARS capability by hiding their armored vehicles and mixing their other military vehicles in among the civilian cars, trucks, and tractors that made up the large refugee convoys. Often we were called to identify a convoy while in the process of setting up to attack a valid target. We would abort that attack, fly to the convoy, confirm it was limited to refugee vehicles, run out of gas, and be forced to return to base or go to the tanker. That approach left the original valid targets undamaged. We agreed within the first two weeks to change tactics. The JSTARS crew members would give us an initial target-area briefing, and then we would “pull” information rather than have them “push” it—unless, of course, they saw a particularly interesting target.
There were too few JSTARS aircraft in-theater to provide around-the-clock coverage. The CAOC directed JSTARS to support daylight KEZ operations, which left the night uncovered. We later concluded that the E-8 would have been much more productive had it flown at night. The Serbs generally stopped moving equipment during the day because we had been successful in acquiring them visually and picking them off. However, they became very active at night—moving and digging like crazy. Every morning artillery had been moved to new pits; more revetments, tunnels, and dugouts had been constructed to protect APCs; and the wreckage of the armor and artillery that we destroyed the day before had been recovered and moved.
Our proposal to switch JSTARS from day to night operations was passed to the CAOC. We reasoned that the JSTARS crews would be able to observe the Serbs’ nightly movements and inform us where they had relocated their weaponry, so that at daybreak we could schwack ’em. JSTARS would also be able to assist the nighttime F-16CG and Guard A-10 FACs. The CAOC rejected our proposal, beliving that JSTARS aircraft were needed during the day, when most small Serbian military targets were being struck.
The lack of satellites or drone imagery was also a serious void in our tool kit for locating targets. We did not have a dedicated imagery-production process to provide our unit with photos of the Serbs’ fielded forces. Perhaps those targets did not have sufficient priority or had been deemed “too hard” to locate. Whatever the reason, the CAOC provided imagery on only a few occasions.
The 40th EOG’s operations-intelligence section contained some of the unsung heroes in the A-10’s OAF success. This outstanding group was comprised of elements from the 52d FW at Spang and the 23d Fighter Group at Pope, and was well led by 1st Lt Stephen “Al” Smith and Capt Kenneth R. “Ken” Uhler. Their creative response to the imagery problem provided us with many good pictures that resulted in numerous confirmed kills. They fought hard to acquire the equipment necessary to access the Web sites that contained classified US imagery and then spent two to three hours every night searching and downloading satellite target pictures of Serbian forces in the KEZ. They also built a solid working relationship with the British intelligence section located across the street from our operations building. Their GR-7 Harriers often carried photoreconnaissance pods, and our intel counterparts would process the morning’s film, analyze and enlarge the best prospects, and then bring them over to our unit. From these two sources, Al and Ken built the eagerly anticipated “Hog menu of the day” target list. Some days were richer than others, but the Hog menu usually included at least three and as many as seven targets.
Sadly, there was another source of beautiful imagery for finding Serb army targets that we were unable to exploit. The French army operated ground-controlled reconnaissance drones and routinely collected imagery during the course of the conflict that could have been very valuable. We were unable to use it because we did not even know it existed. I first became aware of this capability a year after OAF while visiting a large open house in September 2000 that was hosted jointly by the French army and the French Joint Defense College. There I met members of the 61st Artillery Regiment as they displayed their CL 289 and Crecerelle optical and IR-imagery drones. They explained that their drones were able to take both day and night pictures, and that they also had data links for real-time imagery transmission. Curious, I asked their leader if they had participated in any recent operations. A lieutenant said, “Yes, Kosovo.” I then asked them from where, and they replied “Kumanovo”—which had been our Macedonian entry point into the eastern half of the KEZ. I asked them how they employed their drones, and they said they flew them daily to determine the positions of Serb army tanks, APCs, and artillery—exactly in the areas where we flew our missions. They normally flew them at night so that they would have a good ground order of battle for the beginning of the day. I asked if their operations had been effective. They said, “Yes,” and added that they got very good, if not complete, information on Serb army dispositions. I asked them to whom they sent their imagery, and they said to the normal French army-intelligence channels.
I was flabbergasted. As the sun went up each day, we were always the first AFACs across the border, and we desperately needed that kind of imagery to find targets. The French imagery would have been perfect for us since it could have been delivered close enough to real time that the enemy could not have moved his weaponry before we arrived overhead. Someone had it—but we didn’t. I can’t speculate on where it went or why it never got to us, but obviously every possibility for obtaining crucial target information should be exploited.
Another good source of current target information was the CAOC’s force-level execution (FLEX) targeting cell. The FLEX cell fused information gained from various sources, using a variety of methods, and was often able to determine the location of actively operating Serb army units and command posts. For example, when a Serb artillery unit fired on a Kosovar village, the US Army’s counterbattery radar could plot the Serbs’ position. Similarly, when a Serb command post transmitted orders on its radios, our electronic warfare (EW) folks could triangulate its location. The FLEX cell also received target information from Predator drones. The cell fused all of this information and passed it to the ABCCC, which would then contact the AFAC in the area nearest the suspected target. The AFAC would then take a look, and our experience confirmed that the FLEX information was usually quite accurate. The marriage of drones and AFACs to locate and engage targets was a first for either combat or training operations. Sometimes it worked well, and sometimes it didn’t. Our pioneering work with Predator is addressed more fully in chapter 7.
Even with all this high-tech help, we still located about 80 percent of the targets we engaged using our Mk-1 eyeballs, augmented by our trusty 12-power gyrostabilized binoculars. A typical OAF AFAC scenario would begin with thorough flight planning using the best target information available from all sources. After takeoff the AFAC would contact ABCCC and/or JSTARS and integrate their updates into his mission planning while en route to the KEZ. He would then proceed directly to the target areas. The AFAC would focus on either the highest priority target or the one that had the highest likelihood of being found. Failing to find anything to attack at the target coordinates and lacking any other good target information, the AFAC would proceed to areas where his experience suggested that he might find something worth attacking.
When AFACs looked for a particular target around a set of coordinates, their observant, naked eyes could often spot telltale signs of other targets—new revetments, tracks leading into the woods, and unusually configured shapes on a hillside. “Well, looky here!” was our normal reaction. The AFAC would then use his binoculars to get a closer look, and if it proved to be a valid target, he’d set up attacks. After our AFACs returned to the squadron, they compiled their own list of AFAC-located (but not destroyed) targets to be included in the daily mission reports to the CAOC and for follow-on AFACs to use in their flight planning.
Rules of engagement are exactly what the term implies—rules that limit friendly forces’ operations as they engage hostile forces. There were two levels of ROEs during OAF: those imposed by the integrated NATO command authorities and those imposed by nations (or their organic commands) on their own forces. In most cases, national ROEs were more restrictive. For example, out of concern for the vulnerability of their particular aircraft, some nations raised the minimum-attack altitude for their aircraft to above that published in the NATO ROEs.
ROEs were proposed, developed, and changed at several levels within the command hierarchy. In our case, ROEs originated with and were reviewed by political authorities at the North Atlantic Council (NAC) in Brussels; SACEUR at Mons, Belgium; commander, AFSOUTH at Naples, Italy; commander, AIRSOUTH; and the CFACC. We understood that commander of AFSOUTH, at the operational level, was the chief ROE-setting authority.
NATO authorities “published” air-campaign ROEs in several documents that were then disseminated to subordinate headquarters and units. ROEs were published in the air operations directive and special instructions (SPINS). Additionally, because ROEs often changed, the CAOC dedicated a section of the daily ATO to list all of the ROEs applicable to that day’s sorties. All OAF units reviewed that dedicated section during their pre-mission preparation and found it very useful.
ROEs for KEZ operations fell into three general categories: altitude restrictions, restricted (no-attack) zones on the ground, and procedures to lower the risk of collateral damage. Altitude restrictions were designed to minimize the risk to aircrews from SAM and AAA threats. Higher altitude either puts the aircraft above the enemy’s effective capability or provides the aircrew with enough time to react and defeat the surface-to-air threats. On 30 March, at the outset of CAIFF operations, we were given the same minimum-altitude restriction as aircraft flying in interdiction packages—15,000 feet AGL. Although it kept us relatively safe, this altitude made identification of small military vehicles very difficult and rendered A-10 attacks practically impossible. By 6 April the mission-support elements of the standard KEZ package (SEAD, EW, ABCCC, NAEW, and combat air patrol [CAP]) had demonstrated their ability to suppress the most lethal Serb threats, and we were able to convince the CAOC to let AFACs descend to 10,000 feet AGL.
We operated with that minimum ROE altitude until 14 April—the tragic day when a civilian-vehicle convoy was incorrectly identified as a Serb military target and attacked by NATO aircraft. In reaction to this incident, the CAOC changed the minimum altitudes for KEZ operations to improve target confirmation. AFACs could descend to 5,000 feet AGL, and all fighters could descend to 8,000 feet AGL during target attack. These ROEs seemed backwards to us Hog drivers at Gioia. Most enemy missile launches occurred while the AFACs were searching for targets, focused on the ground, and doing very little maneuvering—not during weapons delivery. We immediately directed all of our pilots to conduct target search no lower than 10,000 feet AGL.
No-attack zones were geographically defined areas within Kosovo in which we could not expend ordnance. Restricted areas were also geographically defined areas within Kosovo where weapons could be employed only after receiving CAOC permission. Various authorities had established these zones for a variety of reasons. The first such area to appear in the ROEs was the no-attack zone within 10 nautical miles of the Macedonian border. While at first we did not understand, we later learned that the zone was meant to reduce the risk of possible Serb reprisals on NATO troops in Macedonia. Authorities apparently believed that the Serbs might confuse the source of the ordnance raining down on them. Their concern was that the Serbs would ignore the NATO fighters overhead, credit the attacks to NATO artillery to the south, and expand the war into Macedonia. Other restricted areas were created near the end of the campaign, particularly in areas where it was thought the KLA was operating or where refugees were gathered.
Finally, we followed a variety of rules designed to lower the risk of collateral damage, commonly defined as the unintentional or incidental injury or damage to persons or objects that are not military targets. The various rules included a temporary prohibition on the use of cluster munitions, the requirement for CAOC approval before attacking targets close to civilian structures, and restrictions on the type and color of vehicles that could be engaged.
Each of these rules had intentional and unintentional consequences for target identification and target engagement. ROEs calling for higher minimum made the pilots’ task of judging whether the vehicles on the ground were civilian or military more difficult. The border of a no-attack zone meant the enemy could move from imminent danger on one side of a road to sanctuary on the other side. Other attack restrictions on areas, munitions, and target categories slowed down the prosecution of attacks. During this waiting period targets could escape by moving to a sanctuary area, weather might move in and hide them from attack, or the waiting attack aircraft might run low on fuel and be forced to depart before receiving an attack clearance.
In the beginning of KEZ operations, the ROEs were quite simple: AFACs were their own attack-clearance authority after they had determined that the suspected object was a valid military target. CAOC approval was required only when the AFAC judged that the attack might cause collateral damage—usually to civilian buildings. During the course of the conflict, many additional target types were moved from the discretionary list to one that required CAOC approval. Later, geographic areas were defined, and targets in those new restricted areas required CAOC approval prior to attack.
AFACs often became frustrated when it appeared the CAOC was second-guessing the targets they had chosen to attack. For example, on the second day of KEZ attacks, near the southwestern border of Kosovo, I had located two large, darkgreen military deuce-and-a-half trucks, complete with curved canvas tops. They were parked just off an asphalt road on the north side of high terrain, northwest of Mount Pastrik. As I talked a flight of Dutch F-16s through an attack on these two targets, ABCCC called in-the-clear, on strike frequency, and asked whether I had positively confirmed that the target was military. I responded, “Yes.” The F-16s had missed on their first pass, so I directed them to make another attack. Apparently my answer to the ABCCC wasn’t sufficient for the CAOC. ABCCC then asked me to describe the target—evidently to personally confirm that it was military. Because we had already received AAA fire, I didn’t think we should loiter in this area. Barely able to maintain my composure, I described the trucks in detail. After landing I called our CAOC representative who told me that since JSTARS had not detected trucks in that area, the battle-staff director needed additional assurance on the target. In this case, JSTARS could not see the trucks because a high mountain blocked the line of sight between its radar orbit and the trucks. Likewise, the E-8’s moving-target indicator would not have highlighted the parked trucks.
We appreciated the CAOC’s concern for avoiding civilian nontargets. In a similar manner, I was particularly proud of our pilots’ strict discipline, which ensured that their attacks and the ones they directed avoided collateral civilian damage. If they were in doubt about the military nature of the target or the possibility of damaging a civilian structure, they brought their ordnance home and let the potential bad guys escape. To do otherwise might have cost the lives of innocent people and severely jeopardized the political support for KEZ operations. As CAOC “interference” became routine, I learned to control my blood pressure, jink, and talk to ABCCC.
The CAOC seemed poorly organized to act quickly when the ROEs required its approval. When an AFAC independently found a target in a restricted area or close to a village, it usually took 15 to 20 minutes to obtain approval. For example, using Lt Al Smith’s intel imagery, Maj Goldie Haun discovered what he thought to be a radar-missile launcher in an area south of Pristina. I flew there the next morning and found the small, wooded area just on the southern edge of a square village. From 15,000 feet I looked down through the trees with my binoculars and did not see any launchers. I did, however, see about 10 rectangular metal canisters stacked on a flatbed trailer in the center of the woods. They were about 15 feet long and the size of radar-guided missiles. I had two problems: the canisters were less than 100 meters from the nearest house in the village, and a low deck of clouds had begun to move in and would soon block my view of the target area. The trees prevented success when I tried to lock up the trailer with a Maverick, and the low clouds to the east ruled out an airburst Mk-82 attack flying on a heading parallel to the houses—the only attack axis that would minimize the risk to the occupants of the house. I still had my 30 mm cannon, which would work just fine. Moving off to the west to mask my interest in the area, I called ABCCC on secure radio to request approval to attack the trailer and had to wait 25 minutes for a response. During that time, the low clouds moved into the area around the trailer. I made several more calls to help expedite the clearance. I finally got the CAOC’s approval to attack, and with it came the direction to “use the gun and not hit any houses!”
The delay and the tactical direction were absurd and a clear violation of the principle of “centralized control and decentralized execution.” The expert on the weapon systems’ capabilities and limitations had been sitting in the cockpit, looking at the target, fully capable of making a real-time execution decision consistent with ROEs that centralized control had generated. During those 25 minutes, the low clouds moved in and obscured the northern end of the target area. Because the clouds limited my options, the only attack heading I could then use was south to north—pointed directly at the village. My bullets could have ricocheted off the trailer and into the houses. The policy to centralize execution decisions had caused the delay, which had allowed the weather to change, the Serb missile canisters to escape attack, and the missiles in those canisters to remain part of the enemy’s combat capability. Perhaps those missiles were some of the many that were subsequently shot at our pilots.
The CAOC’s actions indicated that its personnel were under colossal pressure to produce results and avoid collateral damage. They regularly interrupted our attacks—even those authorized by the ROEs. In their zeal to ensure we had the most up-to-date ROEs, the CAOC prompted ABCCC crews to remind us of what and where we could and could not attack. Unfortunately, these calls were sometimes made over the strike frequencies that Serbs monitored, allowing them to hear, “You cannot attack any targets within 10 miles of the Macedonia border,” “you cannot attack any trucks, civilian vehicles, vehicles painted white,” and other similar transmissions. The unintended consequences of these in-the-clear transmissions were that Serbs began moving their forces south to within 10 miles of the Macedonian border and started transporting their troops in civilian vehicles instead of APCs—we saw white charter buses everywhere.
Even when AFACs were directed to enter restricted areas to find specific CAOC-identified targets, they were still required to receive CAOC approval before they attacked. On one occasion the CAOC tasked Capt Jim Meger to find and destroy a Serb command post. After locating it he requested approval to attack. It then took the CAOC 20 minutes to approve the strike against the command post it had sent him to destroy. These delays not only allowed some targets to escape attack, but also put our pilots at greater risk. On this occasion, it took the Serbs a lot less than 20 minutes to locate Jim, draw a bead, and fire two missiles at him (see chap. 4).
Our frustrations grew. We were getting shot at while we waited for CAOC clearance to attack, we watched helplessly as Serbs in “civilian” vehicles burned villages, and we were denied clearance to attack enemy troops and equipment in the ROE-designed sanctuaries. I began briefing visiting dignitaries on our successes and frustrations when they passed through our base at Gioia. In late April I described our situation—from our point of view, warts and all—to two very distinguished visitors, Acting Secretary of the Air Force F. Whitten Peters and Gen John P. Jumper, commander of USAFE. I emphasized that, while we thought the current situation was unacceptable, we also understood there were probably some aspects of the ROEs and target-approval process that might justify the delays. So, we thought that informing us of those constraints could make the whole process more responsive and that we could better fulfill our responsibility for identifying valid targets. They listened carefully. General Jumper took a lot of notes, asked who had been working these issues for us at the CAOC, and asked if I knew whether Lt Gen Mike Short was aware of our problems. I told him that, every day, we passed our concerns to the lieutenant colonel who represented our unit at the CAOC. He replied, “Now you’ve got a four star.”
General Jumper suggested that I go to the CAOC to brief General Short personally. I thought that was a good idea and discussed it with Lt Col Walrus Heise, our representative. I expected to be summoned to the CAOC during the days that followed. A few days later our unit got a call to attend a conference at Tirana, Albania, with General Short and Lt Gen John W. Hendrix, the US Army V Corps commander. The purpose of the conference was to discuss employment options with the US Army Apache helicopters. We hoped to discuss the ROEs with General Short while we were there.
Colonel Thompson, Lt Col Coke Koechle, our two squadron weapons officers, and I flew as passengers in a C-21 from Gioia del Colle to Tirana on 1 May. We did not know that General Jumper had sent General Short an E-mail outlining our grievances with the ROEs and approval process. We soon discovered just how unhappy General Short was when he called Coke and me aside at the end of the Apache conference. Coke had worked for General Short at Headquarters USAFE, knew him well, and was not completely unhinged by the dressing-down that followed. General Short detailed for us—with incandescent clarity—that when our commander, a four-star flag officer (Vice Adm James O. Ellis Jr., AFSOUTH and NATO joint force commander) had determined the ROEs, he expected his field graders to understand and follow them, or seek clarification in appropriate channels. He did not need those same officers taking their gripes to another four star when they failed to understand or disagreed with the ROEs. “If you have a problem,” General Short said, “then come to Vicenza and see me.”
After it was over—and I realized that I hadn’t been fired—we all had gained a new appreciation for the source of General Short’s ire. He had foreign governments second-guessing his conduct of the campaign; he surely didn’t need the same thing from his own squadron commanders and line pilots. Nonetheless, we still hoped something good would result from venting our concerns and our pain.
Something did. Just days later, the CAOC adjusted the ROEs and replaced the 10-mile-deep no-attack border restriction with three zones: from the border to two miles, two to five miles, and five to 10 miles from the border. Attacks on targets inside 10 miles were now possible, and the probability of CAOC approval increased as the distance from the border to the targets increased.
Our AFACs took this change to heart and besieged the CAOC with requests to attack their favorite and previously protected targets. This deluge of AFAC requests generated a negative, unintended consequence. The CAOC began to feel the need to issue a clearance for all attacks—even for those targets where the existing ROEs did not require CAOC approval. Late in the campaign, for example, ABCCC sent Colonel Thompson and me to identify an active Serb artillery site that the US Army’s counterbattery radar had located about five miles west of Prizren. It was exactly at the coordinates given, six big tubes parked around a concrete pad facing south. They were not in a restricted area and were well clear of all civilian buildings. We confirmed the target with ABCCC (call sign Moonbeam) and directed that it move our fighters forward. Moonbeam said to stand by for CAOC approval. It was clear that the controller was reading from a checklist he had been directed to use. The radio conversations went something like this:
“Chili 11, Moonbeam, say distance to nearest civilian buildings.”
“Moonbeam, Chili 11, this target is not in a restricted area. I do not need CAOC approval to attack it.”
“Chili 11, the CAOC will not let you attack the target unless we get clearance.”
“Moonbeam, attacking this target will not cause collateral damage. Say reason for needing target approval.”
“Chili 11, we know the CAOC will insist on the information. We know we’ll get the approval faster if we have it ready.”
“Moonbeam, go ahead.”
“Chili 11, say distance and direction to nearest civilian buildings.”
“Moonbeam, the nearest civilian building is 400 meters to the west.”
“Chili 11, describe the building.”
“Moonbeam, the building is a farm building about 20-meters square.”
“Chili 11, Moonbeam, say type and location of the nearest civilian vehicle.”
(I paused to find one somewhere.)
“Moonbeam, the nearest civilian vehicle is a single car parked in the outskirts of Prizren, at least two kilometers from the target.”
“Moonbeam copies, stand by.”
We productively used the next several minutes (and gas) to locate some mortar pits farther to the south.
“Chili 11, Moonbeam.”
“Moonbeam, Chili 11, go ahead.”
“Chili 11, the CAOC wants to know if you can attack that target without causing collateral damage to the farm building 400 meters away.”
“I can’t believe this!” I transmitted privately to Colonel Thompson on the FM radio.
“Moonbeam, Chili 11, inform the CAOC that I can attack this target without causing collateral damage to a building 400 meters away.”
“Moonbeam copies, stand by.”
We finally got our attack clearance and proceeded to destroy all six of the big artillery pieces, using all the ordnance that was carried by another A-10 flight, a GR-7 flight, and us.
We began to accuse the CAOC folks of trying to FAC from their desks in Vicenza. This time, however, we took a different approach. We submitted a paper to the CAOC on ways to simplify the ROEs and reduce risk to the pilots, while maintaining the zero-collateral-damage safeguards. Col Stu Johnson, the CAOC chief of operations, sympathized with our request but reasoned that the alliance had about all it could take of changing ROEs. He said we’d have to do the best we could under the circumstances and promised to try to accelerate the approval process.
That he did. In the last week or so of the campaign, as the KLA stepped up its pressure on the Serbs, the delay in receiving our target-attack approval was significantly reduced. As noted earlier, this close control by the CAOC violated a basic tenet of air and space power, which calls for centralized control and decentralized execution. Just as centralized planning and control is critical to the proper application of airpower, our Air Force Basic Doctrine (AFDD 1) reminds us that so is the concept of decentralized execution—“delegation of execution authority to responsible and capable lower-level commanders is essential to achieve effective span of control and to foster initiative, situational responsiveness, and tactical flexibility”(23). We experienced centralized control and execution. Consequently, our situational responsiveness and tactical flexibility suffered.
I want to emphasize that our discussion of ROEs, their application in OAF, and the influence that they had on our combat effectiveness is not intended to air gripes or to disparage the fine professionals who worked in the extremely difficult circumstances at the CAOC. When we complain about problems with “the CAOC,” we do not fix blame but merely identify the CAOC as our immediate contact in the command and control conduit. General Short, Colonel Johnson, and all the battlestaff directors who sat in “the chair” during the nonstop combat operations probably saved our unit from immeasurable pain on numerous occasions—and we never knew it. We did know that we had their support throughout OAF, and we gratefully and sincerely applauded their leadership after the NATO victory.
Nevertheless, we hope that we will never again fight a war with similar ROEs. We had no knowledge of the agendas of, or the interaction and dynamics between nations, heads of state, and our leaders (SECDEF, SACEUR, AFSOUTH, CFACC, and the CAOC). Therefore, we are unable to comment on what the ROEs could or should have been. We did understand that any targeting mistakes that took the lives of innocents could easily have unraveled the international support for the campaign. However, it is our conviction that the war was unnecessarily lengthened and that innocent Kosovar lives were lost because we did not have streamlined ROEs which supported decentralized execution. Our recommendations for the future:
• Push responsibility and authority to the lowest possible level—trust those who are well trained to do their jobs.
• Develop unambiguous target-identification procedures that allow for decentralized decision making at the AFAC level.
• Consider calling a “time out” for a day or two to gather the experts and decide on a best way to go forward when the ROEs adversely affect combat operations.
• Ensure that all war fighters fully understand the ROEs.
It was a bright, clear morning. We approached our target from the southwest and visually picked up the revetted APC while we were still more than five miles out. We circled south to compare and confirm the target with the photo. I set up and executed a tip-in from the east. It sounds so simple, and sometimes it was if I could ignore the fact that I was being shot at. To get to that point usually took quite a bit of work.
This story really starts one and one-half days prior to that bright, clear morning. Our A-10 squadron was flying from Gioia del Colle AB, where several Italian squadrons, some British GR-7 Harriers, and we were based. The GR-7s occasionally flew with a reconnaissance pod and took pictures of areas of interest. The Brits were nice enough to share those photos with our intel folks. The photo for this mission came in one night while I was working a 12-hour top-three tour. Those of us who pulled top-three duty, especially the night tours, were not thrilled with the task. But this night we were lucky. Pictures of an APC and several other interesting targets came in just after dark, and I briefed them to all the pilots on the morning schedule. At the end of my tour, I went to bed knowing that the guys walking out the door to fly had several good targets. Like everyone who pulled a night top-three tour, I was given a day off to sleep and then scheduled in the first flight of the following day.
I arrived in our operations area at about 0230, sat down for the intel brief, and was handed a couple photos and the usual target list. To my surprise the same photos of APCs were still in the package. Intel said no one had reported hitting the target, so they were passing it out again. The weather had been poor in the area where the APCs were located, and the AFACs had looked for targets in areas with better weather. Although day-old target photos are often worthless, my wingman and I decided we would take a look since we would be the first planes in Kosovo and we hadn’t worked that area in many days.
When I studied the targets, including the APCs, I plotted them all on a large Joint Operations Graphic (JOG) map, drawn to a 1:250 scale, that I used in the cockpit to work the entire Kosovo area. I plotted the most promising targets on a smaller 1:50 scale map so I could get a good feel for those areas and accurately determine target coordinates and elevations. (While these maps are normally referred to as 1:50, 1:250, or 1:500, their scale is actually 1:50,000, 1:250,000, or 1:500,000.)
The elevations had to be converted from meters to feet, and the coordinates put in several formats to satisfy the various navigation systems used in the different aircraft we might control. We often did this while circling over a target, but it was much easier to do ahead of time on the ground. I then used a computer that contained photomaps of the area in various scales. With this I could look at each target area from different altitudes. Although neither the photos on the computer nor the maps were current, they gave us a good perspective of the surrounding area—the lay-of-the-land, so to speak. With this situational awareness, we could choose an attack axis that would give us the best chance of success while limiting damage to anything but the target. With our planning complete, I briefed my wingman on the plan to attack this target and search for others in our assigned area.
The flight to the tanker was quiet, and I was encouraged that the weather would cooperate. Our tanker track was over southern Macedonia, and from there I could see that the sky over our target area was bright and clear. We got our gas, departed Macedonia, and proceeded northeast into southeastern Serbia above the Kumanovo Valley. While en route to the tanker, I had tested my two Maverick missiles, and now I reapplied power to start what I hoped would be the last warm-up cycle for at least one of them. We headed for our target, which we expected to find just a few miles west of Vranje. The larger target area was easy to see from far off, and the APC came into view as we drew near. I checked with my wingman to confirm that he saw the target, and then I started to circle south to make a final check to ensure that we were at the right spot. I put the jet’s head-up display (HUD) in the proper mode for attack and checked the other settings. The scene below was surrealistic.
A small single-story house was located on a small dirt lane that connected to the valley’s main road. The backyard of the house, which the sun was just beginning to hit, was fenced and contained the usual scattering of possessions one has become accustomed to expect. In the front, a dirt driveway led from the road to a small attached garage. I could not detect any movement in the area—not unusual that early in the morning. This was a scene that could have been repeated anywhere in the world, except for two things. This scene matched my target photo exactly, and just 15 to 20 feet north of the driveway was a recently dug revetment in which sat an APC—not your average family roadster. To arm my weapons, I flipped the final switches. I got a good picture and the green lights that indicated the missiles were ready. I checked the position of my wingman and set up for a tip-in from the east. I made one last check of my switches, rechecked the HUD, and rolled into a dive for the attack.
The IIR Maverick picture that I saw in the cockpit was about what I expected. Different parts of the house indicated various temperatures. The rear wall and one window were cold, while another window and the chimney were hot. Most importantly, the APC was cold—very cold. It had not been run that night. Our observations up to this point indicated that lucrative targets that didn’t move were often decoys. However, I was confident that was not the case this time. I slewed the Maverick seeker to the target and got an immediate lock. I waited a moment for the missile’s brain to signal me it could distinguish the target from the background. It was ready. I took one last look through the HUD and pushed the pickle button.
The sound of a 500 lb Maverick coming off the LAU-117 launcher is exciting. I shot many of these missiles during the war and still was amazed by each launch. Once it was on its way I started to pull out of the dive. I had lost less altitude than expected during the attack since the missile lock-on process had gone so quickly. The first few seconds of the missile’s flight are usually indicative of whether or not it will hit the target. This one flew as expected, and I monitored it until it disappeared below the nose of my jet. I then looked up to locate my wingman to make sure that I would not climb into him. He was where I expected, watching for possible ground fire. I continued to maneuver in the climb although neither of us saw any reaction from the surface. I looked back at the target as it sat there doomed. A Maverick missile never arrives early or even when one expects it to hit. No matter how long I thought the time of flight would be, I would still have to wait when I looked at the target to see the impact. I waited—then it was over. I saw a bright flash of orange and yellow flame followed by a growing cloud of dark gray smoke. A small shock wave emanated from the revetment in a growing circle and quickly dissipated. We saw small flashes at the base of the smoke plume—probably detonating ammunition inside the APC. This was no decoy. The house was untouched.
We proceeded southwest down the valley to move away from the area we had just attacked. The Kumanovo Valley was full of revetments, trenches, and other obstacles the Serbs felt would help them if an invasion came from the south. All of them we saw that day were empty. Forty minutes later, as we departed the area to get more gas, the APC was still burning.
We pilots lived by ROEs. They told us what we could strike and what we could not. We had to become familiar with a stack of documents several inches thick in order to fly missions over Kosovo. The ROEs were a small—but important—portion of these documents. Early in the conflict the rules were fluid and changed quite often for many different reasons. After a couple of weeks into the war and a major shift in focus, the ROEs settled down and remained fairly constant except for a few silly, and often politically driven changes. During the last couple of weeks of OAF, the rules again began to change frequently. It was not uncommon when we flew on two consecutive days to use different rules on the second day—sometimes they even changed during a mission.
In Kosovo and during the period from a few weeks after the war started until a couple weeks before it ended, we AFACs were given a fairly free rein to attack whatever military targets we deemed worthy. It was a bold move by Lt Gen Mike Short, commander of AIRSOUTH in Vicenza, Italy, to relinquish control of tactical decision making to the lieutenants, captains, and majors who were on scene and had the best available information. While it was a big responsibility, it was also a godsend and allowed us to kill many more targets than would have been possible under different rules. Our freedom to observe the situation, apply the ROEs, and make decisions was taken away sometime during the last couple of weeks of the war. That change led to the most frustration I experienced during any mission of the conflict.
My wingman and I were on an AFAC mission. We were proceeding from the tanker track into the KEZ when we got a call from a flight of A-10 strikers. They were searching for targets about five or 10 miles north of Prizren, and the flight lead said he had found what he thought were four APCs. Someone next to the APCs had just fired a handheld SAM at his aircraft. He was running low on fuel and wanted to hand the target off to someone before he departed. I led my flight to the target area to take control. A thin, wispy trail of smoke still indicated the missile’s flight path, from where it had been launched to where its motor had burnt out. After receiving a good hand-over brief, I used my binos to check out the area. I had some bad news and some good news for the departing flight lead. The bad news—he had found only one APC; the good news—the other three were tanks, and relatively modern models at that. What a find! I began the coordination process for an attack.
We were near the end of the conflict, the ROEs had changed, and we were required to get permission before we could direct attacks on any target. We had operated with these rules for the past couple of days and found that the approval process could be either quick or take many minutes. We soon learned that the best course of action was to find a target, call in the request to strike, and then search for other targets while waiting for approval. At any given time we would have several attack requests in the process of being approved or denied. When we received permission to strike, we had to figure out which of the targets had been approved and then relocate it—hoping to find it where we had last seen it. I was sure that the details of these targets would cause the process to be more responsive. It was a large group of tanks, in the open and far away from any man-made structure. I was excited and decided to stay overhead to monitor their activity while waiting for approval.
I had submitted my target to the ABCCC guys orbiting south of the border, who passed it to the CAOC in Italy. The CAOC, I was told later, had to phone someone even farther removed from the fight. That person would look at a map and decide either to grant or refuse permission to attack. That person was probably away from his desk when the call with our request came in; my wingman and I waited. There was activity down below. The tanks were about 100 feet from a paved road, and a small white car was parked on the side of the road. We suspected that the occupants of the white car were part of the Serb forces associated with the tanks. The Serbs’ favorite mode of travel appeared to be stolen Kosovar cars. Their early OAF experience had taught them that civilian vehicles survived when convoys were attacked. At about this time a small civilian flatbed truck arrived from the town located about a kilometer down the road. It stopped near the car, and its occupants proceeded to unload whatever it was carrying.
After five or 10 minutes had passed, I radioed the ABCCC folks to see if they had any news. I told them of the new developments and asked that they check with CAOC to see if more information was needed. From the radio traffic that we could hear, KEZ didn’t sound very busy, but one never knew what might be happening in the CAOC. While we waited, I continued to prepare for the attack. I determined the target coordinates and elevation, and checked the winds at our attack altitude. I rechecked our location on a large-scale map and planned where I could hold strikers, if and when they arrived. I wanted to avoid holding a striker directly over a known radar SAM area—that was poor form. I made sure my flight would be ready to attack when the word came down. We spun up our weapons and briefed a plan of action. And we waited.
A flight of A-10 strikers checked in on frequency. They had been released by their AFAC in the west since he did not have any targets for them and had heard that we had tanks. They were each carrying two CBU-87 combined-effects munitions, which were perfect weapons for my group of four armored vehicles parked right next to each other. I brought the strikers in and held them south of the target, briefing them on the situation and asking how much time they had. The flight lead said they had enough fuel to stay on station about 20 to 25 minutes. I told them the approval process was taking longer than usual and asked that they conserve as much fuel as possible. To prepare them for their attack, I deconflicted our altitudes and had them fly over and take a look at the target. This would enable them to attack as soon as the approval came down. We confirmed that we were dealing with three of the newer tank models the Serbs had in Kosovo. We discussed attack plans, munitions placement, and local threats. I then called ABCCC for an update. And we waited.
I sent the strikers south a few miles to resume holding. I scouted the surrounding area for more targets but didn’t stray more than a mile or two from the tanks. It had been over 30 minutes since we requested approval to strike, and I was beginning to worry about our fuel state. We could spend between 45 minutes and an hour over Kosovo. This relatively short onstation time would leave us with enough fuel to fly to the tanker track, attempt refueling, and—even if we could not refuel—still be able to return to Italy. We were conserving fuel while we were waiting, but how much longer would it take? Movement on the ground brought me back to the main target. One of the civilian vehicles was departing, and, as I flew overhead, it appeared that the armored vehicles were starting their engines. I excitedly called ABCCC to say it was now or never and told them to tell the CAOC that these vehicles would soon be gone. We waited a few minutes more.
Then things started to come apart. The lone APC moved out onto the road and sat there for a few moments. The three tanks started to move in the opposite direction, through the fields, and deployed in a column formation. I called ABCCC with my last message: “They’re moving, they are moving. I’m losing this target.” It took a few moments for them to answer, and I assumed that delay was caused by a call to the CAOC. Finally, 40 minutes after my first request, I got what I wanted—clearance to attack. The tanks were now moving slowly through the fields, but the APC was speeding down the paved road towards the nearest town. I decided that I would try to stop the APC first and then bring the fighters up from the south to deal with the tanks. I was sure the tanks could not get far in the time it would take to make one Maverick attack on the APC. It was probably less than a minute from the time I got permission to attack till I rolled in on the APC. I had the right weapons mode set, Maverick video on the screen, and green ready lights. I was ready to kill something.
As I rolled out, pointed at the APC moving down the road, I looked at the TV screen and watched the picture change from crisp and clear to a wavy-lined mess. I couldn’t believe it. I thought maybe just the one missile had gone bad and had thrown the switches that called up the missile on the other wing. The picture there was just as bad—the 1960s technology in my TV monitor had failed; maintenance guys say this has been a problem for years. The Maverick pass had taken me too low to be able to switch to the gun, continue the attack, and recover before busting the ROEs’ minimum-altitude limit. However, I had developed plenty of speed during that long dive and was able to pull the plane up to near vertical to gain altitude while I set the switches for a gun attack. I was still below the minimum altitude that I generally used for starting attacks—especially for consecutive attacks in a handheld SAM area—but I was mad. I pulled the jet out of its steep climb and pointed its nose at the APC rolling down the road. It was moving faster than any target I had ever tried to attack (moving targets are hard to come by in training). I aimed in front of the APC by a distance I thought would work, and planned to walk my rounds down the road to make sure I got some hits. Because each round from a 30 mm Avenger cannon weighs three quarters of a pound, it doesn’t take many hits to destroy an APC. I squeezed the trigger and heard the gun rotate—something was wrong—no rounds came out. I held the trigger down for a while longer, thinking there might be a blank spot in the shell train. I got nothing and was now far lower than I wanted to be. I pulled off target—again trading airspeed for altitude. I was beside myself. I rechecked switches—all were correct. I could not believe what had just happened. All the effort that had gone into this sortie and all the time it took to get approval for the attack would not matter, it seemed, for today was not going to be my day to kill armor.
I had bled off too much energy, so it took a while to climb back to altitude. That gave me time to cool off and decide that, even though I couldn’t get a kill today, I could make sure that others would. By this time the APC had escaped. It made it to the safety of the town and was off limits. I started looking for the three tanks that I had last seen moving slowly to the northeast. They were gone. I cursed myself for not assigning the job of watching those tanks to the A-10 strikers holding in the south. Too late now—I had to find them. I did not think they could have gotten far, but the area to the north and east was heavily wooded. I searched for a while before finding one. It had pulled about 20 feet off a dirt road and into some trees. I called the strike leader to bring his flight forward. He told me they were low on gas and had only a little time before they had to depart. I quickly deconflicted our flights and brought him to a point over the target while I worked out a new target elevation. It took a few minutes for the strikers to get their eyes onto the target. They were near their bingo (required recovery fuel), so I told them that my wingman and I could cover their attack. That would allow them to roll in, one after the other, and save time. The leader agreed. He rolled in from the west, and his CBU pattern impacted the ground just short of the target. I passed the correction to the wingman as he rolled out in his dive. His two CBU cans also hit a bit short. The tank appeared to be just in the edge of the bomblet pattern. This was certainly not our best day.
As the strikers departed, I decided we had just enough gas to allow my wingman to try a Maverick attack on the target. I had him try passes from several directions, but the trees prevented him from getting a good lock-on with his missile seeker. We were now bingo and had to depart the area. It was a long, quiet flight home—my most frustrating sortie of the war. We had started with one of the best targets we had seen in a long time and ended up with only the possibility of having damaged one tank.
There were numerous failures. Two different weapons systems had failed on my jet and I had failed to have striker flight track the three tanks as they headed off to the northeast while I attacked the APC. Nevertheless, the biggest failure of all was a process that required us to wait more than 40 minutes for approval to attack a lucrative target located far away from any civilians or buildings. The ROEs that had helped us for much of the war had now—for mostly political reasons—been turned against us. On this day they really bit us.
Humor is important. It makes us happy and rejuvenates the mind. We worked hard during the war’s 70-plus days, and our minds and bodies were tired. I saw and heard several things that now might seem insignificant, but at the time they were quite funny. One such event, which struck me as very symbolic, said, “This is a team effort; we will make this war work.”
A normal sortie profile includes taking off, flying to a tanker and refueling, working in Kosovo for about 45 minutes, taking a 45-minute round-trip flight to a tanker for more fuel, and then working another 45 minutes in Kosovo before going home. Although we often used the same tanker track for both refuelings, we might refuel from two different tankers. It was also normal for us to launch on every mission with all the tanker call signs and air-refueling track frequencies, so that we would have the information if it became necessary to switch to a different tanker while in flight. The tankers launched each day with the call sign and the amount of fuel they planned to off-load to each of their scheduled receivers. This enabled them to manage their fuel and deal with unexpected events. The plan was rather simple, flexible, and usually worked well. Occasionally, however, the refueling tracks became little more than controlled chaos.
On a rather stormy day about halfway through the conflict, my wingman and I were flying an AFAC mission in Kosovo using the call sign Ford 11. As I hooked up to our tanker at the beginning of the mission, I was told by its crew that, even though they were supposed to refuel us again during our second visit to the track, they would not have the fuel to do so. Most pilots like to take on a few extra pounds of fuel when the weather is bad; it seemed that today, because of those extra pounds, our tanker would come up short. The tanker pilot said that he was coordinating with the CAOC for a replacement tanker. I was not concerned since this was not the first time this sort of thing had happened. We took our gas and headed north to Kosovo.
We had some luck: the weather over Kosovo was better than that over Macedonia, where the tanker tracks were located. Eventually we had to depart Kosovo to meet our second tanker. We noticed more radio chatter than usual as we checked in with the area controller. This was to be expected with bad weather, but today the controller was also hard to hear. After trying for several minutes, I was able to get one good exchange with the controller. He said that my new tanker’s call sign was Esso 73 and that I would refuel on my original track. That was the last successful contact I had with the controller. That meant we could not have him give us vectors to the tanker, which created a problem since the A-10 does not have an air-to-air radar and the visibility prevented my acquiring the tanker visually. My only other method for finding the tanker was to use my radio’s direction-finding capability. That required my being in radio contact with the tanker. While en route to the track I tried all the frequencies but could not contact the tanker. We were now about to enter the track, and our remaining fuel dictated that we either start taking on fuel in the next few minutes or turn for home.
I entered the tanker track and found some clear air in the western end. The tanker pilots had a fair amount of discretion with regard to where they flew in the large tracks—the good ones tried to stay out of the clouds. Lucky for us, most tanker crews were good. As we were about to turn for home, I saw a tanker fly out of a wall of clouds on a heading that I could intercept. We turned to intercept it and started calling again on all frequencies for the track. Normally we didn’t talk to the tankers prior to refueling, but we were always on the same frequency and knew each other’s call sign. While we were still a mile or two from the tanker, it started a turn towards us, which helped my rejoin geometry and kept us from reentering the clouds before we could join on its wing. We were still about a half mile from the tanker and closing when I saw its flaps come down and the refueling boom lower. This was a good sign. I slid behind the tanker and stabilized next to the boom. From there I could see the boom operator settle into position.
Once connected to the tanker, I was able to talk to the crew on the intercom system. I felt a bit silly but asked, “Hi fellas, are you Esso 73?” As luck would have it, they were, and they asked if I was Ford 11. I guessed they had not received any help from the controllers either. This little exchange made me laugh. In spite of all our planning, our ability to continue with the rest of our mission came down to my plugging into the first passing tanker I saw and asking if it was ours. I then asked the most important question: “Do you have enough gas for two A-10s?” But as I was asking, I looked down and saw that they were already pumping, bringing out an even bigger grin. These guys knew why we were there, how important it was, and got right to it. This was how a war should be run. I had a smile the rest of the day. I took just enough gas to stay above my turn-for-home fuel. I then asked for, received, and passed the tanker’s radio frequency to my wingman. I cleared him to drop down and fill up. When he finished, I got back on the boom and took the rest of my gas. It may not sound like much now, but I chuckled about it all the way back to the border. In peacetime, I rarely got near a tanker unless I had talked to the controller, he had cleared me, and I had checked in with the tanker. On this day, the tanker crew didn’t care that we weren’t talking to them or even who we were. They had a job, and they did it. This war—like all others—was a team effort by all the little guys at the bottom—the ones who made happen those things that the people at the top talked about.
Late one afternoon in mid-May, I flew an AFAC mission in southeastern Kosovo. Thunderstorms had built up in the area east of Urosevac where I had planned to spend my second push for the day, so I took my wingman further east into the Serbian Kumanovo Valley. On the whole we spent less time in that valley than we did in Kosovo. Old Milo had been preparing a layered defense within the valley in reaction to the possibility of an invasion. We thought an invasion was unlikely, but if he was willing to provide us targets, we were willing to provide him with incoming ordnance. The AFACs who flew there usually had good luck.
As we entered the valley from the south, I started looking in and around the fortifications at the southern end of the generally north-south valley. The Serbs had done more work since the last time I had visited. During my previous trip, I had stayed at the northern end of the valley near Vranje. My plan was to start in the south, head north, and search for targets as we went. The sun was getting low, and the shadows being cast by the valley’s western mountains on its southwestern floor were growing. Because the northern half of the valley was oriented northeast-southwest, the shadows up there would not cover the valley floor as quickly as they did in the south. We had plenty of fuel but would be able to stay only as long as we could see targets in the dusk. By working our way northeast, we could use the last bit of available sunlight.
Dawn and dusk, while the sky is bright and the ground dark, are tactically the worst times of the day for a fighter to strike a ground target. The darkness on the ground reduces the effectiveness of the Mk-1 eyeball, and many night-vision devices are not operable because of the bright light at altitude. Unfortunately, the folks on the ground have no trouble seeing an approaching fighter. These factors usually result in high risk and little success for those who attack targets under these conditions. We flew anyway to keep up the pressure on the Serbs. My wingman and I continued searching up the valley. We stopped once down south to check out a square object up against a house, but it was too hard to identify (ID) in the shadows. I decided to go directly to Vranje since there was still some light on the ground there.
We checked out an area about eight miles west of Vranje where I had found some targets several weeks before. Finding nothing there, we moved in closer to the city and spotted movement northwest of the city. I decided to investigate and could not believe what I saw once I arrived overhead. It was the largest gathering of nonrefugees I had seen during the war. On the darkening landscape I counted well over a dozen military vehicles, at least five of which were APCs, and 75 to 100 troops in three clusters. They were only about a mile out of town in a large field that had been chewed up by the vehicles. My first inclination while I circled overhead was to start bombing immediately. Then I saw some green vehicles parked in the area of the APCs that I could not positively identify as military. I did not want to make a mistake with a group this large. I needed help.
I knew there were some Navy F-14 AFACs working just west of us in Kosovo. Ten years ago, I would have laughed at the thought of the Navy using F-14 Tomcats as AFACs, but then—10 years ago—the A-10 was also headed for the boneyard. The Navy upgraded the F-14 when it was assigned the AFAC mission. The fighter had acquired a low-light-capable reconnaissance pod whose capability I now needed. I switched to the eastern-Kosovo working frequency, called the call sign fragged for that time slot, and got a quick response. They agreed to help when I asked if they had time to look at a target and learned that they were not having any luck finding targets where they were. I passed the target coordinates in a format they could use, and they were on the way.
None of our pilots had worked with the Navy F-14 AFACs before the war and were a bit leery of them the first few times they appeared in Kosovo. We soon discovered that while they did not have much experience with the AFAC mission, they weren’t bad. A few had yet to master the requisite skills, but most were doing fine—and some were quite good. I recognized the voice of one of their better AFACs in the jets headed our way. I deconflicted our flight paths, briefed him on the target below, and told him what I needed. The shadows now covered most of the valley floor. With this lighting I would normally head home—but this target was too good to pass up. From our orbit overhead, I was able to talk the Tomcat crew’s eyes onto the targets below. He confirmed that it looked like lots of troops and APCs but was also unsure of the few green vehicles. He decided to make a diving pass near the target to give their pod a closer look. I maneuvered my flight into a position to cover the F-14s as they made their run.
The Tomcats completed the pass and rather excitedly stated they had good pictures of the APCs and confirmed that the few green vehicles were also military. They said there appeared to be more troops in the gathering than first thought. We were running out of both daylight and fuel as I coordinated with the F-14 lead. My wingman and I would make three passes each and then leave the rest to them. He agreed, and they set up an orbit north of us. We made all of our attacks from the south because of some clouds that had entered the area. With the F-14 flight covering us, I reduced the usual attack spacing between my wingman and me. We coordinated our targets and each made two quick Maverick attacks on the APCs. On our third and final pass we each dropped our four 500 lb bombs, offsetting the impact points of our four-bomb sticks so we could cover as much of the area as possible. Anyone dumb enough not to have fled after our four Maverick attacks would not be pleased with the 500 pounders. They were set to explode 15 feet above the ground, and their blast and fragmentation kill mechanisms made them deadly weapons against troops and soft-skinned vehicles.
After our last pass we departed the area for home. The F-14 lead told us that, although our attacks had been right on target, there were still plenty left for his flight to attack. As we departed he was calling over the F/A-18 strikers he was supposed to work with so they could deliver even more ordnance to the area. It was too dark when we left for me to make a valid damage assessment, but I would come back in the next day or so. When I did I was amazed at what I saw. The Navy must have really worked that area over that night—it was a mess—pieces of military vehicles everywhere. I visited the area in the weeks that followed and could see where the Serbs had attempted to clean up the damage, but the ground still bore witness to the beating it had taken. I was puzzled: why, during combat, would anyone without air superiority gather so many troops together in daylight. In war, stupidity rarely fails to get its reward.
I have several memories of my time in Kosovo. Most are good; some I would rather forget, but all of them are slowly losing their details with the passage of time. Still, one of my most vivid memories involves nothing more than a simple truck.
Maj John “Scratch” Regan, our 74th EFS commander, was on my wing. We had flown some rather effective sorties together during the preceding couple of days. On this particular occasion we had both dropped our bombs on some meager targets earlier in the mission. I say meager because we had direct hits with two tons of Mk-82s with limited secondaries. We were patrolling the western side of the province, just west of Dakovica (D-Town) when Scratch called over the FM radio that he saw a mover. Even during the waning period of OAF, we occasionally saw movers on the roads.
An interesting thing happened during the opening days of the conflict that had a direct impact on our ability to interdict movers. ABCCC had heard someone’s description of a convoy of assumed military vehicles that included a white car. ABCCC then said—in the clear over a nonsecure radio—that, in effect, we couldn’t hit any white vehicle. As a consequence, during the next few days it appeared that every can of white paint in Kosovo was used on any vehicle that could move.
It didn’t take Scratch long to talk my eyes onto the mover. It was immediately apparent that this target was different—big, green, and fast moving. More specifically, it was heading deeper into bad-guy land. I made one low pass behind the mover, and even before I grabbed my 15-power binos, I could tell it was a big truck. With the optics I could clearly see it was a green deuce-and-a-half truck with an open back end covered by a solid-brown object. I couldn’t quite make out what was in the back, but I figured it was just a dirty-brown tarp. I immediately brought the nose of my airplane back up to stay behind the truck and to gain back some altitude.
My next move is one that I will regret for quite a while—I called ABCCC for clearance to strike this target. There were rules about what we could and couldn’t attack, and, unfortunately, trucks weren’t approved. ABCCC’s initial reply was to “stand by.” So there I was, carving a lazy figure eight at about 14,000 feet behind this truck with Scratch slightly high and behind me. I was all set to roll in from behind, lock a Maverick onto this truck, and blow it to smithereens. A Maverick costs several times more than a truck, and it was certainly not the most cost-effective weapon choice in this situation. However, it would be most spectacular and have the highest kill probability—and the guy driving this truck deserved it.
I was appalled that someone would have the audacity to drive his truck at 80 miles per hour down an open highway in the middle of the day. At the time I was merely shocked; in retrospect I am almost disgusted. I now think about all of the FACs from an earlier time—the Mistys, the Nails, the Coveys. Those guys flying the Ho Chi Minh Trail in their F-100s, O-2s, and OV-10s just hoping to catch a glimpse of a truck through triple-canopy jungle. There I was watching—just watching—this Serb drive down the road like some teenager trying to see how fast his dad’s pickup will go.
While I waited several long minutes for approval, the truck made it into D-Town and began winding through some of the lesser streets (the ones without the rubble of destroyed homes). Simultaneously, a voice came over the strike frequency and asked if I could positively identify my target as “armor.” Laughingly, I again explained that it was a truck, a military truck, a deuce-and-a-half truck! Immediately, I was informed, “Do not, I repeat, do not strike that target.” This was followed 25 seconds later by another voice telling me not to engage the target. Reluctantly, I broke off the attack.
That was my last sortie of Allied Force. Scratch and I went on to find some other targets. I wound up earning a rather important medal for this particular excursion into combat and someday hope to show it to my grandson. I will, however, remember that truck more than anything else we did that day. Even now I often wonder what that driver is doing. Perhaps he told his buddies—at some cheap Belgrade bar—about the day he gave the finger to two A-10s in broad daylight, or how he thinks we were too scared to attack him. Perhaps he never even knew we were there. I knew that with one push of my right thumb, several hundred pounds of missile would have slammed into the top of that truck and cartwheeled it into a fiery heap. It’s that image—seen only in my imagination—that I will remember the best.
I joined the squadron at Aviano on 28 March and was immediately tasked to go to Vicenza to be one of the squadron’s CAOC representatives. Eight days later I returned to Aviano and flew my first sortie. After a couple of weather days, I was selected to be a member of our advance team and was sent to Gioia del Colle.
I will not soon forget my first opportunity to fly after the squadron moved to Gioia. It was 14 April 1999 and only my second sortie of the conflict. I was the flight lead, Lt Col Surgeon Dahl was my wingman, and our call sign was Cub 31. The first half of our AFAC sortie in the eastern half of the KEZ was uneventful. We were flying above a broken deck, which made our search difficult because it was hard to see targets around and through the clouds. After completing our first vul period, we went to the tanker; by the time we returned the weather had cleared significantly. I checked in with Maj Joe “Dice” Kopacz, the other AFAC on station, and learned he was busy dodging AAA. When he finally got a chance to talk, a couple of expletives were quickly followed by a recommendation not to come over to his area. He had found some targets along the Kumanovo Valley road and had, in his words, “stirred up a bees’ nest.” He suggested I give it a little time to cool down in case I wanted to come over and take a look. I took his recommendation to heart and continued to search for targets south of Urosevac.
Shortly after Dice departed, Moonbeam (ABCCC) transmitted that Bear 21, an F-16 AFAC in the western half of Kosovo, had found and was preparing to engage a large convoy of more than 100 vehicles. Bear 21 had requested that Moonbeam send any available fighters, and I was not about to miss an opportunity to engage some targets. I called Bear 21 and told him that my flight was holding in the eastern half of Kosovo and that we were available as strikers. Since I was fourth in his stack of fighters, I continued to look for targets on my side of the KEZ.
While waiting, I spotted a large truck on the road below me but couldn’t immediately confirm whether it was civilian or military. I used my binoculars to investigate, but I still couldn’t get a good look because he was moving down the road. I decided to try to stop the truck by firing a couple of rockets. I had previously observed rockets being fired on a range and knew they made a lot of noise but created a very small amount of frag or blast. I estimated the distance in front of the moving truck where I needed to aim so that the impact would occur far enough in front of him to get his attention but avoid hitting him with frags. I set up, flew the pass, and launched two rockets that bracketed his position. The truck immediately came to a stop, and I tried to get a closer look. It appeared to be a deuce-and-a-half military truck—but I wasn’t sure.
While I was trying to confirm the truck’s identity, I got another call from Moonbeam. Apparently someone at the CAOC had questioned the composition of the 100-vehicle convoy target being attacked in the western half of the KEZ. Moonbeam wanted an aircraft equipped with binoculars to check out the convoy to make sure it was a valid military target. I contacted Bear 21 to get the location of the convoy. I was not sure that I had sufficient fuel to fly all the way to his target, observe the convoy, and still make it to a tanker with the required reserves. I was required to have enough gas when I reached the tanker to be able to recover home in case the refueling was not successful.
I communicated my concern to Bear 21 on the common strike frequency. He pushed me to a different radio frequency, checked us in, and said he was sure that he was striking a good target and that he really just needed me to come on over and check it out. I figured it would take me at least 45 minutes to get to the tanker and back, during which time the attacks on this lucrative target would be halted. I had refueled many times before without any problems. I considered the consequences of having a malfunction today and decided that even if we had difficulties, we would still be able to land at a friendly airfield—just not at our home base. We would get refueled on the ground, and with a fast turnaround we could quickly be on our way home. We had Moonbeam confirm that tankers would be available and proceeded to Bear 21’s target.
The convoy, which had been heading to the southeast, was stopped near the town of Dakovica, on the main road to Prizren. At 20,000 feet and with my naked eyes, I could tell the target was not military—at least most of it wasn’t. It consisted of lots of colorful vehicles. I could see blues, reds, and a lot of bright silver from bumpers reflecting the sun. I let Arden know that I was going heads-down to use the binos. I could then see that a number of civilian cars and tractors had pulled over to the side of the road. I also noticed a couple of larger vehicles leaving the area to the west. One of them was a large charter bus with some colorful paint marks down both sides. There were one or two other west-moving vehicles whose identity I could not confirm before they entered the town of Dakovica and disappeared.
I then noticed a strange-looking pattern south of the road and took a closer look. I identified it as a crowd of people who had left their cars and tractors and had run away to a safe distance. They were waiting in the fields for their vehicles to stop exploding. I immediately transmitted that the convoy was definitely civilian. Bear 21 ensured that all attacks and weapons expenditures were stopped. The CAOC quickly closed the KEZ and sent everyone home. We had been trying to extend the target-area time to prosecute this target—that was obviously no longer required.
Our day wasn’t over. We flew to the tanker to get a little fuel before going home. I took gas without a problem, but Arden started getting a lot of fuel spray from around the refueling receptacle, the location where the tanker’s boom connects to the nose of the aircraft. He disconnected, reconnected, and tried again with the same result. Since the A-10 tech manual recommended that refueling be discontinued in the event of fuel spray due to the potential danger of fuel pooling in the electronics bays, Arden ended up taking on only about 500 pounds of gas. We did the math to figure out if we had enough fuel to get home. It didn’t look good for flying straight home, but we could make it to our divert base. While we waited for the jets to be refueled at Brindisi, we learned that CNN was already covering the attack on the convoy. Immediately after landing at Gioia del Colle, Arden and I wrote our reports and submitted them to our CAOC rep at Vicenza.
It was not a glorious moment for either the Air Force or NATO. However, we thought it could be a critical lesson learned and a memorable chapter in Hog history. There are plenty of things technology can do, but experience and training in the basics remain critical. When it came time to confirm a target with vast political implications, it was not a high-tech targeting pod but a set of 12-power binoculars in the hands of an experienced AFAC that provided the accurate and critical identification.
I watched the post-incident briefing at Supreme Headquarters Allied Powers Europe (SHAPE) on public television a couple of weeks later. This was one of the most frustrating aspects of the episode. I wouldn’t describe the briefing as misleading, but the description just didn’t quite match what we had experienced. Arden and I had submitted our written reports to the CAOC through our rep at Vicenza—but I didn’t recognize any of those details in the SHAPE brief. I suspect that our reports were not part, or at least not the focus, of SHAPE’s investigation. Its briefing went into much detail on the circumstances of the incident, including identifying Bear 21 and showing his F-16’s cockpit video. The briefing appeared to focus on “how” this tragic incident could have happened.
The A-10’s role in terminating the attack was understated during the SHAPE briefing. A-10 AFAC expertise did not get credit for its critical role in identifying the noncombatant nature of the convoy, terminating the attack, and preventing further civilian casualties. Like the role we played, our call sign was also abbreviated. We were simply referred to as “Cub flight,” which could identify any one of several Cub call signs flying that day. I appreciated our leaders’ desire to explain the circumstances of “how” this could have happened to the public, get this tragic incident quickly behind us, and refocus on our continuing combat operations. However, my hope is that we don’t forget an important lesson—technology cannot completely substitute for training and experience. This is especially critical in basic tasks like looking at the ground and being able accurately to identify and understand what is really there—particularly when fighting a cynical enemy willing to use civilians as human shields.