Joint STARS in Bosnia

Too Much Data--Too Little Intel?

by Lieutenant Colonel Collin A. Agee

On the heels of great success in Operation DESERT STORM, a more mature Joint Surveillance and Target Attack Radar System (Joint STARS) deployed in support of Operation JOINT ENDEAVOR in December 1995. The system was technically a quantum improvement from its desert-tested predecessor. However, the intelligence community struggled to constructively apply the system to peacekeeping operations in Bosnia. This article examines the application of a technologically advanced system to stability and support operations (SASOs) and documents lessons learned specifically for our doctrine and force structure and delves into the complexities of combined operations and the implications for intelligence operations.

DESERT STORM A Tough Act to Follow

Joint STARS was justifiably lauded as one of the star performers during Operation DESERT STORM. U.S. Air Force Chief of Staff General Merrill A. McPeak gushed, We will not ever again want to fight without a Joint STARS kind of system. It is just terrific. Major General Stephen Silvasy, the Deputy Chief of Staff for Concepts, Doctrine and Development at the Army Training and Doctrine Command (TRADOC) seconded, Joint STARS worked. It saved countless Allied lives and proved to be critical to the outcome of DESERT STORM. The VII Corps Commander, General Frederick M. Franks, Jr., provided this ground commander's assessment Joint STARS' power and flexibility provided the United States Army with a new dimension of battlefield surveillance which made a significant contribution to the overwhelming victory in Southwest Asia.
Given that the system was in its infancy, just two days in theater when the DESERT STORM ground campaign commenced, this performance was incredible. But a more favorable scenario could hardly have been conceived:>
This last bullet may be less obvious and certainly less benign than the others, but provided immediate tangible feedback as to the system's contributions to the war effort, usually in the form of burning tank hulks.

A Challenging Operational Environment

By contrast, Bosnia was much more problematic for the Joint STARS application:
At one point, this accumulation of obstacles led the Ground Station Module (GSM) Task Force Commander, Lieutenant Colonel Steve Rotkoff, to call the system a nuclear sub in Kansas.

The Vision

Despite this group of obstacles, the Supreme Allied Commander Europe, General George A. Joulwan, without question wanted Joint STARS at his disposal for peacekeeping operations in Bosnia. His vision was that Joint STARS, in conjunction with cross-cued collectors, most notably the Airborne Reconnaissance Low (ARL) system, would serve as his tools to monitor the ZOS, allowing him to gauge and enforce compliance with the Dayton Peace Accord. He articulated this vision to Colonel Robert DeBusk, the commander of the Joint STARS Squadron, Joint Task Force (JTF) 4500 (Provisional), on 2 January 1996. He made these comments during a rampside visit to the operational base for the Joint STARS aircraft at Rhein Main Air Force Base, Germany.
General Joulwan's intent was to confront factional leaders with hard (photographic) evidence of noncompliance. In fact, he proposed taking them for demonstration rides in the collection platforms to make clear that we were watching. I will discuss the ability to satisfy this intent later.

Multinational Operations

One of the most significant challenges of the Joint STARS deployment was integration into a combined operation, with support rendered to British and French Divisions as well as various combined headquarters. Potential consumers of Joint STARS products included both the North Atlantic Treaty Organization (NATO) and non-NATO members of IFOR, including the Russian Brigade. Upon arrival in theater in late November 1995, Lieutenant Colonel Rotkoff and his staff began a frantic series of coordination visits to plan for a dozen GSMs at eleven sites in four countries. Unfortunately, the soon-to-arrive systems, with more than a dozen personnel and their complement of vehicles at some sites, had not been coordinated at national levels nor their acceptance articulated to the supported units. This placed the GSM coordination teams in the undesirable position of being traveling salesmen.
For those systems in the box (Bosnia), the European Command (EUCOM) assigned logistical responsibility to the U.S. Army Europe (USAREUR), a logical arrangement since these teams were supporting ground units. For the two GSM teams outside the U.S. sector, USAREUR's ability to provide daily supply of the most mundane necessities food, fuel, water, laundry and a place to sleep was virtually impossible due to the distance to the nearest U.S. unit or line of communication (LOC).
The British Headquarters (HQ) at Gornji Vakuf provides the most dramatic example. Though located only about 40 kilometers as the crow flies from U.S. elements in Sarajevo, there was no connecting LOC. Resupply was only available via the British convoy from the port city of Split, Croatia. It soon became obvious that life support could only be obtained from the supported unit. In an immature theater with scarce resources, this initially appeared to be a problem. The USAREUR Deputy Chief of Staff for Intelligence (DCSINT) discovered that there are implementing arrangements for mutual logistics support in effect with our Allies. This means of solving the logistics problem would have entailed tracking all the costs of support provided to the Americans by their British hosts every meal, gallon of gas, and load of laundry. In turn, the Americans could have attempted to quantify the cost of the intelligence provided via their GSM. Fortunately, this bureaucratic, legalistic solution was not necessary as the Brits treated the Yanks as their own.
The Allied reception of the GSM teams was not just a logistic concern, but a force protection issue as well. The French initially rejected the team that arrived to support them in Sarajevo, in the face of severe space limitations. The team soon learned that Sarajevo was not a good place to be homeless. In the words of the GSM Task Force (TF GSM) Commander: We...ended up stranding a team in Sarajevo without a life support sponsor for several days. What made this particularly distressing was that Sarajevo was not a safe place. During our tenure, sniping incidents were almost a daily occurrence and on one occasion GSM soldiers...were in a vehicle hit by shrapnel while driving to the Sarajevo airport.
An additional factor in the background of all dealings with our Allies was the impending NATO selection of a ground surveillance system. The French entrant in the competition is the Horizon, an MTI radar which lacks Joint STARS' synthetic aperture radar (SAR) capability. The Italians have a comparable heliborne platform called CRESO. The British Airborne Stand-Off Radar (ASTOR) is on the drawing board, with a 2001 projected service date.3 With two of these competitors supported by GSMs, tactical considerations potentially conflicted with national-level military and economic agendas. A final hurdle that the United States stumbled over on this deployment was the foreign release of Joint STARS-produced intelligence. Our joint doctrine calls for two categories of shared intelligence:
  • Level 1, which can be shown to but not retained by coalition and United Nations (U.N.) forces.
  • Level 2, which has been properly cleared for release to coalition and U.N. forces.4
  • It further states, The methodology for exchanging intelligence should be conceived and exercised well before operations begin.
    Early verbal guidance from the Joint STARS Squadron to the GSM crews was consistent with Level 1: they could permit their NATO partners to view data on their screens, but were prohibited from providing hard copies, which reduced the utility of the product. The original written guidance in December 1995, promulgated by the U.S. Air Force's Disclosure Policy Branch, was ambiguous.6 EUCOM did not publish definitive guidance until mid-February, a month and a half after initial operating capability was declared.7 In the absence of such guidance, elements at British, French, and multinational headquarters had to make their own, localized decisions on releasability.
    In any coalition undertaking, coordination is more complicated and time-consuming. In this instance, NATO, Supreme Headquarters Allied Powers Europe, EUCOM, U.S. Air Forces Europe (USAFE), and USAREUR were involved in planning for the deployment, not to mention the HQ at each GSM site.

    Lessons Learned

    We learned many lessons for multinational operations. Among them are @BULLET = Task organization should be established at the HQ which controls the military elements of both nations. The arrival of a system should not be a surprise to the recipient. The product must be validated within the intelligence framework at each supported HQ.
    • Logistics arrangements should be agreed upon in advance. Strive for a relationship that approximates our own concept of attachment for logistical support. With this goal in mind, a U.S. element should be as self-sufficient as possible, and arrive prepared to initially self-sustain in an immature operational area or when coordination for support has not been locked in.
    • GSM crews serve as subject matter experts and liaison officers (LNOs) point men for the system.
    • Attachments must be sensitive to other factors which may impact on Allied receptiveness to their intelligence support.
    • Foreign disclosure guidance should be established and disseminated prior to deployment.
    • Planning for coalition operations is inherently more complicated and time-consuming.

    Joint STARS Collection Management A Decentralized Process

    Tasking the aircraft occurs in two modes, not unlike preplanned and immediate calls for close air support. In the latter mode, the GSM crews transmit Radar Service Requests (RSRs) directly to the aircraft. (Although slave sites communicated with the aircraft via the master site, this was merely a relay, with no processing of the request.) In Bosnia, few targets could be identified far enough in advance for inclusion in the Air Tasking Order. More often, Joint STARS would be cross-cued via the supported unit's S2. When these requests reached the aircraft, they were managed by the Airborne Target Surveillance Supervisor, an Army senior noncommissioned officer, and the Airborne Surveillance Officer, an Air Force officer. But they usually dealt with rather cryptic RSRs without knowing the context of many of the requests they were servicing. This was somewhat disconcerting to the air crew, the Combined Air Operations Center, who perceived themselves as the clearing house for all requests, and various entities who sought to document the contributions of the system.
    This decentralized collection management, however, is consistent with our doctrine, which advocates broadcast intelligence to facilitate command, control, communications, computers, and intelligence for the warrior. In essence, the Joint STARS downlink is a form of broadcast intelligence in that the vast quantity of data continuously provided in the MTI mode overwhelms any one element's ability to digest it all. However, it does allow an individual GSM to focus on specific named areas of interest. If the supported commander gets a timely response to his priority intelligence requirements, that is success, even if the air crew or a higher HQ collection management cell is oblivious to the success. If the information warrants, it can be passed to other users or fused into an all-source product that results in a more complete picture of the battlefield.

    Freedom to Maneuver (More Accurately, Freedom to Collect)

    While the skies above the former Yugoslavia would intuitively seem a permissive flight environment, with air supremacy and only minimal risk of engagement from ground weapons, the Joint STARS squadron's use of airspace was significantly restricted. This hampered their efforts to overcome the limitations of the rugged terrain. Even over Hungary (which provided access to substantial ground facilities, airfields, and LOCs), safety concerns for civilian air traffic routes constrained the airspace. Unless we return to the desert (a flat one at that), we will confront radar shadowing by both natural and man-made objects in any future contingency. Flexibility in orbit selection and adjustment, based on a sound risk assessment and early planning, will lessen this obstacle by permitting the sensors to look down at terrain features rather than across them.

    Synchronizing Operations and Intelligence

    One of the seven basic principles of intelligence espoused in our joint doctrine is that we must synchronize intelligence with operations. Our military intelligence doctrine more specifically prescribes that our intelligence analysis is best done in the context of understanding the relative friendly-adversary situation. 10 In the initial stages of Operation JOINT ENDEAVOR, the flow of the IFOR into the theater was of great concern. For S2s, factional reaction or potential interdiction of those movements was of obvious interest. Unfortunately, the Joint STARS mission planning cell and air crews had poor visibility on friendly dispositions, plans, and movements.
    Since battlefield management is one of the major advertised capabilities of Joint STARS,11 we need to integrate operators into the Joint STARS planning team. For this deployment, the integration could have taken the form of LNOs assigned to the JTF 4500 mission planning cell at Rhein Main to provide insight into planned friendly activities and assess changes in the friendly situation during mission execution.

    Force Structure Implications

    The doctrinal size of a GSM team comprises two shift leaders, four operators (military operational specialty 96H), and a maintainer (MOS 33T).12 The actual size of the deployed teams for Operation JOINT ENDEAVOR, including the contractors, LNOs and supervisors, ranged from 9 to 15. That operational complement of six was designed to support 24-hour operations. For this mission, with two aircraft available, data flowed less than 12 hours daily, yet the teams were stretched to the limit. In addition to the time spent at the console, these teams must also conduct their personal hygiene, system maintenance, and pre- and post-mission analysis (including extensive replay and manipulation of data). They must also perform their fair share of guard duty and other details. The JOINT ENDEAVOR crews accomplished this for 56 consecutive days without a day off. In total, the operation lasted three months.
    The projected fleet of E-8 Joint STARS aircraft will number as many as nineteen, depending on budget decisions. This means that sustained, 24-hour data flow is on the horizon. This will overtax our crews as currently comprised, degrading their performance. The days of the GSM company are numbered according to the modified table of organization and equipment (MTOE) due to be implemented in September 1996. It will delete B Company from the Operations Battalion of corps military intelligence (MI) brigades. During JOINT ENDEAVOR, company commanders from both Fort Bragg and Fort Hood were gainfully employed supervising the employment and sustainment of the dozen teams at widely separated locations. Had this deployment occurred a year later, after the MTOE change, the GSM deployment could not have happened, at least not in the fashion executed by TF GSM. Downsizing notwithstanding, this role for the company HQ, while deployed, reinforces what we already knew about the company's role in training at the home station.
    The GSM company is essential for command and control as well as logistical support. Looking ahead, our force structure experts should carefully consider the manning and skill requirements for the impending Common Ground Station (CGS). We learned that the GSM crews were not just system operators, but were expected to interpret their data and synthesize it into the known intelligence situation. This requirement will be magnified by the addition of unmanned aerial vehicles, signals intelligence, aviation, broadcast intelligence, and secondary imagery on the CGS. The Army Intelligence Center advertises the following capabilities for the CGS:
    • Situation development.
    • Battle management.
    • Targeting.
    • Force protection.
    • Limited battle damage assessment and target damage assessment.
    • Operations planning.
    This impressive access to intelligence cries for an information manager who is part system expert, part analyst, and part collection manager. The information manager should be trained and employed on the battlefield by the MI brigade at corps level, and by the MI battalion at division. The abundance of information carries with it the risk that it will not be massaged into intelligence, and this incredible technology will fail to deliver to the commander a coherent, tactically useful product. As we evolve toward Intel XXI, we must be mindful that the intelligence task present is just as dependent on our expertise as it is on emerging technologies. The right combination of staffing, training, and equipment will result in our goal of helping the commander to understand the impact of what is being presented and use that understanding to make knowledge-based decisions.
    This operation also suggests reassessing the composition of the mission planning cell within the Joint STARS squadron. The majority of this cell is comprised of air crew members who are not flying, on a rotational basis. Doctrinally, 3 to 5 members of a 21-person crew are Army.15 A small, permanent intelligence cell had no Army representation and was augmented by two Army analysts from the USAREUR Combat Intelligence Readiness Facility. While the Air Force is likely to continue to outnumber the Army on the air crew, we need a permanent Army presence in the intelligence cell, in the form of three to five 96B analysts. This modification will better equip the mission planning cell to provide accurate mission briefs. In this contingency, despite the wings on the collector, the collection targets were almost exclusively ground targets.

    Digesting the Data

    Perhaps the biggest challenge to effective use of the system was data overload: the inability to distinguish significant MTIs from the voluminous data stream. From the first training mission, it became apparent that the screen would quickly fill with MTIs, making it difficult to determine those of military significance. As the comparison of DESERT STORM and JOINT ENDEAVOR revealed, the classic mass formations of armored vehicles moving across uncluttered terrain simply did not exist.
    Questions posed to the system had to be more specific. Use of Joint STARS confirmed or denied movement in towns thought deserted. It monitored suspected mass burial sites, often in isolated areas where any movement in the dead of winter was suspect.16 Recalling that General Joulwan saw Joint STARS as the cornerstone of peace treaty compliance, the JTF 4500 Commander had each ZOS digitized so that they were readily available on every GSM monitor as well as aboard the aircraft. An effort is ongoing to apply automation to the process to help sort the wheat from the chaff.
    Lieutenant Colonel David Anhalt of JTF 4500 is developing software enhancements that would use automation to monitor traffic at selected points.17 Over time, traffic patterns can be discerned; deviations from historical averages will alert analysts to unusual activity.18 Unfortunately, this capability was not available in time to impact on JOINT ENDEAVOR, though it holds great promise for future missions. One capability that is wedded to conventional warfare is Joint STARS' ability to distinguish between wheeled and tracked vehicles. Using differential Doppler, an algorithm detects the proximity of radar reflective surfaces moving at different speeds (like a track on a tank) and marks the MTI as a track. In Bosnia, this resulted in many more false icons than accurate ones. For example, two vehicles that pass in opposite directions on a highway provide the requisite radar returns to generate a track symbol. In this environment, we would have been well served to turn the capability off and remove the misleading information.
    Despite the aggressive efforts of all elements of JTF 4500 and their supported S2s, the raw volume of hard intelligence provided to commanders was not impressive, certainly not in comparison to DESERT STORM. For GSM operators living in austere conditions and interacting with their supported units daily, this was a hard pill to swallow. But the real measure of success was not directly proportional to the number of hard reports rendered.

    Success or Failure?

    At the tactical level, it would be difficult to justify the expense of this sophisticated system and its 450-person squadron. At the operational level, and not readily apparent to these beleaguered soldiers, there was a more significant contribution. We must keep in mind the ultimate goal of the commander: enforcement of the Dayton Treaty. If selected violations could be documented, IFOR would benefit from the deterrent effect of the factions' knowledge that they were under observation. General Joulwan strove to present an aura of omniscience. That the aircraft did not have 24 hour coverage was not critical complete coverage was not required. The use of Joint STARS in conventional warfare may have a chilling effect on enemy commanders because their desire to achieve mass and mobility will be counterpoised by the certainty that such formations will be immediately detected in all conditions of light and weather. So too will a potential Bosnian combatant think twice about traversing a forbidden area if he thinks that someone is observing him.
    Joint STARS' recording capability also alleviates the requirement to catch him in the act. An MTI history can be replayed to track a suspected mover, identifying his origin, his destination, and his rate and route of movement. The Joint STARS' SAR was used to verify the existence of equipment collection sites, define to a limited extent the contents of those sites, and track changes over time.
    In the Cold War era, intelligence professionals constantly sought to distinguish capabilities from intentions. In the Post-Cold War era, Operation JOINT ENDEAVOR suggests that we must now understand the difference between perception and reality. Intelligence has become a force multiplier, not only in allowing the commander to effectively employ his assets to parry the enemy, but now also serves an important deterrent role. If our goal is to provide perfect intelligence, SASO will give us perfect frustration. Even the term enemy has become outdated in the context of rival ethnic groups in a region grasping for peace. But if Joint STARS or any system can contribute to preventing hostilities, it has accomplished the mission. The challenge for intelligence professionals is to learn the dissimilar lessons of both Operations DESERT STORM and JOINT ENDEAVOR. They must understand that their commander's role may be equal parts political and military, and that the ultimate consumer of their intelligence may well be a potential belligerent that the commander strives to keep in check. By synthesizing and moving beyond the experiences of these two contingencies, we will be better prepared to provide intelligence for the next.
    Lieutenant Colonel Collin A. Agee is the Deputy Commander of the 525th MI Brigade (Airborne) at Fort Bragg, North Carolina. During the early stages of Operation JOINT ENDEAVOR, he was the 525th MI Brigade S3, serving as an augmentee to USAREUR. He has a bachelor of arts degree in National Security and Public Affairs from West Point and a Masters of Military Arts and Sciences from the School of Advanced Military Studies. Readers can reach the author at (910) 396-3209/6084, DSN 236-3209/6084, and via E-mail at