B.  Department of Defense Modeling and Simulation Capability

1.  Modeling and Simulation Conducted During the Gulf War

The modeling for the Gulf War air campaign began in late 1990 in the Pentagon under the direction of an air staff organization known as Checkmate, supported by the Air Force Studies and Analysis Agency.[19] Our investigation found that this modeling arose not from concerns related to Coalition bombing of chemical and biological warfare targets, but from concerns about Coalition aircraft and aircrew losses. Computer-driven simulations were derived from the Command, Control, Communications, and Intelligence Simulation Model, known as C3ISIM. This model is a classic attrition model which simulated flying specific aircraft missions during which the aircraft were exposed to enemy air defenses. The objective of this model was to estimate aircraft losses during a particular mission and then to adjust the mission by changing the aircraft routes to avoid enemy air defenses, by suppressing enemy defenses with other aircraft, or by applying other tactics to reduce the estimated aircraft losses. A team of analysts deployed to Saudi Arabia with C3ISIM and supported the air campaign planners during combat operations.[20]

According to a newspaper interview with the senior Gulf War air campaign planner, Air Force Brigadier General Buster Glosson, computer modeling and simulation honed the attacks against the Iraqi chemical and biological warfare weapons targets. "It’s fair to say we did a significant amount of modeling trying to determine the safest—if there is such a thing—way to destroy the storage areas associated with areas of mass destruction and research and production areas…. Never has a war been so programmed, so modeled. I guarantee it."[21] The newspaper interview went on to report that "the simulations were conducted by analysts in the ‘Checkmate’ planning group … and by number crunchers using the Air Force C3ISIM system."[22]

An officer who worked with C3ISIM in Saudi Arabia told us that C3ISIM was only an aircraft attrition model used to improve aircraft attacks with respect to US and Coalition aircraft survivability; at no time did he or his team model attacks against specific targets.[23] Consequently, this modeling did not evaluate the potential hazards of the dispersion of chemical and biological warfare agents resulting from bombed Iraqi facilities. General Charles Horner, the senior US Central Command Air Force commander, recalled other modeling, but that modeling focused on estimating the time required to achieve certain levels of attrition of Iraq’s ground forces.[24] At our request, the Air Force Deputy Chief of Staff, Air and Space Operations, researched the planning and modeling of bombing missions against chemical and biological targets during the Gulf War. "No records were found of modeling … for the effects of releasing these agents during the bombing missions…."[25]

Checkmate also employed modeling to estimate Iraqi casualties that might result from Coalition bombing. This effort focused exclusively on attacks by Coalition aircraft against non-chemical and non-biological warfare agent targets in Baghdad. Sponsored by the US Air Force Human Systems Division at the Brooks Medical Facility, San Antonio, Texas, the Threat Related Attrition model was used to conduct casualty estimations.[26]

Our investigation discovered one joint study of Coalition attacks against Iraqi weapons facilities. In January 1991, the US Army Engineers at Waterways Experiment Station, Vicksburg, Mississippi, conducted explosive model testing on a mock Iraqi munitions storage bunker. Using a 1/3-scale model of the Iraqi bunker, Waterways Experiment Station engineers evaluated the effects of attacking the facility with precision-guided, high-explosive weapons. In conjunction with the Waterways Experiment Station modeling, a scientist from the Wright Laboratory at Eglin Air Force Base, Florida, conducted tests of the expulsion of simulants from the exploded Waterways Experiment Station model. Neither the Waterways Experiment Station nor the Wright Laboratory results were available before the end of the very short Gulf War air campaign.[27]

We also found that, in 1990, scientists at the Defense Nuclear Agency modeled scenarios of Iraq's possible use of chemical and biological warfare agents on the battlefield against Coalition forces, but they did not extend these models to examine potential releases caused by Coalition bombing of chemical and biological warfare targets.[28] In December 1990, the Central Intelligence Agency and Defense Intelligence Agency assembled the Iraq Interagency Biological Warfare Working Group to examine the capabilities of the Iraqi biological warfare program and Iraq’s intentions.[29] The group developed a methodology to estimate collateral damage to Iraq’s population using a series of Calder Tables (a tabular estimate of biological warfare agent dispersal) to estimate casualties.[30] This methodology included assumptions concerning Iraqi population distribution and biological warfare agent distribution.

The Defense Nuclear Agency (DNA) had developed a modeling and simulation capability built on existing hazard prediction and weapons effects models.[31] This capability allowed the agency to develop more than 16,000 templates[32] to model the effects of Iraq’s possible chemical and biological attacks against Coalition forces[33] under a variety of generic meteorological and terrain conditions. At the same time, a contractor working for the DNA estimated possible Iraqi agent dispersion using different amounts of agent under varying weather conditions.[34] In addition, the Defense Nuclear Agency operations center provided around-the-clock predictions and analysis to US military leaders in Saudi Arabia.[35] The DNA operations center staff conducted numerous "what if" analyses of hypothetical Iraqi chemical weapons attacks against Coalition forces.[36] The DNA staff officers discussed the question of potential chemical releases caused by Coalition bombing of Iraqi chemical warfare facilities. Although the staff officers did not have the resources to properly analyze the question, their professional experience and knowledge of chemical agents led them to conclude that any release of chemical warfare agents from Coalition bombing had little likelihood of significant collateral damage, would result in little agent dispersal, and most of the agent would be incinerated as a result of the attack.[37]

Our investigation also researched Defense Intelligence Agency and Central Intelligence Agency modeling and simulation capabilities. Until after the Gulf War, the Defense Intelligence Agency had no in-house capability to assess the effects of weapons of mass destruction, but relied on Defense Nuclear Agency modeling, which, as stated earlier, was not used to predict hazards to Coalition forces from Coalition air attacks on Iraqi targets. According to an experienced modeling contractor, the Central Intelligence Agency had some capability to conduct hazard prediction modeling,[38] but we found no evidence that the agency did any such modeling before or during the Gulf War.  Even so, the Central Intelligence Agency and the Defense Intelligence Agency shared the concerns of the various air planners about biological warfare agents, and those concerns led the Iraq Interagency Biological Warfare Working Group to examine the biological warfare agent issue.[39]

We contacted other government agencies (shown in Table 1) known for modeling and simulation to determine if there were any other instances of modeling and simulation of Gulf War-related scenarios. The table identifies the Department of Defense agencies with known modeling and simulation capabilities interviewed by our investigators, whether or not the agencies did modeling in support of the Gulf War, the type of modeling done, and if the results of the modeling were available to appropriate decision makers during the Gulf War. More details of the specific interviews are summarized in Tab C.

Table 1. Results of modeling and simulation research

Agencies Contacted

Conducted
Modeling
and
Simulation

Type
Of
Simulation
Conducted

Results
Available
for
Planning

Aberdeen Proving Ground

No

Casualty Prediction Yes
Air Force Human Systems Division

Yes

Casualty Prediction

Yes

Air Force Technical Applications Center

No

Casualty Prediction Yes
Air Force Studies and Analysis Agency

Yes

Aircraft Attrition

Yes

Central Intelligence Agency

No

Casualty Prediction Yes
Defense Intelligence Agency

No

Casualty Prediction Yes
Defense Nuclear Agency

Yes

CW/BW Defense

Yes

Defense Nuclear Agency

Yes

Dispersion of Iraqi attack

Yes

Dugway Proving Ground

No

Casualty Prediction Yes
Iraq Interagency Biological Warfare Working Group

Yes

Casualty Prediction

Yes

Lawrence Livermore National Laboratory*

Yes

Demonstration

No

Waterways Experiment Station

Yes

Weapons Effects

No

Wright Laboratory

Yes

Dispersion

No

* Source of modeling of chemical releases reported in USA Today

2.  Modeling and Simulation Capability Since the Gulf War

Between 1990 and 1999, several government agencies improved their chemical and biological warfare agent modeling and simulation capabilities, emphasizing agent dispersal scenarios of facilities targeted for bombing. Many DoD scientists consider the Defense Threat Reduction Agency (DTRA) (formerly called the Defense Nuclear Agency) to be the focal point for such activities.[40] The DTRA runs HASCAL (Hazard Assessment System for Consequence Analysis) - a model developed for the DTRA to support analysis of chemical, radiological, and biological incidents. HASCAL consists of the Second-Order Closure, Integrated Puff (SCIPUFF) ) and its associated tool kit, the Hazard Prediction and Assessment Capability (HPAC).[41] It is used to estimate contamination and doses at nuclear, chemical, and biological facilities incidents. SCIPUFF is a model for projecting the transport of released agent.[42] It projects the increasing diffusion of overlapping wind puffs as a cloud grows and moves over time and takes into account variations in concentration during dispersion.

In addition to DTRA, the Naval Surface Warfare Center (NSWC) in Dahlgren, Virginia, and the US Army Edgewood Research, Development, and Engineering Center at Aberdeen Proving Ground, Maryland, possess chemical warfare agent modeling and simulation capabilities.[43] NSWC runs the Vapor Liquid Solid Tracking (VLSTRACK) model. VLSTRACK provides approximate downwind hazard predictions for many currently known chemical and biological agents.[44]

The US Army Edgewood Research, Development, and Engineering Center is the proponent of the Non-Uniform Simple Surface Evaporation Model (NUSSE).[45] The latest version, NUSSE-4, predicts the hazardous environment created by a single liquid-filled chemical munition.[46] It calculates the coverage (footprint) of the warhead’s lethal area on the ground by following the transport and diffusion of the agent vapor cloud from release to impact with the ground. The Central Intelligence Agency also uses NUSSE-4 and has used it to support the Office of the Special Assistant.[47]

The Department of Energy supports the operation of the National Atmospheric Release Advisory Center at Lawrence Livermore National Laboratory for emergency response assessments and projections of hazardous release accidents and related events. The Department of Energy also supports the development of advanced systems for chemical, nuclear, and biological release assessment at Lawrence Livermore, Los Alamos, Lawrence Berkley, and Argonne National Laboratories. These programs are providing new, sophisticated tools spanning the broad spectrum of hazard dispersion within/from buildings and subways to regional and transnational scales.

The Office of the Special Assistant for Gulf War Illnesses has sponsored several modeling and simulation efforts. In coordination with the intelligence community, the chemical warfare agent demolition incident at Khamasiyah was modeled using an ensemble of weather and dispersion models. Currently, the Office of the Special Assistant is re-modeling the Khamisiyah demolition based on more accurate information about the winds and the Central Intelligence Agency’s revised assessment of the amount of agent released in that demolition. We are also sponsoring modeling efforts to assess the possible dispersal of contaminants from the Iraq’s chemical warfare agent facilities at Al Muthanna and Muhammadiyat that were bombed during the Gulf War. The approach taken for these efforts will follow the same ensemble approach used in the Khamisiyah modeling. The results of these simulations will be documented in their respective case narratives.

The modeling ensemble combines weather prediction models and transport/diffusion models to determine the extent of a hazard created by a possible chemical warfare agent release. Transport/diffusion models use data (e.g., wind and temperature) generated by weather prediction models to display how an agent released into the atmosphere would be dispersed and define the extent of any subsequent contamination. Additionally, this ensemble approach combines the results of multiple weather models and multiple transport/diffusion models to assure the optimum identification of areas in which veterans could have been at risk.


| First Page | Prev Page | Next Page |