II.  INVESTIGATION

Section A discusses the investigation of the report mentioned in the USA Today article. Section B discusses our review of the Department of Defense modeling and simulation capabilities as they existed before and after the Gulf War.

A.  The USA Today Article

The August 13, 1997, edition of USA Today featured an article entitled, "Pentagon Was Aware of Gulf Chemical Threat."[1] Less than two weeks later, on August 25, 1997, another article appeared in the Army Times recounting the same information in the USA Today article.[2] According to the story, a report produced by the Lawrence Livermore National Laboratory informed the US Air Force that bombing Iraqi chemical weapons plants and storage bunkers was certain to release deadly nerve agents that could endanger American troops. According to the article, the report was delivered to the Air Force in October 1990, three months before the beginning of the Gulf War air campaign.

To begin our investigation, a team from the Special Assistant's office traveled to Lawrence Livermore National Laboratory in California and interviewed the scientists, analysts, the project director, and the division director who were involved in the development of the report.

The division director recalled a request for assistance—specifically, a request from the Air Force 5th Weather Wing[3] to depict the area that a chemical warfare agent plume might cover—had been made informally by telephone and had been undertaken because the director of Lawrence Livermore National Laboratory had offered to provide whatever support necessary to the Department of Defense for Operation Desert Shield/Desert Storm.[4]

The project manager, who led the effort and authored the document identified in the USA Today article, obtained data from the Laboratory intelligence division and also from the Air Force 5th Weather Wing officer who had requested the effort.[5] A photo interpreter from the intelligence division calculated the maximum storage capacity of a typical storage site. The photo interpreter stated that his efforts were not in-depth analyses. Due to the short turnaround time and without specific guidance from the requestor, the photo interpreter devoted only three to four working days to this project.[6] The requestor provided data pertaining to the weapons to be used in the attack simulation.[7] The project manager received no data on the type or quantity of agents or their locations. He made assumptions on the particle size of the agent, the weather, terrain, and other environmental factors.[8] He used the Mass Consistent Wind Field/Atmospheric Diffusion Particle in Cell model for this project. Previously this model had been used only in simulations of nuclear incidents.[9] The project manager constructed a generic template to demonstrate how the model could be applied to a wide range of targets and summarized his findings in a memorandum for the record.[10] The project manager insisted he had never submitted a formal report and that he did not consider this effort a study.[11]

According to the USA Today article, the Livermore report was initially delivered to the Air Force 5th Weather Wing at Langley Air Force Base, Virginia.[12] Through our investigation, we learned that the report was delivered to the same officer who made the request for assistance. During our interview, this officer recalled that the original request had come from an agency in Saudi Arabia (probably deployed members of the 5th Weather Wing), and that, because she believed the request could not be answered locally, she pursued agencies that could help. The officer eventually contacted Lawrence Livermore National Laboratory for assistance. The 5th Weather Wing officer also recalled receiving the report from Livermore, and after review, believed the plumes depicted were so large that they were unrealistic. The report was passed on to her supervisor.[13]

The staff at Lawrence Livermore National Laboratory received no additional requests from the Air Force and no feedback regarding the model itself.[14] There is no evidence that any civilian or military leader outside of the 5th Weather Wing was made aware of Livermore's work. It was not until 1997, after a Freedom of Information Act request,[15] that the subject of a Livermore model surfaced in the national media. The USA Today article claimed that the Department of Defense was aware of possible release hazards from Coalition bombing, citing the Livermore report as proof.[16]

As interpreted by the person who gave the information to the newspapers, the study showed that US bombing of chemical warfare-related facilities would spread chemical warfare agents up to 378 miles downwind, into areas containing Coalition forces.[17] However, when our investigator showed the plumes associated with the document released under the Freedom of Information Act request to the Livermore scientists, they did not recognize them. The Livermore file copy of the demonstration effort contained a completely different set of plumes. After some discussion, the Livermore scientists concluded that the Pentagon had provided the wrong potential hazard area graphs in response to the Freedom of Information Act request. What had been delivered instead were potential nuclear fall-out predictors developed for the Department of Energy. The predictors of a hypothetical nuclear incident covered a much larger area than those developed for Livermore's generic chemical warfare agent demonstration.[18]

With our investigation of the USA Today article completed on the matter of Lawrence Livermore National Laboratory's role—or absence of a role—in the planning process, there remained the possibility that other Department of Defense agencies may have conducted modeling and simulation analyses of bombing chemical or biological warfare facilities. To be as thorough as possible and to determine who may have conducted similar efforts, we interviewed experts from agencies known for modeling and simulation expertise (Tab C). At the same time, we wanted to identify the state of the art of modeling and simulation during the Gulf War and how such capabilities have evolved since then.

We learned that modeling did, in fact, occur prior to the air campaign, but the goal of the modeling was not to predict potential health risks to US and Coalition forces resulting from bombings of chemical or biological warfare agent sites, but to predict aircraft attrition or the possible results of a chemical warfare attack by Iraq on Coalition forces. Section B, which follows, describes combat and dispersion models that were available before and during the Gulf War. While none of these models was specifically designed or executed to predict the health effects of chemical or biological warfare agents released during a Coalition bombing attack on Iraqi facilities, we nonetheless tried to determine whether any of the models may have helped Gulf War planners reach some conclusions about the risks to Coalition forces associated with bombing chemical or biological weapons facilities. Details on our research are located in Tab C.


| First Page | Prev Page | Next Page |