CONCLUSIONS AND RECOMMENDATIONS
None of the UXO risk assessment methods that we evaluated fully meets the Army's need for sound technical analysis to inform decisionmaking, either for the purpose of setting priorities among UXO sites or for detailed analysis of explosion and munitions constituents risks at individual sites. Table 5.1 summarizes the methods' strengths and limitations. As shown, each method falls short in one or more of the key criteria necessary for an effective method: technical soundness of risk calculations, effectiveness of implementation, or ease of communication. Furthermore, stakeholders and regulators involved at the various sites have not uniformly accepted these methods as credible elements of the decisionmaking process, and continued reliance on them is likely to delay the UXO response process further.
A fundamental reason why none of the modeling methods evaluated meets the Army's needs is that the UXO problem is not reducible to a single, objective measure of risk. Risk methods must address the risk of explosion of the munitions but also consider the risk of chemicals from exploded munitions and UXO that leach into the soil and groundwater. Further, the methods used for analyzing these two broad categories of risk (explosion and munitions constituents), while different in substance, both depend on subjective judgments about modeling assumptions and data. For example, assessing the explosion risk requires, among other types of information, estimates of the probability that humans will come in contact with UXO. These estimates require assumptions about human behavior and predictions of future population and land use; the density and distribution of UXO items that cannot be seen because they are buried; and the