Application of Statistically Based Site Characterization Tools to Victorville Precision Bombing Ranges Y and 15 for the ESTCP Wide Area Assessment Demonstration.

摘要

Efficient characterization and remediation of sites potentially contaminated with unexploded ordnance (UXO) remain a high priority for the U.S. Department of Defense (DoD). Recent estimates of the amount of land that is potentially contaminated with UXO are as high as 10 million acres (4 million hectares). This total land area is comprised of as many as several thousand individual sites. Characterization efforts to date have shown that at a typical site the UXO contamination is concentrated in a small portion of the site area, often only 10 to 20 percent of the entire area. Therefore, efficient site characterization should be focused on identifying the location and extent of these smaller areas within a site. Toward this goal, Pacific Northwest National Laboratory (PNNL) and Sandia National Laboratories (SNL) have developed efficient and defensible, statistically based approaches for UXO site characterization. The Environmental Security Technology Certification Program (ESTCP) established several demonstrations of UXO site characterization technologies under a Wide Area Assessment (WAA) Project. This report focuses on the application and performance of statistically based tools to the Victorville Precision Bombing Range located near Victorville, California. PNNL and SNL have developed statistical algorithms to create transect designs based on desired Data Quality Objectives (DQO) and then identify potential target areas based on the surveyed transects. The transect design tools provide a statistically defensible method that uses transect survey data for only a small proportion of the total study area (i.e., 1 to 3 percent) to identify target areas of a specific size, shape, and anomaly density. Target area density estimates, probability estimates, and density flagging routines are applied after the data are gathered from the established transect design to separate potential target areas from areas that require no further remediation.