Representative sampling for energetic compounds at an antitank firing range

Abstract : Field sampling experiments were conducted at the CFB-Valcartier Arnhem antitank rocket range to investigate various sampling schemes that would yield representative soil samples at firing points and impact areas of antitank ranges. Three sampling strategies were evaluated. Between the firing point and the target, 10-m x 10-m grids were established and 30-increment composite soil samples were collected. In two of these grids, one near the firing point and one at the target, other grids were divided into 100 1-m x 1-m minigrids. Within each minigrid a discrete and a 10-increment composite soil sample were collected and analyzed for energetic compounds. In the target area, an alternative strategy was also evaluated using concentric halos around the target. Each halo was subdivided into increasing numbers of segments at increasing distances from the targets. Multi-increment composite samples were collected within each halo segment. Behind the firing line, nine line (linear) composites were collected at various distances from 0 to 25 m from the firing line. Results from the 100 1-m x 1-m minigrids near the firing line and the target demonstrated that the distribution of analyte concentrations in the discrete samples was non-Gaussian and the range of concentrations varied over two orders of magnitude. The distributions of data for multi-increment composite samples with various numbers of increments were simulated by averaging the concentration estimates from randomly selected discrete samples. For the firing line area, the distribution of NG computed composites exhibits increased normality as the number of increments is increased and the resulting tolerance range declined substantially. This was also true for HMX in the target area. Recommendations are made for appropriate sampling strategies to collect representative surface soil sample for antitank rocket ranges.