Investigations of Contaminated Fluvial Sediment Deposits: Merging of Statistical and Geomorphic Approaches

Abstract

Concentrations of contaminants in sediment deposits can have large spatial variability resulting from geomorphic processes acting over long time periods. Thus, systematic (e.g., regularly spaced sample locations) or random sampling approaches might be inefficient and/or lead to highly biased results. We demonstrate the bias associated with systematic sampling and compare these results to those achieved by methods that merge a geomorphic approach to evaluating the physical system and stratified random sampling concepts. By combining these approaches, we achieve a more efficient and less biased characterization of sediment contamination in fluvial systems. These methods are applied using a phased sampling approach to characterize radiological contamination in sediment deposits in two semiarid canyons that have received historical releases from the Los Alamos National Laboratory. Uncertainty in contaminant inventory was used as a metric to evaluate the adequacy of sampling during these phased investigations. Simple, one-dimensional Monte Carlo simulations were used to estimate uncertainty in contaminant inventory. We also show how one can use stratified random sampling theory to help estimate uncertainty in mean contaminant concentrations.