Quantifying uncertainties in the assessment of sediment quality: Statistical criteria and guidelines for sediment quality assessments

Abstract

Current sediment quality guidelines generally adopt a tiered approach in order to assess sediment quality more cost‐effectively. The uncertainties involved in the tiered approach of an integrative assessment, however have not been quantified resulting in a risk of committing type I error or type II error at the final confirmatory stage. This study develops statistical criteria and guidelines for the sediment chemistry component of an integrative assessment of sediment quality. At the tier 1 screening stage, historical data or an initial survey is required to determine the minimum sample numbers that will be required to be representative of the study site. Understanding the guiding principles which have underpinned the setting of sediment quality criteria for contaminant is an important factor in tier 1 evaluations. To reduce cost and uncertainty in data, sampling should include the least number of samples necessary to minimise uncertainties by estimating the probability distribution function (which represents the variability of the natural environment at the location of concern), with subsequent application of kriging and sequential simulation methods on data obtained. Implementation of statistical criteria and guidelines in sediment quality assessments can provide a foundation for a further quantitative cost / benefit analysis and decrease the risk of committing type I and type II statistical errors at later classification stages.