Uncertainties related to sampling and their impact on the chemical analysis of groundwater

It is necessary to take into account the impact of the field stages on the chemical analysis of natural water, in terms of both sampling representativeness and spatial and temporal variability. We demonstrated this need in studies conducted at two industrial sites, which involved groundwater contaminated with metals and organochlorine compounds. We developed a methodology for estimating the uncertainties linked to the entire analytical chain - from sampling to handing over results to the client. This methodology is based on the use of groundwater chemical logs acquired in the field for evaluating field heterogeneity and identifying water levels to be sampled. We validated the methodology through specifically-designed repeatability and reproducibility tests carried out in the field. The results illustrated that a geochemical understanding of the natural environment and the changes it undergoes is possible, but requires an international consensus on the methodology best suited to application in field tests. If not, major inconsistencies and/or erroneous interpretations may arise in environmental monitoring programs. (C) 2003 Elsevier Ltd. All rights reserved.