Using high-frequency water quality data to assess sampling strategies for the EU Water Framework Directive

摘要

The EU Water Framework Directive (WFD) requires that the ecological andchemical status of water bodies in Europe should be assessed, and actiontaken where possible to ensure that at least "good" quality is attained ineach case by 2015. This paper is concerned with the accuracy and precisionwith which chemical status in rivers can be measured given certain samplingstrategies, and how this can be improved. High-frequency (hourly) chemicaldata from four rivers in southern England were subsampled to simulatedifferent sampling strategies for four parameters used for WFDclassification: dissolved phosphorus, dissolved oxygen, pH and watertemperature. These data sub-sets were then used to calculate the WFDclassification for each site. Monthly sampling was less precise than weeklysampling, but the effect on WFD classification depended on the closeness ofthe range of concentrations to the class boundaries. In some cases, monthlysampling for a year could result in the same water body being assigned tothree or four of the WFD classes with 95% confidence, due to randomsampling effects, whereas with weekly sampling this was one or two classesfor the same cases. In the most extreme case, the same water body could havebeen assigned to any of the five WFD quality classes. Weekly samplingconsiderably reduces the uncertainties compared to monthly sampling. Thewidth of the weekly sampled confidence intervals was about 33% that of themonthly for P species and pH, about 50% for dissolved oxygen, and about67% for water temperature. For water temperature, which is assessed as the98th percentile in the UK, monthly sampling biases the mean downwards byabout 1 °C compared to the true value, due to problems of assessinghigh percentiles with limited data. Low-frequency measurements will generallybe unsuitable for assessing standards expressed as high percentiles.Confining sampling to the working week compared to all 7 days made littledifference, but a modest improvement in precision could be obtained bysampling at the same time of day within a 3 h time window, and this isrecommended. For parameters with a strong diel variation, such as dissolvedoxygen, the value obtained, and thus possibly the WFD classification, candepend markedly on when in the cycle the sample was taken. Specifying this inthe sampling regime would be a straightforward way to improve precision, butthere needs to be agreement about how best to characterise risk in differenttypes of river. These results suggest that in some cases it will be difficultto assign accurate WFD chemical classes or to detect likely trends usingcurrent sampling regimes, even for these largely groundwater-fed rivers. Amore critical approach to sampling is needed to ensure that managementactions are appropriate and supported by data.