Quality Assurance

Data-collection and analytical procedures used in this study incorporated practices designed to control, verify, and assess the quality of sample data. Methods and associated quality control for collection and field processing of water-quality samples are described by Knapton (1985), Edwards and Glysson (1988), Ward and Harr (1990), Knapton and Nimick (1991), and Horowitz and others (1994).

The quality of analytical results reported for water-quality samples can be evaluated with data from quality-control samples that were submitted from the field and analyzed concurrently in the laboratory with routine samples. These quality-control samples consisted of duplicates and blanks, which provide quantitative information on the precision and bias of the overall field and laboratory process. In addition to quality-control samples submitted from the field, internal quality-assurance practices at the NWQL were performed systematically to provide quality control of analytical procedures (Pritt and Raese, 1995). These internal practices included analyses of quality-control samples such as calibration standards, standard-reference-water samples, replicate samples, deionized-water blanks, or spiked samples at a proportion equivalent to at least 10 percent of the sample load.

Duplicate samples were obtained in the field to provide data on precision (reproducibility) for samples exposed to all sources of variability. Precision of analytical results for field duplicates is affected by many sources of variability within the field and laboratory environments, including sample collection, processing, and analysis. For this study, two duplicate samples were obtained in the field by splitting a single composite sample into two subsamples, which then were analyzed separately. Analytical results for field duplicates are presented in table 6. Precision of analytical results for any constituent can be determined from the relative percent difference (RPD) of the concentrations of the constituent in duplicate analyses. The RPD is calculated for a constituent by dividing the absolute value of the difference between the two concentrations by the mean of the two concentrations. RPD values for dissolved and total-recoverable metals were almost all less than 10 percent, with no systematic exceedance of this value, indicating good precision for analytical results.

Two field blanks were analyzed to identify the presence and magnitude of contamination that potentially could bias analytical results. A field blank is an aliquot of deionized water that is certified as essentially free of the measured trace elements and that is processed through the sampling equipment used to collect stream samples. The blank is then subjected to the same processing (filtration, preservation, transportation, and laboratory handling) as stream samples. Blank samples were analyzed for the same constituents as those of stream samples to identify whether any detectable concentrations existed. Analytical results for field blanks are presented in table 6. With the exception of dissolved aluminum and copper, concentrations of all metals were less than the minimum reporting level. Consequently, analytical results (except for dissolved aluminum and copper) for the synoptic samples are assumed to be free of significant or systematic bias from contamination associated with sample collection and processing. Dissolved aluminum was detected at low concentrations (2-6 μg/L) in field blanks as well as in water samples from some surface inflows that otherwise had dissolved metal concentrations less than the minimum reporting level. These data suggest that all water samples may have been affected by low-level contamination from sampling and processing equipment, sample bottles, or the nitric acid used for sample preservation. Dissolved copper was detected in one of the two field blanks. The dissolved copper concentration in this blank was 3.8 μg/L while the total-recoverable concentration was <1 μg/L, likely indicating some cross contamination from the sample processed prior to the blank. This contamination probably resulted from incomplete rinsing of the filter-plate apparatus. On the basis of these data, the possibility exists that cross contamination affected the dissolved copper concentration in the few synoptic samples that had concentrations of dissolved copper that were low (<5 μg/L) but higher than the total-recoverable concentration. However, cross contamination was not evident in the other field blank or the majority of synoptic samples, particularly samples from surface inflows with low or undetectable copper concentrations.

Previous: Synoptic sampling

Next: Tracer injection results