USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2009–5025

Back to Table of Contents

Appendix A. Quality Control and Quality-Assurance of Water Samples

Eighteen percent of the chlorophyll a samples, 33 percent of the total phosphorus samples, 32 percent of the total nitrogen samples, 36 percent of the dissolved nutrient samples (orthophosphate, ammonia, and nitrite-plus-nitrate), 38 percent of the particulate samples (total particulate nitrogen, total particulate carbon, particulate inorganic carbon, and particulate organic carbon), and 37 percent of the dissolved organic carbon samples collected were used for quality-assurance purposes, which included field blanks (the first sample collected every week) and either a split sample or a method replicate (each type every other week). Field blanks are samples of reagent-grade inorganic blank water processed onsite through clean sampling equipment, before an environmental sample is collected. Analysis of blank samples determines whether the processes of collection, handling, transport, and analysis cause measurable contamination. Split samples are environmental water samples collected once and divided with the use of a churn splitter into two or more samples that are used to determine the variability in the analytical methods. Replicate samples are environmental samples collected twice in rapid succession from the same location and analyzed to determine the variability of the system, the sampling method, and the analytical methods.

The results of the quality-assurance sampling indicated that precision and accuracy were acceptable and the variability in sampling and processing was smaller than the week-to-week variability (table A1). Most of field-blank concentrations of total phosphorus and orthophosphate were less than the laboratory’s minimum reporting level. The median values of the blank samples greater than the minimum reporting level were 0.006 and 0.013 mg/L for total phosphorus and orthophosphate, respectively. The field-blank samples of total nitrogen and ammonia revealed some contamination, where almost one-half of the samples were greater than laboratory minimum reporting level, although it was minimal compared to the concentrations of the environmental samples collected the same week. The number of nitrite-plus-nitrate blank sample values greater than the laboratory minimum reporting level was less than for total nitrogen and ammonia, indicating that the contamination in total nitrogen samples was dominated by the ammonia-nitrogen species. Only one blank sample collected for total particulate nitrogen was greater than the minimum reporting level. Contamination of ammonia blanks in a 2003–04 study was attributed to atmospheric deposition (Wood and others, 2006). The use of capsule filters since then has diminished but did not eliminate the problem completely. Total particulate carbon blank samples were all below minimum reporting level. The median concentration of DOC blank samples was twice the minimum reporting level, indicating an increased chance for contamination.

Variability of analytical methods as measured by split samples generally was low. The median relative percentage difference of chlorophyll a, ammonia, total particulate nitrogen, total particulate carbon, and particulate organic carbon was greater than 10 percent, but less than 13 percent in split samples. Relative percentage difference of the other analytes was less than 10 percent in split samples. Variability of the environmental system and analytical methods as measured by replicate samples generally was low. The median relative percentage difference of replicate samples generally is slightly higher than split samples. This is expected because the variability of the environmental samples is added to the variability of the sampling and analytical methods. The median relative percentage difference of chlorophyll a, total particulate nitrogen, total particulate carbon, and particulate organic carbon was greater than 10 percent and less than 16 percent in replicate samples. Relative percentage difference of the other analytes was less than 10 percent in replicate samples. The large relative percentage difference in ammonia split samples was high due to a large difference (40 percent) in a set of low concentration samples (6 and 10 µg/L). The relative percentage difference of chlorophyll a, total particulate nitrogen, total particulate carbon, and particulate organic carbon split and replicate samples was slightly greater than 10 percent due to the variability of the filtration process. Overall, variability of quality-assurance samples was minimal relative to the variability of the environmental samples.

Back to Table of Contents