9409 - Laserfiche WebLink

Home Browse Search

Table 3. Distribution of field-blank samples [--, no sample or not applicable; lab, office laboratory] Site Land Previous Gener I use site chemic al constitu nts State Line Mixed Reed h Dotsero Mixed Baker 1 Gunnison Mixed Gunnison near m near Grand Grand Junction Junction Gunnison Mixed Lab cleaned Tunnel reference2 Lab cleaned Lab cleaned East Rural urban, Gunnison at recreation 32 Road Gore Rural urban, French recreation French Mining French Dry Agriculture Lab cleaned Baker Reference, Lab cleaned srm3 Lab cleaned Concentration range at previous site 1 Dissolved Suspended Nutrients organic organic carbon carbon Trace elements h 1 h 1 h m m m m m m m h m h lConcentration ranges with respect to all sites within the surface-wa er network, I = low, m = medium, h = high. 2Downstream from major reservoir system. 3Physiographic province, srm = Southern Rocky Mountain. different combinations of sources of variability. The pre-processing split replicate, the concurrent and sequential replicate, and the post-processing split-replicate samples collected in surface water of the UCOL study unit for water years 1995-96 are discussed in the following sections. Data published by the U.S. Geological Survey are rounded. For example, a calcium value of 150 mglL (milligrams per liter) is reported (rounded) to the nearest 10 mglL. A calcium value of 6.5 mglL is rounded to the nearest 0.1 mglL. Data rounding is incorporated in the assessment of replicate analysis in this report by using a unit called the lowest rounding unit. A lowest rounding unit of I would represent 10 mglL for the 150 mglL sample and would represent 0.1 mglL for the 6.5 mg/L sample. A lowest rounding unit is equal to the magnitude of the least significant figure reported by the USGS National Water Quality Laboratory. In this report, the differences between the environmental and replicate values are given in the units of the constituent and in lowest rounding units. The use of lowest rounding units for comparison avoids skewing comparisons when values of constitu- ents span orders of magnitude. For example, consider the two pairs of magnesium values for the French and Dry sites (table 4). The environmental and split- replicate values for French are 5.5 and 5.6 mglL, resulting in a difference of -0.1 mglL. The Dry envi- ronmental and split-replicate values are 36 and 37 mg/L, resulting in a difference of -1 mglL. Both of the differences are -1 unit of the least significant figure or -1 unit of the lowest rounding value. The practical difference between the environmental and replicate values is similar in both of these cases (the only smaller difference available for each pair of samples would be zero). A comparison of the actual magnitude difference between the environmental and split-replicate pairs (-0.1 to -1 mg/L) is of little use. A difference between the environmental and split- replicate sample for the Dry site could not be 0.1 mglL because the data are reported to the nearest whole milligram per liter. Percent differences between the environmental and replicate values do not yield practical significance for many of the constituents in this data set. For example, the chloride environmental and replicate values for French have a -0.1 mglL difference (table 4), which represents a -10 percent difference. QUALITY-CONTROL SAMPLING PROGRAM 5