New reporting conventions were applied to four methods beginning October 1, 1998 (U.S. Geological Survey National Water Quality Laboratory Technical Memorandum 98.07, 1998) and will be applicable to selected high-demand NWQL water methods beginning in fiscal year 2000. For all methods, the new reporting conventions allow reporting of analytes detected at concentrations less than the LRL and as low as the LT–MDL, whereas non-detections are reported as <LRL. [Note: the LRL previously was called non-detection value or NDV (Connor and others, 1998: U.S. Geological Survey National Water Quality Laboratory Technical Memorandum 98.07 1998), a term that is no longer used.]

The new low-concentration reporting conventions are outlined in figure 10. Conventions are shown for situations where the analyte is detected (shading) in the quantitative region and in the semiquantitative region above and, for information-rich methods, below the LT–MDL. Also shown are situations where the instrumental response is in the non-quantitative region where analyte detection is unclear and <LRL is reported. Conventions are shown for methods where the lowest calibration standard (LS) concentration is greater than or less than the LRL. Decreasing shading represents regions of increasing measurement uncertainty.

At concentrations less than the LRL, the risk of a false negative increases rapidly, and calibration standards with concentrations less than the LRL often may not be detected. Therefore, the LS used for quantitation typically will be at a concentration equivalent to or greater than the LRL. Realistically, data users should expect that analyte concentrations much less than LRL would not be consistently detected in samples over time. Additionally, the lower the reported result in relation to the LRL or the LS, the less accurate the measurement is likely to be. For this reason, the NWQL is qualifying all reported concentrations less than the LRL or the LS, whichever is higher, as estimated using an “E” remark code. Reported concentrations that are greater than the highest calibration standard also will receive an “E” remark code. The “E” remark code will be added to all values falling outside (above or below) the calibration range because of increased measurement uncertainty. The “E” remark code will not be used for values reported for analytes that are diluted into the calibration range and re-analyzed.

Figure 10.  New low-concentration reporting conventions showing the reported value and associated.

image size 40KB-gif

Figure 10. New low-concentration reporting conventions showing the reported value and associated qualifying remark code in relation to the long-term method detection level (LT–MDL), the laboratory reporting level (LRL), and the lowest calibration standard (LS). [> greater than or equal to;< , less than; E, estimated]


Information-Rich Methods

Analytical results for information-rich methods will follow slightly different reporting conventions. The NWQL has categorized the following instrumental techniques as information-rich methods for organic chemical analyses.

  1. Gas chromatography with mass spectrometry (GC/MS); for example,
  1. High performance liquid chromatography with mass spectrometry (HPLC/MS); new methods under development.
  2. High performance liquid chromatography with photodiode array ultraviolet/visible spectroscopy (HPLC/DAD); for example,

These method types have qualifying information provided by the detector, in addition to signal and retention time matching, that enhance analyte identification. For mass spectral methods, the presence of characteristic mass spectral ions with correct ion ratios augment analyte identification. Characteristic absorption spectra are used to enhance analyte identification in photodiode array detection.

Because qualitative identification is required before a concentration is reported, information–rich methods will not be restricted to censoring all measurements below the LT–MDL. Therefore, information-rich methods will report estimated (“E” remark code) concentrations for all positively identified analytes below the LT–MDL if other quality control criteria, especially laboratory blank-related criteria, are met. Because data below the LT–MDL have an increased risk of false positives, data users should carefully examine these reported concentrations with respect to both laboratory and field blank data.

Additional reasons underlying the reporting changes are detailed in previous sections. The effect of potential yearly changes of LT–MDL’s and LRL’s in the USGS National Water Information System (NWIS) data base are discussed in the sections, “Considerations for Data Interpretation” and “Storing Data in the National Water Information System” that follow.

Use of the “E” Remark Code by the National Water Quality Laboratory

The “E” remark code currently (1999) is used to signify that a measured concentration is estimated by the NWQL. The availability of remark codes in the NWIS data base is limited. A wide variety of conditions can justify evoking the “E” remark code other than those conditions shown in figure 10. Furthermore, some “E” coded values are more uncertain than others. For example, the “E” code also is used by the NWQL under the following conditions:

next arrow Planned implementation at the National Water Quality Laboratory
back arrow Setting The Reporting Level
up arrow U.S. Geological Survey Open-File Report 99-193