Data validation of PAMS auto-GC VOC data: More lessons learned
- Sonoma Technology, Inc., Santa Rosa, CA (United States)
Data validation of continuous volatile organic compound (VOC) data collected with PAMS auto-GCs at several sites in the US including the Northeast was performed. Using a graphical software platform from which to display the data, more than 1.7 million data entries were reviewed quickly and efficiently. The hourly concentrations of the PAMS target species were displayed in several ways including scatter plots, fingerprints, and time series. Scatter plots were useful for showing species misidentification and relationships among species. Fingerprint plots were used to inspect the concentration of each species in a sample (in the order in which the species elute from the chromatography). Time series plots were useful in showing the diurnal behavior of a species, species misidentification, sample contamination, etc. From our experience on various projects, invalid and suspect hydrocarbon data can be found in AIRS for nearly all PAMS sites. At some sites, the questionable data account for a small percentage of the data reported while at other sites data problems may be a more significant concern. In general, data resulting from calibration runs, operating problems, etc. had been removed from most data sets by the state monitoring people resulting in better data quality for the 1995 PAMS data collected in the Northeast than for 1994. This paper provides examples of the types of data that were identified as suspect or invalid; these results will aid other PAMS operators in evaluating their continuous VOC data and aid data analysts in their use of the data.
- OSTI ID:
- 617792
- Report Number(s):
- CONF-9704195--
- Country of Publication:
- United States
- Language:
- English
Similar Records
Spatial and temporal variations of PAMS VOCs in the northeast
Receptor modeling of VOC data