Water quality monitoring is done for different reasons, two of the most important of which are to: a) determine compliance near point discharges and to b) assess affects of pollution on a watershed from point and non-point sources. In principle, most comprehensive data sets for the first objective could be used for the second, although most data are not captured into a central, publicly accessible database. Compliance monitoring can be limited to sites of waste discharges, and thus inadequate for broader water quality assessments. The largest water quality archive in the USA is currently STORET, maintained by the US Environmental Protection Agency. It incorporates data from other government agencies such as the US Geological Survey, volunteer data, and any other programs or agencies that choose to submit data. As a long-term archive, STORET has at least two limitations: undocumented quality assurance and quality control procedures and large spatial and temporal gaps, as large as 15-20 years at some sites.

Large, coordinated K-12 science and educational programs such as GLOBE offer very good potential to help bridge some spatial and temporal gaps present in government agency databases such as STORET and NWIS, gaps that inhibit regional water quality assessments. To test this, we investigated the temporal and spatial relationships between the databases in. three states, Massachusetts, California and Colorado. In Colorado data were also available from Rivers of Colorado Water Watch Network, a statewide program that concentrates on water quality data collection for grades K-12. We found that accessing the data was the most time consuming task for many of the databases — GLOBE’s database being the most easily accessible because it is housed on the web and does not require special permission to view the data. Temporally, we found that many of the gaps in STORET and NWIS occur before the inception of the GLOBE and Rivers of Colorado programs, however, they helped to fill in some of the more recent temporal gaps.

Comparing GLOBE with other water quality data sets, one sees some striking differences. Colorado Rivers sites have the lowest mean temperatures, which may be a result of factors such as altitude, snowmelt, etc. GLOBE sites have higher means, and have several outliers above 30°C. USGS sites have means approximately 3 - 5°C warmer than the other two data sets, which may be the result of site locations. GLOBE mean pH values are generally lower than the other two data sets, with Colorado Riverwatch highest. There is much greater variability among the three data sets in respect to alkalinity, with almost 20% of the GLOBE sites reporting a mean alkalinity of 0. Since the other two data sets do not have a similar distribution; it is unlikely that these represent actual water conditions. Removing the sites with zero alkalinity makes the GLOBE data correspond much better with the other two sets.