Continuous Water-Quality Monitoring—A Valuable Tool for TMDL Programs

By Teresa J. Rasmussen, Andrew C. Ziegler, Patrick P. Rasmussen, and Thomas C. Stiles

Abstract

Sensor technology currently is not available to directly measure many chemicals of interest in streams; therefore, regression models are used to estimate stream chemical concentrations from the relation between laboratory-analyzed samples and in-stream sensor measurements such as turbidity and specific conductance. Continuous hourly values from in-stream sensors (turbidity, specific conductance, dissolved oxygen, pH, and water temperature) and regression models make it possible to estimate concentrations and loads for different time periods--daily, weekly, monthly, or annually. Hourly values help characterize load fluctuations under changing streamflow and seasonal conditions and in response to different contributing areas within a watershed. Uncertainty in regression-estimated concentrations is defined using prediction intervals and typical regression diagnostic statistics including mean square error (MSE) and the coefficient of determination (R2). The continuous estimated data, associated uncertainty, and probability and duration curves for selected stream monitoring sites in Kansas are available on the World Wide Web at URL: http://ks.water.usgs.gov/rtqw/.

Continuous data provide the foundation for a more comprehensive evaluation of the variability in loading characteristics and water-quality degradation than provided by discrete water-quality samples. Continuous concentration estimates can be used to construct cumulative frequency distribution (duration) curves to determine percentage of time that estimated concentrations exceed water-quality criteria. Estimated concentration and load duration curves can be used to evaluate current water-quality conditions and estimate the duration and magnitude of potential water-quality degradation. Examination of the differences in regression-estimated concentrations and loads at a series of sensor stations along a stream allows the analysis of upstream-to-downstream changes in water quality. In situations where discrete samples and constituent concentration data are necessary for regulatory requirements, monitoring by continuous sensor data allows regulatory agencies to optimize sampling efforts. When continuous estimates are considered over the long term, it may be possible to identify changes in water-quality conditions resulting from land-use changes and implementation of best management practices in the watershed.

Rasmussen, T.J., Ziegler, A.C., Rasmussen, P.P., and Stiles, T.C., 2003, Continuous water-quality monitoring—a valuable tool for TMDL programs, in Proceedings of National TMDL Science and Policy 2003 Speciality Conference, Chicago, Illinois, November 16–19, 2003: Water Environment Federaton, compact disk, session 11, 17 p.

Additional information about real-time water-quality and TMDL studies in Kansas can be found at: http://ks.water.usgs.gov/rtqw/ and http://ks.water.usgs.gov/studies/TMDL/

For additional information contact:

Andy Ziegler
U.S. Geological Survey
4821 Quail Crest Place
Lawrence, KS 66049-3839
Telephone: (785) 832-3539
Fax: (785) 832-3500
Email: aziegler@usgs.gov