A simple method for detection of changes in relations between solute concentration and stream discharge

Abstract:

The relation between constituent concentrations and stream discharge (C/Q relations) are fundamental to the estimation of fluxes or loads in biogeochemical studies. C/Q relations are useful for understanding nutrient, trace element, and contaminant behavior in response to storm and snowmelt related changes in discharge. The shape and seasonal variation of C/Q relations provides information about availability, mobilization, and release of solutes to streams. The properties of C/Q relations can allude to flowpaths, antecedent moisture conditions, and solute availability. Changes in C/Q relations over time for certain constituents like dissolved organic carbon (DOC) may be indicative of changes in supply that may have resulted from changes in climate, vegetation, or land use and land cover. The focus of this presentation is on a simple method for detection of change in C/Q relations using the LOADEST regression model. The LOADEST model fits a seasonally variable C/Q relation to discrete water quality data. For a continuously gauged stream or river, a relatively long record of C/Q data can be partitioned into distinct periods and regression models can be determined for each period. By running each model with the same discharge record and subsequently plotting each flux time series, differences between models can be visualized graphically. Plotting differences between periods (models) illustrates at what times of year the differences are largest. Running each model with a range of discharges for each day of the year provides additional insight into whether the changes in C/Q relations are evident at all levels of discharge or only at specific levels of discharge. The DOC record (1991 to 2014) from a research watershed at Sleepers River in Vermont was used in this analysis. The analysis showed that there have been increases in DOC concentration for certain seasons and rates of discharge.