Dating base flow in streams using dissolved gases and diurnal temperature changes

Abstract

A method is presented for using dissolved CFCs or SF 6 to estimate the apparent age of stream base flow by indirectly estimating the mean concentration of the tracer in the inflowing groundwater. The mean value is estimated simultaneously with the mean residence times of the gas and water in the stream by sampling the stream for one or both age tracers, along with dissolved nitrogen and argon at a single location over a period of approximately 12-14 h. The data are fitted to an equation representing the temporal in-stream gas exchange as it responds to the diurnal temperature fluctuation. The efficacy of the method is demonstrated by collecting and analyzing samples at six different stream locations across parts of northern Virginia, USA. The studied streams drain watersheds with areas of between 2 and 122 km 2 during periods when the diurnal stream temperature ranged between 2 and 58C. The method has the advantage of estimating the mean groundwater residence time of discharge from the watershed to the stream without the need for the collection of groundwater infiltrating to streambeds or local groundwater sampled from shallow observation wells near the stream.

Baseflow Age Distributions and Depth of Active Groundwater Flow in a Snow‐Dominated Mountain Headwater Basin

“…When T r and A e were known, estimated values of C gw for SF 6 and CFC‐113 were used to calculate their atmospheric concentrations at the time of recharge, which in turn provides a mean age for baseflow through use of an assumed travel time distribution (Busenberg & Plummer, 2002). Sanford et al (2015) assumed a Weibull distribution in which the cumulative distribution function (CDF) is defined as follows: in which the scale parameter ( k ) and shape parameter ( n ) are adjusted to match the computed atmospheric concentration to the historical record (USGS, 2017). The Weibull distribution is equivalent to the exponential mixing model when n = 1.…”

“…Stream water dissolved gas samples were collected following the approach presented by Sanford et al (2015). Samples were collected hourly for two relatively inert atmospheric gases (N 2 and Ar) and two gas age tracers (SF 6 and CFC-113) over a 12-hr period.…”

“…Values of C gw for each gas were estimated using a modified version of the Excel spreadsheet calculator developed by Sanford et al (2015), which minimizes the misfit between measured and modeled values of C s employing the automated General Reduced Gradient solver tool (provided as supporting documentation).…”

“…When T r and A e were known, estimated values of C gw for SF 6 and CFC-113 were used to calculate their atmospheric concentrations at the time of recharge, which in turn provides a mean age for baseflow through use of an assumed travel time distribution (Busenberg & Plummer, 2002). Sanford et al (2015) assumed a Weibull distribution in which the cumulative distribution function (CDF) is defined as follows:…”

“…This suggests that (a) the age of groundwater discharging into the stream was sufficiently large that the difference between C gw and C e was nontrivial (on the order of years rather than weeks/months) and (b) though τ g values were relatively small for Copper Creek, they were still large enough so that the age tracer C gw signal was maintained in the stream. Therefore, despite Copper Creek's high gradient, v g for the age tracer gases were sufficiently slow to permit application of the method proposed by Sanford et al (2015). Although SF 6 and CFC-113 were generally well mixed in the atmosphere, local atmospheric concentrations on the day of sampling could be slightly below the Northern Hemisphere 6-month average values used to compute C e (USGS, 2017), such that C s was actually equal to C e to invalidate the method.…”