Abstract
Abstract. Baseflow recession analysis and groundwater dating have up to now developed as two distinct branches of hydrogeology and have been used to solve entirely different problems. We show that by combining two classical models, namely the Boussinesq equation describing spring baseflow recession, and the exponential piston-flow model used in groundwater dating studies, the parameters describing the transit time distribution of an aquifer can be in some cases estimated to a far more accurate degree than with the latter alone. Under the assumption that the aquifer basis is sub-horizontal, the mean transit time of water in the saturated zone can be estimated from spring baseflow recession. This provides an independent estimate of groundwater transit time that can refine those obtained from tritium measurements. The approach is illustrated in a case study predicting atrazine concentration trend in a series of springs draining the fractured-rock aquifer known as the Luxembourg Sandstone. A transport model calibrated on tritium measurements alone predicted different times to trend reversal following the nationwide ban on atrazine in 2005 with different rates of decrease. For some of the springs, the actual time of trend reversal and the rate of change agreed extremely well with the model calibrated using both tritium measurements and the recession of spring discharge during the dry season. The agreement between predicted and observed values was however poorer for the springs displaying the most gentle recessions, possibly indicating a stronger influence of continuous groundwater recharge during the summer months.
Highlights
Spring baseflow recession analysis has a long history in hydrology, starting more than a century ago when Boussinesq and Maillet proposed using quadratic or exponential laws to describe spring recession (Boussinesq, 1904; Maillet, 1905)
As was shown from the statistical analysis of the recessions of 100 karstic springs (Drogue, 1972) and using numerical techniques (Dewandel et al, 2003), a quadratic law describes much more truthfully spring recession than an exponential law, the latter is more popular in groundwater hydrology
We show that the discharge recession can be used to reduce parameter uncertainties in a model predicting atrazine concentration in spring water over time.The parameters of the model are estimated from tritium measurements and baseflow recession, and predictions compared with the observed atrazine time series in a second testing step
Summary
Spring baseflow recession analysis has a long history in hydrology, starting more than a century ago when Boussinesq and Maillet proposed using quadratic or exponential laws to describe spring recession (Boussinesq, 1904; Maillet, 1905) Following this seminal work, a number of more complex models have been developed combining two or more reservoirs and considering non-linear responses (Horton, 1933; Brutsaert and Nieber, 1977; Brutsaert, 1994; Coutagne, 1948; Mangin, 1970; Padilla et al, 1994). Dewandel et al (2003) give an excellent review of the subject, and classify all recession studies into two approaches: the first considering drainage in both saturated and unsaturated zones, and the second concentrating on the recession of the saturated zone only For the latter, only Boussinesq’s quadratic solution is both analytically exact and interpretable in terms of hydraulic parameters of the aquifer (hydraulic conductivity and effective porosity). Boussinesq quadratic law was preferred over an exponential law to describe spring flow recession in this paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
