Effect of Hyporheic Exchange on Conservative and Reactive Solute Transport in Streams : Model Assessments Based on Tracer Tests

Understanding of the processes affecting solute transport in flowing water is important for the possibility to predict the evolution with time of polluted stream systems. This thesis presents tracer experiment methodology and model developments for solute transport in streams, with special focus on retention processes and their effect on solute stream transport. Results are presented from a tracer experiment in the Säva Stream, Uppland County, Sweden, where both a conservative (3H as tritiated water) and a reactive (51Cr as trivalent chromium ion) tracer were injected simultaneously. The time and length scales of the experiment were prolonged compared to previous studies, which allowed for new critical tests of different model concepts. It was found that the hyporheic exchange greatly affected the solute transport of both tracers. However, the retention of chromium was significantly more pronounced. About 76% of the injected chromium was lost from the stream water phase directly after the passage of the pulse 30 km downstream of the injection point. The inventory of chromium in the sediments indicated that the main part was retrieved in the hyporheic zone. Both a diffusive and an advective hyporheic exchange model were developed and evaluated versus independent observations in the stream water and hyporheic zone. Analytical expressions for the central temporal moments of the breakthrough curve and semi-analytical solutions for the solute concentration in the Laplace domain were derived. Both models were found useful in representing the observations. For the transport of the reactive solute, it was found essential to consider a kinetic sorption on to particulate matter in the hyporheic zone. The time needed for a wash-out of 75% of the maximum uptake in the hyporheic zone was found to be ~85 times longer for the reactive solute, compared to the conservative solute. Neglecting the sorption kinetics in the transport model yielded significant errors in the central temporal moments, which implied an incorrect description of the wash-out process from the hyporheic zone. Independent observations in the stream water and hyporheic zone as well as choice of evaluation method are essential for a correct interpretation of the processes. A first attempt was also made to link model parameters such as the residence time in the hyporheic zone with measurable parameters of the stream. Such a relationship offers the possibility to generalize results for other streams and stream conditions. A generic study of the transport of an inert solute in the Lule River, Sweden, using this kind of relationship, indicated that the hyporheic exchange can have a large practical implication