invited Talk, Research Seminar at the Dept. of Hydrogeology, UFZ Leipzig: 2012-02-15
The interaction between stream discharge and the riparian zone during high flow events is still poorly understood. Chemical data support the existence of defined flow paths controlling the exchange between riparian zone and streams. After rainfall events, waters rich in DOC, resembling the chemical composition of uppermost soil pore waters, is rapidly mobilized and mixes with waters from deeper layers and groundwater in various proportions. Monitoring and modeling of the spatial and temporal dynamics and influence of the different flow pathways on runoff generation and water quality is therefore very difficult. Chemical and hydrometric data taken for a minerotrophic fen located in North- Eastern Bavaria show that water quality and runoff generation during intensive rainfall events is controlled by complex interactions between surface flow and groundwater. Processes controlling runoff generation and water quality seem to predominantly occur within the riparian zone. The signature of runoff during intensive storm flow events often shows a similar chemical composition as the water stored in the riparian zone In this study, the hypothesis is tested that micro-topographical structures, which are typical for this types of wetlands, induce heterogeneous sub-surface flow patterns. Furthermore, they allow a complex interplay of storage and rapid mobilization of surface near pore waters. Thus, the influence of these subsurface-flow patterns on the biogeochemical processes within the wetland is investigated. Herby, a process based surface/sub-surface flow model is used which is capable of representing the interaction between precipitation, riparian zone and stream. Microtopographical structures are generated using a geostatistical approach and are integrated into the hydrological model. Along the flow paths, different biogeochemical processes are modeled, using field and literature data for initial concentrations, reaction rates, and switching points between processes. Processes were implemented as rate expressions in PhreeqC and are controlled by availability of substrates. The different flow paths with individual characteristic residence times allow for turnover of organic matter and electron acceptors, depending on rate and flow. Following field observations, we included aerobic respiration, denitrification, iron reduction, sulfate reduction, and organic matter degradation, leading also to ammonium release. All processes are resolved in high spatio-temporal resolution and cause different chemical characteristics of the waters mobilized during high flow events. Complementary to the simulations, we show field data of different soil pore water profiles. The measured data supports our hypothesis that mainly rapidly mobilized waters from the near surface layers are reflected in the chemical signature of stream water. However, there is also a chemical imprint from waters mixing in from deeper groundwater, as nitrate, sulfate, and e.g. chloride show different mixing behavior during high flow events. This supports our model hypothesis that different flow paths with different residence times contribute to the overall runoff, depending on the event and pre-event conditions.