Search

There is a need for groundwater modelling approaches that can incorporate both the complexity of riverbeds and the influence of lateral flows through the river banks in the simulation of river–aquifer interaction. A simple and straight-forward approach is proposed for modelling river–aquifer interaction in gaining rivers based on existing building blocks available in MODFLOW. The river water itself is modelled as constant-head cells, while the riverbed is modelled as one or more separate layers. The Horizontal Flow Barrier (HFB) Package is used to simulate the effect of a river bank on lateral exchange fluxes between river and aquifer. This approach has several advantages over conventional MODFLOW streamflow packages: (1) riverbeds can be modelled as several layers with different hydraulic properties; (2) both vertical and horizontal flow through the riverbed can be modelled; and (3) lateral flow through the river banks can be simulated. The capabilities of the new approach are demonstrated in two applications based on a MODFLOW model of the Aa River, Belgium. Application I demonstrates that the approach can be used to model heterogeneous multi-layered riverbeds. Results show that neglecting this heterogeneity can result in an overestimation of river–aquifer exchange fluxes. Application II shows that the approach can be used to simulate the effect of lateral fluxes through river banks. It demonstrates that lateral fluxes can be a major contributor to total river–aquifer exchange fluxes. The proposed approach provides the necessary tools to accurately model river–aquifer interaction for gaining rivers in MODFLOW. [ABSTRACT FROM AUTHOR]

The spatial distribution of hydraulic conductivity (K) in riverbeds is essential to understand and model river-groundwater interactions. However, K in riverbeds varies over several orders of magnitude and its spatial distribution is closely linked to complex geological and fluvial processes. Investigating the local distribution and spatial heterogeneity of K is therefore a challenging task. The use of direct current (DC) and time-domain-induced polarization (IP) geoelectrical methods to map qualitatively the spatial distribution of K within riverbeds is described. The approach is demonstrated for a test site situated in a typical lowland river in Belgium. Inverted geophysical parameters (resistivity, chargeability and normalized chargeability) are compared with estimates of K obtained through slug tests. In general, high values of K are observed in the middle of the river and lower values towards the banks, while the opposite is true for chargeability and normalized chargeability. Therefore, there exists an inverse correlation between K and IP geophysical parameters. Furthermore, geostatistical analyses using variograms show that all parameters have ranges of similar magnitudes. The strong correlation between K and chargeability or normalized chargeability can be explained by the fact that all three parameters are mainly controlled by clay and organic matter content. [ABSTRACT FROM AUTHOR]

Among the advances made in analytical and numerical analysis methods to quantify groundwater/surface-water interaction, one methodology that stands out is the use of heat as an environmental tracer. A large data set of river and riverbed temperature profiles from the Aa River in Belgium has been used to examine the spatial-temporal variations of groundwater/surface-water interaction. Exchange fluxes were calculated with the numerical heat-transport code STRIVE. The code was applied in transient mode to overcome previous limitations of steady-state analysis, and allowed for the calculation of model quality. In autumn and winter the mean exchange fluxes reached −90 mm d−1, while in spring and early summer fluxes were −42 mm d−1. Predominantly gaining conditions occurred along the river reach; however, in a few areas the direction of flow changed in time. The river banks showed elevated fluxes up to a factor of 3 compared to the center of the river. Higher fluxes were detected in the upstream section of the reach. Due to the influence of exchange fluxes along the river banks, larger temporal variations were found in the downstream section. The exchange fluxes at the river banks seemed more driven by variable local exchange flows, while the center of the river was dominated by deep and steady regional groundwater flows. These spatial and temporal differences in groundwater/surface-water exchange show the importance of long-term investigations on the driving forces of hyporheic processes across different scales. [ABSTRACT FROM AUTHOR]