3D surface-subsurface modeling of a bromide tracer test in a macroporous tile-drained field: Improvements and limitations

The assessment and forecasting of nutrient loss by tile drains in agricultural areas often rely on physically based models that have adequate representations of macropores and tile drains. Macroporosity has been adequately represented in hydrological models using a dual continuum approach. However, its implementation in hydrological and solute transport models is limited to plot-scale or to one- and two-dimensional models due to the large number of parameters that are rarely available and the long computational times. The purpose of this study is to simulate a tracer test using a 3D coupled surface-subsurface model to improve the representation of the tracer concentration at the drainage discharge. A three-dimensional HydroGeoSphere model was developed and calibrated to simulate tile drainage discharge, Br mass discharge, and hydraulic heads from a Br tracer test in a densely tile-drained field. The conductivity of the drain was one of the most important parameters for drain discharge and solute transport simulations. The model accurately simulated drainage discharge and Br transport to tile drains. However, most of the Br peaks and the late-time Br mass in the drain outflow were underestimated. Our simulation results indicate that explicitly representing tile drains with seepage nodes allows for a physically based, yet computationally efficient representation of Br transport behavior surrounding tile drains at field scale. However, we cannot confirm that the single-porosity model with immobile zone is suitable for simulating the Br peaks at the drain outlet and the late-time Br mass. Improvements to the model include the implementation of heterogeneous soil layers and the inclusion of more measured data to reduce uncertainty during calibration.

Automated summary

In this study, a three-dimensional coupled surface-subsurface model was used to simulate a tracer test in a densely tile-drained field, aiming to improve the representation of tracer concentration at the drainage discharge. The model accurately simulated drainage discharge and tracer transport to tile drains, but underestimated the tracer peaks and the late-time tracer mass at the drain outlet. Improvements to the model include the implementation of heterogeneous soil layers and the inclusion of more measured data to reduce uncertainty during calibration.