Main content area

Controlled laboratory experiments and modeling of vegetative filter strips with shallow water tables

Fox, Garey A., Muñoz-Carpena, Rafael, Purvis, Rebecca A.
Journal of hydrology 2018 v.556 pp. 1-9
Cynodon dactylon, agrochemicals, algorithms, drainage, filter strips, hydrograph, laboratory experimentation, lowlands, model validation, models, overland flow, pollution, probability, runoff, sandy loam soils, silt loam soils, streams, surface water, vegetation, water table
Natural or planted vegetation at the edge of fields or adjacent to streams, also known as vegetative filter strips (VFS), are commonly used as an environmental mitigation practice for runoff pollution and agrochemical spray drift. The VFS position in lowlands near water bodies often implies the presence of a seasonal shallow water table (WT). In spite of its potential importance, there is limited experimental work that systematically studies the effect of shallow WTs on VFS efficacy. Previous research recently coupled a new physically based algorithm describing infiltration into soils bounded by a water table into the VFS numerical overland flow and transport model, VFSMOD, to simulate VFS dynamics under shallow WT conditions. In this study, we tested the performance of the model against laboratory mesoscale data under controlled conditions. A laboratory soil box (1.0 m wide, 2.0 m long, and 0.7 m deep) was used to simulate a VFS and quantify the influence of shallow WTs on runoff. Experiments included planted Bermuda grass on repacked silt loam and sandy loam soils. A series of experiments were performed including a free drainage case (no WT) and a static shallow water table (0.3–0.4 m below ground surface). For each soil type, this research first calibrated VFSMOD to the observed outflow hydrograph for the free drainage experiments to parameterize the soil hydraulic and vegetation parameters, and then evaluated the model based on outflow hydrographs for the shallow WT experiments. This research used several statistical metrics and a new approach based on hypothesis testing of the Nash-Sutcliffe model efficiency coefficient (NSE) to evaluate model performance. The new VFSMOD routines successfully simulated the outflow hydrographs under both free drainage and shallow WT conditions. Statistical metrics considered the model performance valid with greater than 99.5% probability across all scenarios. This research also simulated the shallow water table experiments with both free drainage and various water table depths to quantify the effect of assuming the former boundary condition. For these two soil types, shallow WTs within 1.0–1.2 m below the soil surface influenced infiltration. Existing models will suggest a more protective vegetative filter strip than what actually exists if shallow water table conditions are not considered.