Evaluation of the Root Zone Water Quality Model (RZWQM) for Southern Ontario: Part I. Sensitivity Analysis, Calibration, and Validation

Abstract

Abstract This study focuses on the performance of the Root Zone Water Quality Model (RZWQM) for corn production in southern Ontario. The model was used to simulate the amount of subsurface tile drainage, residual soil nitrate-nitrogen (NO3-N), NO3-N in subsurface drainage water, and crop yield. A precalibration sensitivity analysis of the model was conducted for several key parameters using field data collected at the study site. The RZWQM's hydrology component was most sensitive to the Brooks and Corey fitting parameters and saturated hydraulic conductivity (Ks), while the tile drain flow and the water table depth were sensitive to the Brooks and Corey fitting parameters of bubbling pressure (ψbp) and pore-size-distribution index (λ). The fraction of dead-end pores had relatively little effect on tile drain N loss. The crop yield is most affected by N uptake, age, and evapotranspiration rate. RZWQM simulated evapotranspiration was within the range (568 ± 55 mm) of the observed evapotranspiration. The model simulated corn yield very well (-0.1% difference) at the calibration site; however, it underestimated yield (-14.1%) at the validation site. Overall, the RZWQM simulated tile drain flow, NO3-N loss to tile drainage water, and crop yield with reasonable accuracy, but tended to underestimate the amount of soil NO3-N (mean deviation, -0.971). The inability of the model to handle the spatial and temporal variability of the soil may have affected its prediction accuracy. The model also needs improvement in simulating early spring snowmelt hydrology.