Evaluation of Drainmod Nii and SWAT in Simulating Subsurface Drainage and Nitrate Losses in Minnesota.

Sensor-based variable rate nitrogen fertilizer application allows for fertilizer application at fine spatial and temporal resolution, and is a promising approach for reducing nitrate-N loadings to surface waters.  However, little research has been done to determine the effectiveness of this approach in reducing nitrate-N losses for fields with subsurface (tile) drainage in the Upper Midwest.  Hydrologic models provide an effective way to simulate the impacts of management practices on soil and water resources, and represent a viable way to examine the long-term effect of variable rate fertilizer management practices on nitrate losses from tile-drained agricultural fields.  Drainmod--and its companion model Drainmod NII--are widely used in simulating subsurface drainage systems.  New tile drainage algorithms recently added to SWAT (the Soil Water and Assessment Tool) have made it a candidate for simulating nitrogen losses from subsurface drainage systems as well, though few studies have been published to date using the new algorithms.  The work presented here evaluates both Drainmod NII and SWAT based on how well they represent tile flow and nitrate loadings from corn/soybean fields with varying drainage intensities in Waseca, MN.  Early results indicate that both models are more sensitive to infiltration parameters than to drainage parameters.  The timing of drainage flow simulated by Drainmod NII more closely matches field measurements than the timing of drainage simulated by SWAT, however both models show similar performance based on Nash-Sutcliffe modeling efficiencies.  Factors that would make a model suitable for use in a variable-rate fertilizer management study are evaluated with both models, including: the seasonal timing and magnitude of drainage and nitrate loads, ability to spatially and temporally vary fertilizer management practices within a field, and ease of gathering model input data.