2013 journal article

Comparison of the performances of DRAINMOD-NII and ADAPT models in simulating nitrate losses from subsurface drainage systems

AGRICULTURAL WATER MANAGEMENT, 129, 21–30.

co-author countries: United States of America 🇺🇸
author keywords: DRAINMOD; Hydrology; Nitrate losses; Tile drainage; Water quality
Source: Web Of Science
Added: August 6, 2018

Adequate knowledge on the movement of nitrate-nitrogen (NO3-N) under different subsurface (tile) drain configurations and management practices in the U.S. Midwest is essential for developing remedial measures for reducing hypoxic conditions in the Gulf of Mexico. In this study, DRAINMOD-NII, a daily time-step soil carbon (C) and N model, was calibrated and validated for subsurface drainage and associated NO3-N losses, and crop yield. Long term (1983–1996) monitoring data measured on three experimental plots under continuous corn (Zea mays L.) with conventional tillage practice at the University of Minnesota's Southern Research and Outreach Center near Waseca, southern Minnesota was used for this purpose. Nash-Sutcliffe efficiency (NSE), Percent Error (PE) and Index of agreement (d) were used for assessing the model performance. DRAINMOD-NII predicted monthly subsurface drainage matched well with measured data during calibration (NSE = 0.81, PE = −7.8% and d = 0.94) and validation (NSE = 0.67, PE = −0.7% and d = 0.88) periods. Performance of DRAINMOD-NII for predicting monthly NO3-N losses in subsurface drainage was also good for both calibration (NSE = 0.64, PE = 0.8%, and d = 0.85) and validation (NSE = 0.62, PE = −5.3%, and d = 0.83) periods. DRAINMOD-NII predicted average (1983–1992) annual corn relative yield (93%), a ratio of crop yield in a year to the long-term average crop yield, was close to the observed relative yield (92.5%). DRAINMOD-NII simulation results were also compared and contrasted with those obtained by the Agricultural Drainage and Pesticide Transport (ADAPT) model with the same dataset. Both models performed equally well in predicting monthly subsurface drainage. However, DRAINMOD-NII performed slightly better in predicting monthly NO3-N losses and annual N budget, in addition to showing potential to simulate the effects of excess and deficit water stresses on crop yield. Studies comparing performances of different drainage models in the U.S. Midwest are useful to select an appropriate model for devising various strategies for reducing NO3-N losses from subsurface drainage systems, and thereby minimizing hypoxic conditions in the Gulf of Mexico.