Modelling nitrate losses in drainage water using DRAINMOD 5.0

• Published in: Agricultural water management Vol. 56; no. 2; pp. 153 - 168
• Main Authors: Helwig, Trevor G., Madramootoo, Chandra A., Dodds, Georges T.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 30.07.2002; Elsevier Science; Elsevier
• Series: Agricultural Water Management
• Subjects: Agronomy. Soil science and plant productions; Applied sciences; Biological and medical sciences; Continental surface waters; Drainage; DRAINMOD; Environment; Exact sciences and technology; Fundamental and applied biological sciences. Psychology; Leaching; Modelling; Natural water pollution; Nitrogen; Pollution; Soil and water pollution; Soil science; Subirrigation; Water table management; Water treatment and pollution; Canada; North America; America; Quebec; Canada, Quebec; Drainage; Environment; DRAINMOD; Water table management; Modelling; Subirrigation; Nitrogen; Leaching; Macronutrient(mineral); Nitrogen fertilization; Underground leaching; Model study; Nitrates; POLLUTION BY AGRICULTURE; Modeling; Pollutant; Subsurface irrigation; Water pollution; Simulation model; Nitrogen fertilizer; Drainage water; Environmental protection
• ISSN: 0378-3774; 1873-2283
• DOI: 10.1016/S0378-3774(02)00005-7

The appropriateness of the DRAINMOD-N computer model for the southwestern Quebec region was assessed by comparing field data and simulated results. Four years of field data were collected from a 4.2 ha field research facility located at St. Emmanuel, Que. The experimental layout contrasted two different fertilizer and water table treatments in a split plot design. Nitrogen fertilizer was applied at rates of 120 and 200 kg N/ha. The water table was either maintained at 0.6 m by a subirrigation (SI) system or was under free drainage (FD) at a drain depth of 1.0 m. The model was calibrated using data from 1996 to 1997. Simulated output for 1998 and 1999 were compared to measured values based on the mean residual error (MRE), coefficient of efficiency ( C eff=1.0 represents a perfect match; C eff≤0 a poor match) and other statistical parameters. Except for FD plots in 1998 (MRE=−0.225 m, C eff=−0.198), water table depths (WTDs) were predicted to 0.17 m (23%) with C eff ranging from 0.035 to 0.846. DRAINMOD correctly predicted the pattern, but underestimated the magnitude of daily drainflow in both years. On a monthly basis total drain flow was underestimated by up to 5 mm in 1998 and 1.2 mm in 1999. Except for the 1998-SI plots ( C eff=−0.021) simulations were excellent ( C eff>0.8). With the exception of two FD-year-N combinations, DRAINMOD overestimated leached nitrogen by 0.01–1.09 kg ha −1 month −1 (0.02–5.01%). Nitrate leaching predictions were poor in 1998 (−0.20> C eff>−7.83), but good to excellent in 1999 (0.95> C eff>0.56). The poor performance in 1998 compared to 1999 is attributed to significant rainfall events occurring soon after surface fertilizer application in 1998, but not in 1999. DRAINMOD largely overpredicted monthly total denitrification. Predictions were poor in 1999 (−0.25> C eff) and very poor in 1998 (−139.7> C eff) and mostly attributable to poor simulation of June and July, the peak months of measured denitrification. DRAINMOD performed well in the prediction of hydrological parameters, but fared less well in terms of nitrate leaching and denitrification.