ENHANCED SYSTEM MODELS AND DECISION SUPPORT TOOLS TO OPTIMIZE WATER LIMITED AGRICULTURE
Location: Agricultural Systems Research Unit
Title: Predicting Unsaturated Zone Nitrogen Mass Balances in Agricultural Settings of the United States
Submitted to: Journal of Environmental Quality
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: December 2, 2009
Publication Date: February 22, 2010
Citation: Nolan, T., Puckett, L., Ma, L., Green, C.T., Bayless, E.R., Malone, R.W. 2010. Predicting Unsaturated Zone Nitrogen Mass Balances in Agricultural Settings of the United States. Journal of Environmental Quality. 39(3):1051-1065.
Interpretive Summary: The U.S. Geological Survey’s National Water Quality Assessment Program, in collaboration with the USDA ARS, used automated calibration methods with a new version of the Root Zone Water Quality Model (RZWQM2) to estimate nitrogen mass balances in diverse agricultural settings: an almond orchard in the Merced River study basin, California; and corn-soybean rotations in study basins at Maple Creek, Nebraska, and at Morgan Creek, Maryland. The model was also applied with minimal, manual calibration at a cornfield in the Merced River basin. The work is novel for several reasons. First, RZWQM typically is calibrated manually by adjusting model parameters one at a time. The autocalibration method used here is more objective and allows simultaneous adjustment of many parameters. Second, previous versions of RZWQM were limited to a simulation depth of 3 m. RZWQM2 enables predictions to 30 m and this is the first test of the deeper capabilities of the new model. Lastly, RZWQM2 was shown to effectively simulate nitrate transport and fate at diverse agricultural settings spanning the conterminous U.S. Results indicate that predicted nitrate concentration and soil characteristics agreed reasonably well with measured data throughout the range of simulation profiles (to 10 m). Thus, the model shows promise for studies of the deeper unsaturated zone. Predicted nitrogen losses occurred primarily through plant uptake (144 – 237 kg N/ha) and deep seepage out of the profile (40 – 102 kg N/ha). Large reservoirs of organic nitrogen (up to 22,600 kg N/ha at Nebraska) were predicted to occur in soil humus pools, which may have implications for future transfer of nitrate to groundwater.
Unsaturated zone N fate and transport were evaluated at four sites to identify the predominant pathways of N cycling: an almond orchard and cornfield in the lower Merced River study basin, California (CA); and corn-soybean rotations in study basins at Maple Creek, Nebraska (NE) and at Morgan Creek, Maryland (MD). We used inverse modeling with a new version of the Root Zone Water Quality Model (RZWQM2) to estimate N mass balances throughout the unsaturated zone; previous versions were limited to 3 m depth and relied on manual calibration. The root mean square error (RMSE) and index of agreement (d) between predicted and observed nitrate concentrations in lysimeters were 8.9 – 12.3 mg/L and 0.60 – 0.71, respectively, for CA and NE, indicating that RZWQM2 showed promise for deeper simulation profiles (6.5 – 10 m). For the shallow simulation profile (1 m) in MD, RMSE and d for nitrate were 3.54 mg/L and 0.86. Predictions of average nitrate concentration at the water table agreed reasonably well with measured concentrations in nearby monitoring wells. The largest additions of N were predicted to come from inorganic fertilizer (153 – 195 kg N/ha in CA) and N fixation (99 – 131 kg N/ha in MD and NE). Predicted N losses occurred primarily through plant uptake (144 – 237 kg N/ha) and deep seepage out of the profile (40 – 102 kg N/ha). Large reservoirs of organic N (up to 22,600 kg N/ha at NE) were predicted as storage in soil humus pools, which may have implications for future transfer of nitrate to groundwater.