Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: WATER QUALITY IMPROVEMENT FROM MANAGEMENT PRACTICES IN AGRICULTURAL WATERSHEDS Title: Verifiable metamodels for nitrate losses to drains and groundwater in the corn belt, USA

Authors
item Nolan, Bernard -
item Malone, Robert
item Gronberg, Joann -
item Thorp, Kelly
item Ma, Liwang

Submitted to: Environmental Science and Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 30, 2011
Publication Date: January 17, 2012
Citation: Nolan, B.T., Malone, R.W., Gronberg, J., Thorp, K.R., Ma, L. 2012. Verifiable metamodels for nitrate losses to drains and groundwater in the corn belt, USA. Environmental Science and Technology. 46:901-908.

Interpretive Summary: Nitrate leaching in the U.S. corn-belt poses a risk to groundwater, and potentially to surface water. Computer models such as the Root Zone Water Quality Model (RZWQM2) have been found to accurately estimate nitrate leaching under different soils, weather, and management practices. However, application of mechanistic agricultural system models such as RZWQM2 at large spatial scales is difficult because of the numerous input variables. In this research, we simplified RZWQM2 using a metamodel (MM) for application to the Corn Belt region, USA, evaluated the MM with field measured data, and extrapolated the MM to assess the vulnerability of streams and groundwater in the corn-belt. Metamodels are simplified representations of mechanistic models and exploit relations between model inputs and outputs. The MMs in our case used readily available data comprising fertilizer nitrogen (N), weather data, and soil properties; therefore they were well suited for regional extrapolation. The MM accurately predicted nitrate concentration compared with measured nitrate in 38 samples of recently recharged groundwater. Predicted nitrate generally was higher than that measured in groundwater, which was expected of shallow subsurface nitrate concentrations. The groundwater samples reflect a mixture of ages at the well screen, and older groundwater that likely contains less nitrate. This research will help scientists with United States Geologicial Survey (USGS) identify locations most vulnerable to nitrate leaching for determing more intensive ground- and surface- water sampling and compliment USGS efforts at quantifying and interpolating water quality data across the U.S. corn belt (e.g., SPARROW model).

Technical Abstract: Metamodels (MMs) consisting of artificial neural networks were developed to simplify and upscale mechanistic fate and transport models for prediction of nitrate losses to drains and groundwater in the Corn Belt, USA. The two final MMs predicted nitrate concentration and flux, respectively, in the shallow subsurface. Because each MM was calibrated to tile drainage and leaching data, they represent an integrated approach to vulnerability assessment. Nitrate leaching in the unsaturated zone poses a risk to groundwater, and nitrate in tile drainage is conveyed directly to streams. The MMs used readily available data comprising fertilizer nitrogen (N), weather data, and soil properties; therefore they were well suited for regional extrapolation. The MMs effectively related the outputs of the underlying mechanistic model Root Zone Water Quality Model (RZWQM) to the inputs (R2=0.986 for the nitrate concentration MM). Predicted nitrate concentration was compared with measured nitrate in 38 samples of recently recharged groundwater, yielding a Pearson’s r of 0.466 (p=0.003). Predicted nitrate generally was higher than that measured in groundwater, which was expected of shallow subsurface nitrate concentrations. The groundwater samples reflect a mixture of ages at the well screen, and older groundwater likely contains less nitrate. In a qualitative comparison, predicted nitrate concentration also compared favorably with results from a previous regression model that predicted total N in streams.

Last Modified: 7/30/2014
Footer Content Back to Top of Page