Quantifying the uncertainty in nitrogen application and groundwater nitrate leaching in manure based cropping systems

Abstract Uncertainty in application rates of nitrogen (N) limits the effectiveness of nutrient management plans that aim to realize yield goals while minimizing environmental N losses. Measuring N application rates accurately is especially difficult in manure dependent cropping systems. California dairies use both solid manure and process wastewater (PWW) to fertilize forage crop fields. California dairy farmers measure the ratio of N applied to N removed from each crop field, the N Ratio, based on many uncertain measurements of the quantity and concentration of applied N sources. In this study, a Monte Carlo based model was used to propagate the uncertainty of each application measurement to the overall N application rate. Then the Monte Carlo model was used to stochastically vary the N application rate around farmer measured values for two forage fields (A: silty clay soil, B: sandy loam soil) over a 5 year period. The N Ratios and groundwater nitrate concentrations were simulated for 1000 stochastic N application scenarios using the HYDRUS-1D water and solute transport model and the results were compared to monitoring well measurements. Results showed that uncertainty in measurements of N applied in PWW were the dominant source of uncertainty in total N application rates, contributing 64 to 94% of overall uncertainty. N Ratio uncertainty varied widely between harvests. The average upper and lower limits of the 95% confidence interval deviated from the median simulated N Ratio by 0.25 (range: 0.13 to 0.48) and 0.27 (range: 0.12 to 0.55), respectively, among all harvested crops. Simulated groundwater nitrate concentrations were lower than monitoring well measurements by an average of 20 and 2 mg L−1 for Fields A and B, respectively. All simulated and observed nitrate concentrations in groundwater recharge from both fields were three to eight times the federal maximum contaminant level for nitrate as N of 10 mg L−1. Determining and reducing uncertainty in measurements of N applied in PWW is likely the most impactful way to improve accuracy of overall N application rates. Within current measurement and management systems, measurements of N application in manure dependent cropping systems are not accurate enough to identify application rates that will both support yield goals and reduce environmental N losses.