This project was conducted in collaboration with GNS Science in New Zealand. Nitrite contamination in groundwater from intensive agriculture can degrade groundwater quality making it unusable as a drinking water source. This project involved a large-scale model predictive uncertainty quantification with nonlinear, calibration-constrained, uncertainty analysis, on a coupled MODFLOW/MODPATH/MT3DMS numerical model that simulated groundwater flow, contaminant travel times and chemical dispersion pathways. More than 1,000 adjustable parameters were considered and the nitrate transport risk at various receptors was quantified.