Impacts of spatial and temporal recharge on field-scale contaminant transport model calibration

Summary This paper investigates the role of spatial and temporal averaging of recharge upon the calibration of trichloroethylene (TCE) contaminant transport parameters for a transient, three-dimensional, variably saturated, physically based, field-scale groundwater flow system in Toms River, New Jersey. Coarse vertical discretizations of unsaturated groundwater systems are prone to numerical difficulties requiring, in some cases, the linearization of the van Genuchten or Brooks Corey equations. A flux-based method for calculating relative permeability was developed and shown to improve numerical stability. To reduce the computational burden of calibration, the simulation time of the transport models was reduced by 95% by decoupling the transport model from the flow model. Multi-start PEST was used to calibrate four models with recharge specified as: constant, spatially varying, temporally varying, and both spatially and temporally varying. For the transient flow model with spatially and temporally varying recharge, the estimated transverse dispersivity coefficients were estimated to be significantly less than those calculated from simulations with spatial and/or temporal averaging of recharge. The calibrated TCE retardation value of 1.58 is within 5% of the laboratory determined value of 1.65. Furthermore, the calibrated TCE retardation values are not as sensitive to the spatial and/or temporal averaging of recharge as compared to transverse dispersivity.