PNNL

Abstract

The large and diverse body of subsurface characterization data generated at a field research site near Oyster, VA provides a unique opportunity to test various approaches for characterizing field-scale heterogeneity in aquifer properties and modeling subsurface flow and transport. We are using observed bromide breakthrough curves (BTCs) from an injection experiment conducted in 1999 as a baseline for evaluating data worth and model effectiveness. BTCs are available at 24 multi-level samplers, eight ports each (192 total sampling points). Each BTC is a time series of measured concentrations, spaced two to twelve hours apart over the seven-day field experiment. A detailed model, implemented using the RAFT code, is used to simulate breakthrough curves at the sampler locations. This model requires the specification of spatial distributions of hydrologic parameters such as hydraulic conductivity. This in turn involves the integration of data of various types and amounts into a conceptual model framework. The number of possible conceptualizations and methods for data integration is nearly limitless, and each gives rise to a different prediction of bromide breakthrough at sampling points. To evaluate the relative appropriateness of each approach, and the value of the data utilized therein, we simulate BTCs at each sampler location and quantitatively compare them to the observed BTCs.