DATA ASSIMILATION AT THE HANFORD 300 AREA (Invited Presentation)

The persistent uranium plume at the Hanford 300 Area presents a challenging real-world scenario for inverse modeling and uncertainty quantification due to the complexity of transient groundwater flow driven by an extremely dynamic Columbia River stage and uncertainty in uranium distribution within the vadose and saturated zones. At the Integrated Field Research Challenge (IFRC) site within the 300 Area, researchers have conducted several field experiments injecting tracer and alternate water chemistry to invert for hydrologic properties and evaluate the geochemical response of sorbed uranium to stimulus. This work presents the application of a Bayesian data assimilation framework employing ensemble Kalman filters (EnKF) and high performance computing (HPC) to improve the hydrologic and geochemical conceptual models through assimilation of field characterization data (e.g. pump test, wellbore flow meter test) and injection experiment results (observed tracer and uranium concentrations). This research demonstrates the power of data assimilation with EnKF and HPC to reduce uncertainty in field applications.