WHAT'S IN A PLUME MODEL AND WHY SHOULD WE CARE?

Much of contaminant hydrogeology is fixated on the problem of predicting plume movements toward receptors. Models that appear to mimic field observations of oblong shaped plume geometries and smoothly decreasing concentrations away from a source or a central maximum concentration are assumed to be broadly reasonable. Fickian or similar models based on homogeneity or modest heterogeneity (e.g., ln K variance < 2) can produce such geometries, and manipulation of transport parameters and source terms in these models can even provide good looking fits to sparse field data. An important but often overlooked characteristic of field plumes is that they take decades longer to remove by pumping (remediation) than they took to migrate into the groundwater systems, even when the contaminants are conservative. A strong indicator that homogeneous or modestly heterogeneous models may be generally inappropriate is the fact that they are usually not capable of simulating plume recovery via pumping because they cannot adequately account for sequestration of plume mass in low-conductivity media. Thus the appropriateness of plume modeling approaches needs to be based not only on capability for modeling downgradient migration, but also capability for modeling mass sequestration and remediation. This change in research emphasis is needed not only for better plume modeling, but also for moving beyond the plume into the realm of groundwater quality sustainability.