of the source and plume migration, and are not easily measurable. In particular, measured groundwater concentrations provide only limited insight into the processes responsible for the persistence of dissolved contaminant plumes because it is difficult to distinguish the relative influence of flow field heterogeneity, back-diffusion, and desorption.
The potential importance of back-diffusion is supported by conceptual and modeling analysis (e.g., MacKay and Cherry, 1989; Wilson, 1997; Parker et al., 2008) and a limited number of field investigations that have directly sampled aquitard material (Ball et al., 1997; Chapman and Parker, 2005). Sorption processes are typically included in contaminant transport models and estimates of time to remediate, although the common use of the retardation factor reflects the optimistic assumptions of a single sorbent and rapid linear partitioning. A considerable body of research over the past two decades has demonstrated that, for many aquifer materials, sorption processes are in fact spatially heterogeneous, nonlinear, and potentially limited by solute diffusion to sorbent material located within the interior of soil particles (e.g., as reviewed by Allen-King et al., 2002). As with back-diffusion, conceptual and modeling analyses have shown that nonlinear and/or rate-limited desorption can potentially contribute to plume persistence over decades (e.g., Ball and Roberts, 1991; Rabideau and Miller, 1994; Rivett et al., 2006). However, at the time of this writing, there is a lack of field data and characterization techniques to distinguish desorption processes from other nonideal effects. A modest step toward better understanding the potential role of sorption processes would be to routinely characterize the organic content of collected soil samples (Simpkin and Norris, 2010), a task that could be accomplished at relatively low cost.
Understanding whether back-diffusion and desorption are occurring at a site is challenging because the relative importance of each process is highly dependent on the site-specific contamination history and the presence and distribution of low-permeability and/or strongly sorbing materials. And yet, current site characterization techniques typically do not fully delineate the structure of these materials, particularly when they are distributed over small spatial scales within the plume interior. Furthermore, there are no proven remedial techniques to preferentially target and accelerate the removal of contaminants from localized sites that are desorption/diffusion limited. Finally, currently used mathematical models are difficult to configure to provide realistic predictions of time to remediation when desorption/diffusion processes are the limiting factor because of the need to assign initial conditions that properly represent the mass located in immobile compartments. Additional research is needed to develop strategies for long-term management that focus on plume zone processes that contribute to plume longevity rather than the processes that occur in the source zone.