Comments on 'Pore-Scale Visulization of Colloid Transport andRetention in Partly Saturated Porous Media'

The recent study by Crist et al. (2004) attempted to provide pore scale insights into mechanisms responsible for controlling colloid transport in unsaturated porous media. However, because they relied on images obtained along surfaces that were open to the atmosphere, artificial evaporation resulted in 2 more critical artifacts; formation of air-water-solid (AWS) contact lines, and advection/deposition of colloids to AWS contact lines. These evaporation-related artifacts need to be addressed because they account for most of the colloid deposition at AWS contact lines reported in Crist et al. (2004)...As stated in Crist el al. (2004), ''... the front panel was removed to avoid light reflections that obscured the view and, thus, exposed one side of the sand column to air''. Although a more recent paper (Crist et al., 2005) also presents results using the same methods and is therefore also affected by evaporation, we will restrict our present comments to Crist et al. (2004). Here, we show that removal of the front panel results in a sequence of three critical artifacts; (1) significant evaporation, (2) drying of thin films and formation of air-water-solid (AWS) contact lines, and (3) advection of colloids to AWS contact lines where they are deposited. As explained below, these artifacts so drastically disturbed their system that the magnitude of their observations are not likely to occur anywhere except within the most superficial few cm of soils. Before explaining these artifacts, we note that although trapping of colloids at AWS contact lines reported in Crist et al. (2004) is largely an artifact of evaporation, colloid filtration within perimeters of pendular rings is in fact a main prediction of the film straining model (Wan and Tokunaga, 1997). In that model, colloid filtration is predicted to be more efficient below a critical water saturation, when capillary connections between pendular rings become separated by adsorbed water films. In that paper we stated that ''Retardation of ideal, nonsorbing colloids can occur at two locations: trapped within individual pendular rings due to exclusion from entry into surrounding thin films and within film...'' (Wan and Tokunaga, 1997). Thus, while Crist et al. (2004) implied that the film straining model applies only to retardation of colloid transport within thin films, colloid retention within perimeters of pendular rings is a main feature of our model