• journal_title：Vadose Zone Journal
• Contributor：J. E. Smith ; E. J. Henry ; R. Bashir
• Publisher：Soil Science Society of America
• Date：2011-
• Format：text/html
• Language：en
• Identifier：10.2136/vzj2009.0081
• journal_abbrev：Vadose Zone Journal
• issn：1539-1663
• volume：10
• issue：1
• firstpage：403
• section：ORIGINAL RESEARCH

Most organic compounds, when dissolved in water, decrease the surface tension of the water and many change the wettability of porous media. This directly affects capillary pressures and can induce water flow and solute transport in the vadose zone. This study investigated a coupled unsaturated flow and transport phenomenon, solute-dependent capillarity-induced focused flow (SCIFF), that has significant implications for infiltration into contaminated unsaturated zones. A two-dimensional laboratory flow cell and the two-dimensional hysteretic saturated–unsaturated flow and transport numerical model HYDRUS-2D, which was modified to include solute concentration-dependent hydraulic functions, were used to demonstrate and quantify the effect of SCIFF on point-source infiltration of clean water into contaminated sand. The difference between the SCIFF infiltration event and traditional water infiltration into uncontaminated sand is striking. Highly focused vertical flow with very little horizontal migration was characteristic of the SCIFF infiltration event from a constant-flux point source. The SCIFF resembles a single finger as in classic flow instability fingering; however, the SCIFF system has the added complexities of aqueous solute concentration-dependent hydraulic functions. The SCIFF mechanism may well be active at many sites where the vadose zone has significant contamination with organic compounds. Another significant finding was that the magnitude and trend in moisture content changes measured by time domain reflectometry (TDR) did not match point values from the simulations because the TDR sample area extended beyond the width of the SCIFF zone. A good match between TDR and simulated values was obtained by determining the average simulated moisture content within the same region sampled by the TDR.