Effective Soil Hydraulic Parameters for Transient Flows in Heterogeneous Soils

This study investigated the use of effective soil hydraulic properties (expressed in terms of hydraulic parameters) applicable to large-scale transient infiltration problems in a landscape with horizontally heterogeneous soil hydraulic properties. The heterogeneous landscape was conceptualized as an equivalent homogeneous medium with effective hydraulic properties. The main objectives were to investigate: (i) which effective soil hydraulic property schemes are suitable to represent average behavior of large-scale infiltration processes, (ii) how the effective hydraulic parameters are sensitive to the process time frame, and (iii) how hydraulic parameter variability and correlation impact the effective hydraulic parameters. The heterogeneous landscape was represented by a series of vertically homogeneous stream tubes or parallel columns. Large-scale average infiltration behavior in the heterogeneous soils was quantified through Monte Carlo simulations of multiple realizations (stream tubes) of local-scale infiltration. The optimal effective hydraulic parameters were then calculated with an inverse procedure that minimized the difference between average cumulative infiltration and cumulative infiltration based on a single set of effective parameters. Three scenarios were used to optimize either two hydraulic parameters simultaneously or only one hydraulic parameter while using the arithmetic mean for the other parameter. Results indicate that while the effective hydraulic parameters could simulate average infiltration more closely when multiple parameters were optimized together, the effective parameter values were more variable as time evolved. Optimizing only one hydraulic parameter while keeping the arithmetic mean for the other parameter produced more uniform effective hydraulic parameters with time, but this approach did not represent average infiltration behavior of the heterogeneous soils as well as when multiple hydraulic parameters were optimized simultaneously.