Techniques and Methods 6-A19
U.S. GEOLOGICAL SURVEY
Techniques and Methods 6-A19
The kinematic-wave approximation used in the UZF1 Package was compared to the USGS’s Variably-Saturated Two-Dimensional Flow and Transport (VS2DT) Model (Healy, 1990). VS2DT solves Richards’ equation numerically in terms of pressure head. The results were limited to the simulation of vertical flow through a column.
Percolation through the unsaturated zone was simulated with VS2DT as a 30-m long vertical column of materials having hydraulic properties typical of a sandy clay loam, whereas percolation through the unsaturated zone was simulated with the UZF1 Package and MODFLOW-2005 for a single cell (row 10, column 4) from the model in test simulation 2, described herein. The initial water table in row 10, column 4 was 22 m below land surface and the water table declined less than 1 m during the simulation period. The same variables that describe the unsaturated zone were used in both model simulations (table 1). The wetting front did not reach the water table for both simulations because the comparison was done to test differences in the propagation of the wetting front with varying ET demand rates. VS2DT simulates evaporation separate from transpiration by root uptake; only transpiration was used in the simulation. Total simulation time was 17.7 days. A variable time step was used in the simulation with VS2DT, whereas a constant time step was used in the UZF1/MODFLOW-2005 simulation.
A constant infiltration rate was applied at the upper boundary of the column and was 2.0 ×10-6 m/s. The infiltration rate was half the saturated vertical hydraulic conductivity (table 1). The constant infiltration rate at land surface was converted to water content within UZF1 on the basis of equation 19. The initial water content was set equal to the residual water content throughout the unsaturated zone. The ET extinction depth was set at a depth of 2 m below land surface, and an ET demand rate of 5.0 ×10-8 m/s, 5.0 ×10-7 m/s, and 1.0 ×10-6 m/s was specified for simulations 1, 2 and 3, respectively (table 1). Although the two highest ET demand rates used in this example are too high to represent any actual conditions, they serve to demonstrate the model’s accuracy during extreme situations.
Advancement and magnitude of the wetting front for all three simulations of differing specified ET rates were nearly the same for both models (fig. 4). The kinematic-wave approximation in UZF1 slightly under-predicted the advancement of the wetting front at early time and over-predicted advancement at later time. The effect was more pronounced when the ET demand rate was 1.0 ×10-6 m/s (fig. 4C). Discrepancies between UZF1 and VS2DT for each simulation were related to neglecting the negative pressure gradients with the kinematic-wave approximation.
U.S.
Department of the Interior | U.S. Geological
Survey
Persistent URL: https://pubs.water.usgs.gov/tm6a19
Page Contact Information: Publications Team
Page Last Modified: Friday, 02-Dec-2016 15:45:20 EST
