Problem in Defining the Water Table

Honghong,

For a sloped water table, attach Dirichlet boundary elements (i.e., elements with very large volumes) at the upstream and downstream boundaries of your model domain at the desired elevations of the water table, and give them fully liquid-saturated conditions and pressures of one atmosphere.

For fluctuating water tables, you need to make these boundary conditions time dependent. This is somewhat tricky in regular TOUGH2, where you need to combine large element volumes with appropriate, large injection/production rates (see Section 7.2, specifically Equation 27 in the TOUGH2 manual). You can specify time-dependent Dirichlet boundary conditions very conveniently in TOUGH3 and iTOUGH2 (just provide a table of pressures vs. time; see Section 6.4 in the iTOUGH2 "Enhancement" manual).

Of course, if the fluctuations are caused by time-dependent recharge, injection, or pumping rates within the model itself, the pressure changes should occur "naturally", i.e., without the need to specify time-dependent boundary pressures.

Stefan