lake modeling

???
Is there a way to model the filling of a lake or reservoir in feflow? Modflow has a lake package that can be used to model lakes.

I realize that you could model the flux out of the groundwater, and use an external water balance program. However, I would like to model the effect of the lake on the groundwater system. So, I would prefer to use a coupled system in feflow.

Another possibility would be to include the lake in the model mesh, and then set the lake porosity to 1 and the K to some high value.

The coupled surface/groundwater modeling capability would be useful for river modeling, too.
Thanks,
8)

There are several possibilities. The classical one is using a high k and a porosity/storativity of 1. A general package for doing a water level - volume calculation for the lake is currently not available. However, I attach an example source code for an additional module for FEFLOW (via the interface manager IFM), which deals with such an external, simplified lake. Maybe that could be helpful. The pdf file contains an article with the same topic, but unfortunately this is only available in German.

I am modelling the filling of a lake/mine pit, as the watertable reaches the floor, I change the permeability and porosity to high values.
I don't worry about the porous media below the lake, as the high K and n of the lake control the behaviour.
Previously I have done this in two stages with two models, copying the head from the first to the second. 
I tried doing this in one model, using time varying K and porosity.  Changing the porosity appears to collapse the head to reflect the greater pore space to conserve water mass.  Is there a way to avoid this, by altering the water storage in the pore space by the correct amount to not change the hydraulic head?

reply to KC: an article in english attached

reply to Mark: I wonder if you have a timing problem in which the K and S of the pit are increased too soon. Consider using more gradual changes in K and S, and remember that FEFLOW averages material properties over time, so that for a simulation time step that falls between one value of K and another, FEFLOW computes an intermediate value of K that is a linear interpolation between the two specified K's (this depends on which averaging method you choose, and the averaging approach applies to all materials).

Pete
www.aquageo.com