Hi Gurus,

I try to simulate the accumulative impact of soak hole-disposal of septic tank effluent on groundwater quality (NO3 and faecal coliform) through a stony soil profile and coarse gravel aquifer. In a 4-km area, there are many soak holes (about 4 m deep and 1 m diameter) and lots of drinking water wells (20-40 m deep).

The size of model domain will be 4000 m wide and 100 m deep. Compared to the size of the model domain, sockholes and wells are very small. I wish to generate a mesh that could be appropriate for Hydrus calculations. My questions are:

1. Should sockholes and wells are best inserted as circles in the model domain at the depths where their bottoms are?

2. For the boundary points: what would be the desirable numbers of points on the outside curve and around the sockholes and wells, respectively?

3. What would be the desirable density of boundary points (left and right) for outside curve and around sockholes and wells?

4. I am thinking to use atmosphere upper BC, no flux lower BC, and constant pressure heads for the L and R BCs. For the inner boundaries, I will use constant flux for sockholes and wells. Does this sound right?

5. Whether it would be easier to split the problem with two models? The first model has a depth of 4 m with free drainage at lower BC. The second model gets rid of soil layers above 4 m and thus the sockholes can be assigned at the top boundary as constant flux upper BCs. The areas between sockholes are atmosphere upper BCs using the cumulative fluxes at the bottom of the first model as “rainfall”. There will have then no evaporation in the second model. I hope this approach may avoid the irregular mesh problem that I question in 1-3. How do you think?

6. The sockholes are developed at 4 m deep. Should I still consider denitrification? How could HYDRUS model denitrification?

Wish to hear from you. Thanks a lot in advance.

Best regards

Liping

liping pang