Time stepping with 'infinite volume' boundaries

Hello,

I was wondering if you all could help me understand the time stepping of a soil model I’m building. I would like to model annual soil carbonate formation and dissolution dynamics in a soil. However, my model has a time step of <60 s, which seems very slow from the practice problems.

I believe that the small time steps have something to do with my ‘infinite volume’ node at the bottom of the soil column (model described below), because when I set it to a non-infinite box size the model runs at a much faster time step. This works until the column fills with water and then the model reasonably crashes with nowhere to put water. However, I have to have the ‘infinite volume’ bottom node so that water has somewhere to go. Is there something I am missing about initializing the model that causes the very small time steps?

The model is attached and the setup is as follows:

Setup

The soil is modeled as a 1D column with 3 parts

-           Atmosphere: 1 ‘infinite’ node

-          Soil: Differentiated into three horizons (physically the same for now, but chemically distinct in calcite and gypsum content)

-          Bottom of soil: 1 ‘infinite’ node so water has somewhere to go

GENER inputs

-          Water production: intermittent ‘rainfall events’ (input to the top soil block)

-          Heat: A diel cycle that mimics the observed temperature swings (input to the top soil block)

-          CO2: A diel cycle that mimics respiration rates (input at 11 cm depth)

-          Water extraction (transpiration): Unused at the moment and always set to 0 production

-          7 years of inputs.

Thank you,

Ty