Is that ground surface, at top edge of the brown?
We're told not to try to model soil in Therm, because its temp changes dynamically with the seasons, and under the influence of the heat input from the building.
But I don't see why not, if you're sure you know that the soil holds a stable, unchanging temp, at a certain distance down from the interior floor surface.
Or distance back in from the interior wall face, if it's your retaining wall.
For example, if you know there's a water table below, or permeating water in the soil outside the retaining wall, then what temp is that water? Say 5C seasonal average?
In the case of the underfloor water table, say seasonal average 1.2m below floor level, then that could be your boundary condition - 5C at 1.2m below FL. This would extend 'to infinity' beyond the external walls - along with the boundary condition that is the ground surface, close above.
In the case of the retaining wall built against soil full of permeating water, perhaps the entire ground would be 5C, right up to back of the lower part of the retaining wall - but as you come up to the surface, ground temp against the upper part of the retaining wall will increase (or decrease), more influenced by surface temp. I could prob explain how to model that?