That would be one of the drawbacks of using finite element analysis.
You could assume uniform loading under the base plate. Or linearly varying loading if there is a net moment.
For a small plate on larger concrete slab, you could assume that the mid-surface of the concrete was fixed, and treat the concrete above that as elastic in the vertical direction only.
You could include the concrete as part of the finite element problem also, which introduces other complications.
In a lot of cases, for the base plate design, assuming uniform loading on the bottom will be conservative.
Note that treating a subgrade as an elastic foundation (elastic vertically) can give considerably different results from treating it as an elastic solid. So it's an approach that is useful, but not necessarily an ideal solution, either.
If you have to use very crude approximations to get a solution to a problem, you should consider using very crude approximations that also simplify the problem. So perhaps uniform bearing and a yield-line analysis would be in order.