Rework irrigation coupling for the sake of water tracers

There is special-purpose mapping of irrigation here:

https://github.com/ESCOMP/CMEPS/blob/806cce1de93ac96dcdc977e30f18c1206e8b6bf6/mediator/med_phases_prep_rof_mod.F90#L441-L680

This currently doesn't handle water tracer fields.

I started thinking about how we want to do the volr-based normalization for tracer fields – such as whether each tracer should normalize by its own tracer's volr or by the bulk volr. But then I realized that we probably should do a more substantial rework of irrigation coupling to ensure the correct handling of tracers - see https://github.com/ESCOMP/CTSM/issues/512#issuecomment-4898296928.

When we address that CTSM issue, we'll also need to rework the CMEPS irrigation coupling, both to handle water tracers and to handle any other changes to the LND-ROF irrigation coupling.

Start with mediator/med_phases_prep_rof_mod.F90 lines 441-680 and read the linked CTSM issue comment about LND-ROF irrigation coupling. Clarify the intended volr normalization and coupling changes before implementation. Done means the reworked CMEPS irrigation coupling handles water tracer fields and the related CTSM coupling changes.

Written by the indexing model from the issue text.