Net enthalpy budget diagnostic

We would like to have a "NET ENTHALPY BUDGET" in the diags.log. For now, it should include all common enthalpy terms accounted for by both `ocn` and `atm`, as follows:

| | atm | ocn | **SUM** |
|--------------|-----------|------------|------------|
| hlatvap | XXXXX | XXXXX | XXXXX |
| hlatfus | XXXXX | XXXXX | XXXXX |
| hrain | XXXXX | XXXXX | XXXXX |
| hsnow | XXXXX | XXXXX | XXXXX |
| hevap | XXXXX | XXXXX | XXXXX |
| hcond | XXXXX | XXXXX | XXXXX |
| **SUM** | XXXXX | XXXXX | YYYYY |

These terms should only be computed over ocean points (i.e., they should have `* ofrac(n)`).
`hlatvap` and `hlatfus` are already computed for `atm` and `ocn`. The `atm` values must be modified to only account for ocean points.
Currently, `hrain`, `hsnow`, `hevap`, and `hcond` are only computed for `ocn` (see [here](https://github.com/ESCOMP/CMEPS/blob/d6dc143cad5f965ec4bf1b1979dae6f0e041b9f3/mediator/med_enthalpy_mod.F90#L150-L153)). Corresponding terms must be introduced for `atm` (e.g., `hrain_atm`, `hsnow_atm`, `hevap_atm`, and `hcond_atm` ). These will have a different formulation than the `ocn` counterparts (exact formulation will be provided in the near future), but for now (implementation phase), they can follow `ocn` (see [here](https://github.com/ESCOMP/CMEPS/blob/d6dc143cad5f965ec4bf1b1979dae6f0e041b9f3/mediator/med_enthalpy_mod.F90#L150-L153)).

The goal is to have the sum of these terms balanced (YYYYY = 0).

cc'ing: @jedwards4b, @PeterHjortLauritzen