Correction of convective cloud condensate in prog closure for HR2

[lisa-bengtsson]

This PR updates the prognostic closure in samfdeepcnv.f and samfshalcnv.f for HR2:

1. correction (bug-fix) of convective cloud condensate passed to progsigma_calc.F90 routine
2. Update to only consider positive buoyant levels in integral (to make it general if a convection scheme can have negative buoyant levels between cloud base and cloud top)
3. Update to make the prognostic closure general to pass in q-tendency due to turbulence from any PBL scheme.
4. Move out computation of q-tendency due to dynamics from progsigma_clac.F90 (to samfshalcnv and samfdeepcnv) because other convection schemes compute this term in a pre/post step. I will update samf* routines to use pre/post convection in the future.

[grantfirl]

@lisa-bengtsson ufs-community/ccpp-physics#57 was merged. This is ready to revert the changes in .gitmodules (which it looks like you already did) and update the ccpp/physics submodule pointer to the latest commit (after the ccpp-physics PR was merged) in ufs/dev.

[jkbk2004]

@lisa-bengtsson ccpp pr was merged . can you update ccpp submodule pointer? correct hash is ufs-community/ccpp-physics@57c444f

[jkbk2004]

@lisa-bengtsson branch is pointing to ccpp 03acf73 hash. correct one is 57c444f

[lisa-bengtsson]

@jkbk2004 Now it looks to be correct. Thank you.

[lisa-bengtsson]

@jkbk2004 is this merged?