gravit not defined in ice_dyn_shared?

[JFLemieux73]

In ice_dyn_shared we have:

#ifndef coupled
! calculate tilt from geostrophic currents if needed
strtltx(i,j) = -fm(i,j)*vocn(i,j)
strtlty(i,j) = fm(i,j)uocn(i,j)
#else
strtltx(i,j) = -gravitumass(i,j)ss_tltx(i,j)
strtlty(i,j) = -gravitumass(i,j)*ss_tlty(i,j)
#endif

I run CICE6 non-coupled so my code compiles and works fine. However, I have the impression it would not work in coupled mode because gravit is not defined.

Do we need a 'call icepack_query_parameters'?

[eclare108213]

@apcraig, would you check this in the version of RASM that includes CICE v6, please? Maybe you already fixed it.

[apcraig]

It looks like neither RASM nor CESM set the "coupled" cpp. They both use the CESMCOUPLED cpp. The coupled cpp is used in limited places in CICE5 and CICE6. We should review that cpp and decide whether it should be removed, changed, or otherwise. The cpp appears in

cicecore/cicedynB/infrastructure/comm/serial/ice_communicate.F90
cicecore/cicedynB/dynamics/ice_dyn_shared.F90
cicecore/cicedynB/general/ice_init.F90
cicecore/drivers/cesm/*/*.F90

[dabail10]

Actually, we do define "coupled" in the CESM build and use both CESMCOUPLED and coupled. We do the following in ice_init.F90:

#ifdef CESMCOUPLED
#define coupled
#endif

We need both for various pieces of code. I agree that we do need to add the parameter gravit for the coupled case. I will do this.

Dave

[dabail10]

This is addressed in PR #182.

[dabail10]

Resolved in PR #191.