[MITgcm-devel] changes with external_forcing_surf.F

[Jean-Michel Campin]

Hi,

I made few changes in and around external_forcing_surf.F
(remove bi,bj from list of arguments, and move the call outside bi,bj
loop in do_oceanic_phys.F). I hope I did not miss something
(but if you see something strange, let me know).

Should not change the results, except:
 - when using pkg/offline, we now call EXTERNAL_FORCING_SURF (was not
   called before, so that PTRACERS_FORCING_SURF call was missing).
 - with ALLOW_BALANCE_RELAX (and balanceThetaClimRelax=T or 
    balanceSaltClimRelax=T), the adjustment is now seen by KPP and
   also by PTRACERS_FORCING_SURF (for the case PTRACERS_addSrelax2EmP=T)
 - it seems that few AD experiments are giving slightly different results,
   but it's at the machine truncation level:
    global_ocean.90x40x15               (AD-grad: 13 digits)
    global_ocean.cs32x15.seaice_dynmix  (AD-grad: 11 digits)
    global_ocean.cs32x15.thsice         (AD-grad: 15 digits)
    tutorial_global_oce_biogeo          (AD-grad: 13 digits)
will update the output of global_ocean.cs32x15.seaice_dynmix tomorrow.

Cheers,
Jean-Michel