[MITgcm-devel] initialisation and mnc

Jean-Michel Campin jmc at ocean.mit.edu
Thu Oct 7 19:28:44 EDT 2004


Hi Ed,

I have a little comment about the initialisation required for mnc:
1) about cd_code:
 it should go (and call from) packages_init_fixed.F
 and I would prefer not to have this piece of code in 
 cd_code_ini_vars.F (called from packages_init_variables.F)
 It's also an advantage for TAF that does not see most of the
 init_fixed part (Patrick, do you agree ?).
2) about time_ave:
 I would call the routine you added "timeave_init.F" with a name that
 clearly reflect that it's mnc or I/O or fixed initialisation.
 The reason is that with this name, someone who see this pkg 
 would guess that the initialisation of the time-average arrays 
 is done within timeave_init.F , which is not the case.

Does someone agree or disagree with those 2 points ?

I know that this ini_fixed / ini_var is a little subtil and hard
to figure out first, but it makes sense in some way.
  
See you,

Jean-Michel



More information about the MITgcm-devel mailing list