[MITgcm-devel] exf #undef ALLOW_ATM_WIND
Jean-Michel Campin
jmc at ocean.mit.edu
Sat Apr 14 18:27:16 EDT 2007
Hi Martin and others,
I have few questions about exf pkg (most of them related to
the case with #undef ALLOW_ATM_WIND):
it's stil not clear to me when ustress,vstress is on C-grid and
when it's on A-grid. (did not find much documentation about this).
It seems to depend on #ifdef stuff (e.g. in exf_mapfields.F,
#if (defined (ALLOW_BULKFORMULAE) || defined (USE_EXF_INTERPOLATION))
and also to evolve with time:
In the old version of exf_wind.F & exf_bulkformulae.F
(before I made the changes which broke Martin's set-up, i.e., before Jan 10)
seems that they were on A-grid (and I did not change this),
but looks like now they moved to C-grid (#ifndef ALLOW_ATM_WIND)
in those 2 S/R, but not consistently every where (e.g., not in exf_mapfields.F).
Also, EXCH are applied to lot of fields before exf_mapfields.F
and also after (seems to me that, e.g., Short-Wave heating does not need
to be exchanged 2 times), but some EXCH are missing:
with #undef ALLOW_ATM_WIND, need to exchange ustress,vstress before
calling exf_bulkformulae.F
Right now, with #undef ALLOW_ATM_WIND, the restart is broken (1+1<>2)
But if I add the missing exch (before exf_bulkformulae.F)
and do not go from A to C (in exf_mapfields.F), I got a good restart:
from tst1+1, with every things turned on:
-- compare cg2d_init_res :
run 1iA:
3.48171824684734E+00
3.12627120041265E+00
run 1iB:
2.80140114995309E+00
2.51996292952868E+00
run 2it:
3.48171824684734E+00
3.12627120041265E+00
2.80140114995309E+00
2.51996292952868E+00
Jean-Michel
More information about the MITgcm-devel
mailing list