[MITgcm-devel] global_ocean_ebm verification exp.

[Jean-Michel Campin]

Hi Patrick and others,

I was trying to fix the multi-threading for experiment global_ocean_ebm  
(not yet done) and got few surprises:
1) useRealFreshWaterFlux=.TRUE., but the Run-Off is not part of
 EmPmR (which stands for Evap mins Precip minus RunOff) so that
 the run-off does not contribute to local sea-level changes.
2) I am used to see run-off for the continents that flows into
 the ocean. But here, the run-off file "Runoff.bin" contains positive 
 values but it's added with the same sign as Evap, as if water
 was flowing upward the rivers:
 (scale_runoff = 3.170979198E-11 > 0 and in ebm_forcing_surf.F,
lines 80 to 83:
          surfaceForcingS(i,j,bi,bj) =
     &      EmPmR(i,j,bi,bj)*convertFW2Salt*mass2rUnit
     &      +Run(i,j,bi,bj)*scale_runoff
     &         *convertFW2Salt*rhoConstFresh*mass2rUnit
3) input file SST.bin is not used (but it's read-in several times): 
 to test it, I changes in "data"
> thetaClimFile=  'SST.bin',
 to:
> thetaClimFile=  'Topog.bin',
 and got the same results.

I am thinking of fixing (1) and putting the right sign for (2),
which implies updating the output. But at the same time, 
I wonder if it's not better to use binary input files from 
tutorial_global_oce_latlon (also used in global_ocean.90x40x15)
since: 
a) in global_ocean_ebm/input/data periodicExternalForcing
  is set to TRUE (expecting 12 records in forcing fields) but the
  global_ocean_ebm input files only contain 1 record.
b) compared to global_ocean.90x40x15, the bathymetry file "Topog.bin" 
 from global_ocean_ebm seems to be shifted by 1/2 grid point towards
 the South (an old problem that was fixed in global_ocean.90x40x15
  a long time ago).
I could just use the same "prepare_run" script from 
 global_ocean.90x40x15/input and keep the Runoff.bin file.

Patrick, what is your opinion on this ?

Cheers,
Jean-Michel