[MITgcm-support] questions

[Jean-Michel Campin]

Hi Andreas,

In short:
On Wed, Oct 10, 2012 at 11:30:41AM +1100, Andreas Klocker wrote:
> Jean-Michel,
> Two questions for you...
> 1) you told me in your last email that I shouldn't be using netcdf
> for pickup files - why's that??
a) Since the pickup files are just for internal use (to restart the model,
   but not to do output analysis or plots), there is no real advantages
   to have them in NetCDF format.
b) writing NetCDF file is slower than plain binary file.
c) MITgcm cannot read global NetCDF files and does not write
   global file neither with "globalFiles=TRUE" nor with useSingleCpuIO=T. 
   As a consequence, if you change the tile size (sNx, sNy), not only
   you will have to gather all tiled pickup from the old tile-size decomposition
   but you will have also to split the global fields into new-tiled pickup file.
d) because of a,b, and c (+ the fact that I still have problems to 
   figure out some of the pkg/mnc pieces), some improments in the restart
   procedure have been implemented for binary pickup but not for NetCDF pickup.

> 2) is the submesoscale parameterisation of Baylor and Raf
> implemented in MITgcm?
I believe yes (in fact, twice: 1rst time in Contrib, 2nd time in main
 repository):
 in pkg/gmredi: submeso_calc_psi.F
 run-time logical switch: GM_useSubMeso
 tested in experiment: verification/front_relax/input.bvp

> see ya,
> Andreas


Cheers,
Jean-Michel

> 
> -- 
> ===============================================
> Dr. Andreas Klocker
> Research Fellow
> 
> Center of Excellence for Climate System Science
> &
> Geophysical Fluid Dynamics Group
> Research School of Earth Sciences
> Room J7.221, Jaeger 7 Building
> The Australian National University
> Canberra ACT 0200 Australia
> 
> T:     +61 2 6125 0920
> M:     +61 437 870 182
> W:     http://web.mit.edu/aklocker/www/
> skype: andiklocker
> ===============================================
>