bathymetry file ETOPO2_510x6x510_ver5_filled.bin
for cs510 integrations was generated using mk_bathy3.m
followed by mk_bathy5.m

Levitus initial T and S and monthly SSS files for
cs510 integrations were generated using mk_input.m

These routines are hard-coded for running in
nireas:/home/dimitri/mitgcm/setups/cube/510
Contact menemenlis@jpl.nasa.gov for questions, etc.

=========================================================================
=========================================================================

Files contributed by Jean-Michel on 3/27/2007 and used for computation of
barotropic streamfunction:

psiLine_N2S_cs510.mat
mk_psiLine_CS.m
use_psiLine.m
v_integrate.m
psi_baro.ps.gz

>>>>>>>>>>>>>>

Hi Michael,

I put the *.m files that I was using in:
http://mitgcm.org/~jmc/
(Dimitris: might put them on CVS server later on, 
 may be in MITgcm_contrib/high_res/matlab/something ?)

"mk_psiLine_CS.m" is not really different from 
MITgcm/utils/matlab/cs_grid/bk_line (except kgr=0)

I added the psiLine_N2S_cs510.mat file (from matlab 7, 
don't think it will work with matlab 6).

v_intergrate.m (was part of use_psiLine.m) read 12 
monthly mean + integrate vertical to get the 
annual mean, vertically integrated transport.

and use_psiLine.m is just computing the stream-function 
+ making a plot.

I've just checked-in grph_CSz.m so that it works also for
CS-510 grid.

Jean-Michel

On Tue, Mar 27, 2007 at 06:37:12PM -0400, Michael Schodlok wrote:
> > 
> > Hi Jean-Michel,
> > 
> > thanks for checking this out, just got in and am still a bit
> > puzzled why mk_psiLine_cs.m didn't work here.
> > The *.mat file and modified m-files would be great.
> > Do you want put them on the ECCO2 or  the
> > MITgcm webpage?
> > 
> > cheers
> > 
> > Michael
> > 
>> > >
>> > >I tried "mk_psiLine_CS.m" on cs-510 grid, and it seems
>> > >to work (appart from a stupid plot which you can turn off
>> > >by changing kgr=3; to kgr=0; on the 1rst line).
>> > >I can put the *.mat file (psiLine_N2S_cs510.mat) somewhere 
>> > >if you want to take it.
>> > >
>> > >After I tried to compute a stream-function, using:
>> > >utils/matlab/cs_grid/bk_line/use_psiLine.m
>> > >(had to change the loading of UVELMASS & VVELMASS,
>> > > that was taking too much memory), and it works.
>> > >But the plot at the end is not right (a problem in
>> > >grph_CSz.m that I will look at later).
>> > >Again, if you want me to put those modified script somewhere, 
>> > >just tell me.
>> > >
>> > >I attached here a postscript (but gzip) of the plot, 
>> > >it's average over 1 year.
>> > >
>> > >Jean-Michel
>> > >
>> > >On Tue, Mar 27, 2007 at 09:47:25AM -0400, Jean-Michel Campin wrote:
>> > >  
>>> > >>Hi Michael,
>>> > >>
>>> > >>I will have a look. I remember that I checked the Meridional
>>> > >>stream-function, but no idea regarding this one.
>>> > >>Jean-Michel
>>> > >>
>>> > >>PS: I cc to others who might be interested.
>>> > >>
>>> > >>On Mon, Mar 26, 2007 at 06:07:25PM -0400, Michael Schodlok wrote:
>>> > >>    
>>>> > >>>Hi Jean-Michel
>>>> > >>>
>>>> > >>>I tried to use work out the streamfunction for cube510
>>>> > >>>using calcHorizPsiCube.m and run into all sorts of
>>>> > >>>problems. Well, the main problem I guess is using
>>>> > >>>the mk_psiLine_CS.m routine which seems to be
>>>> > >>>for cube32. I get reshape errors when I use it in
>>>> > >>>split_Z_cub.m. We didn't save the meta files, so
>>>> > >>>I tried to work around that but am a bit lost what
>>>> > >>>routine does what etc. Do you have any suggestions
>>>> > >>>how else to calculate streamfunction in cube?
>>>> > >>>
>>>> > >>>thanks
>>>> > >>>
>>>> > >>>Michael