[MITgcm-support] how to write

Jean-Michel Campin jmc at ocean.mit.edu
Thu Oct 28 14:36:01 EDT 2010


Hi Nataliya,

It looks like you did not specify which region you want
in output file "u2" (and in this case, the default is just 
the region "0", i.e., the global statistics).
adding a line like:
> stat_region(1,1)= 1,
 (i.e., the 1rst region you defined) 
in the same namelist should do it.

I can also mentionned that the output file is either 
ASCII file (easier may be to start with) or NetCDF. 
And the netcdf file always contains all the regions, 
but only selected regions have meaningful values.

Jean-Michel

On Thu, Oct 28, 2010 at 01:01:35PM +0100, Nataliya Stashchuk wrote:
> Hi Jean-Michel,
> 
> I tried to follow your advice and found that I still have some questions.
> I am interesting what value of velocity would be in point x=199500 with time.
> Here is my file:
> 
> &DIAG_STATIS_PARMS
> #- regional mask: band: 1 : x <= 199500 ; 2 : x == 199500
>  diagSt_regMaskFile='region1.bin',
>  nSetRegMskFile=1,
>  set_regMask(1)= 1,  
>  val_regMask(1)= 2., 
> #- an example just to check the agreement with MONITOR output:
>  stat_fields(1,1)= 'UVEL    ',
>  stat_fname(1)= 'u2',
>      stat_freq(1)= 25,
>     stat_phase(1)= 0.,
> &
> 
> 1) It seems to me that the programme does not understand where my point is.
> 
> My matlab file is:
> 
> prec='real*8';
> ieee='b';
> nx=1024;
> ny=1;
> nz=85;
> xc=rdmds('XC');
> regMsk=ones(size(xc));
> regMsk(find (199500 == xc))=2;
> fid=fopen('region1.bin','w',ieee); fwrite(fid,regMsk,prec); fclose(fid);
> 
> 2) My output file gives following: Average, Std.Dev, min, max, Vol. Where should be the data for my point?
> 3) I use 85 levels in vertical. In my output I need to know velocity at 2 or three levels. How I can organize this?
> 
> Many thanks,
> 
> Nataliya
> 
> -----Original Message-----
> From: Jean-Michel Campin [mailto:jmc at ocean.mit.edu] 
> Sent: 27 October 2010 20:45
> To: mitgcm-support at mitgcm.org
> Subject: Re: [MITgcm-support] how to write
> 
> Hi Nataliya,
> 
> You can use the STATDIAGS (part of diagnostics pkg) with a region
> that cover only 1 grid point and, setting the stat_freq equal to
> the time-step, will produce output at 1 grid point every time-step.
> This is not well documented in the manual, but there is an example
> using STATDIAGS in: verification/aim.5l_cs/input.thSI
> can check data.diagnostics there and matlab script mk3regions_mask.m 
> 
> Thanks,
> Jean-Michel
> 
> On Wed, Oct 27, 2010 at 07:28:20PM +0100, Nataliya Stashchuk wrote:
> > Dear mitgcm users,
> > 
> > Can you give me an advise how to arrange an output file which should contain a physical value, i.e. velocity, for instance, in a single specific point, but written at every temporal step, i.e.  with delT frequency?  I need this time series for a spectral analysis.
> > 
> > Thank you,
> > 
> > Nataliya
> 
> > _______________________________________________
> > MITgcm-support mailing list
> > MITgcm-support at mitgcm.org
> > http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support



More information about the MITgcm-support mailing list