[RegCNET] Deciphering OUT_HEAD.CTL

[Maurice.McHugh at noaa.gov]

Hi Xunqiang, Imran, and Chris:

Thanks so much for the concise description of the OUT_HEAD.CTL issue I had. I didn't catch that when reading the pdef page.  I have two follow-up questions:

Is there code within the RegCM package to post-process the domain data contained in OUT_HEAD.CTL to the actual 58x58 grid in netCDF?  When I converted using CDO I received the interpolated data (166x126).

Imran suggested using the regular postproc routine, but this seems to only process ICBC, SFC, ATM, RAD etc. files.  Has anyone hacked postproc.f to do so?  I guess I could always regrid the 166x126 data back to the original 58x58 grid.

On a different topic, using openmpi I have a master node running 12 instances of RegCM3 in parallel on a different 12 core server/node; I want to distribute processing over multiple servers/nodes.  It looks like I need to use a shared filesystem like GFS or NFS for the model to be run in this manner, so that all instances of the model on different servers can write to the same output files.  Does anyone have experience in distributing RegCM's processing over multiple servers/nodes? 

Thanks once again!

Regards,

maurice


----- Original Message -----
From: bixq <bixq at ictp.it>
Date: Thursday, November 19, 2009 9:44 am
Subject: Re: [RegCNET] Deciphering OUT_HEAD.CTL
To: Maurice.McHugh at noaa.gov
Cc: regcnet at lists.ictp.it


> Hi, Maurice:
> 
> The answer is in GrADS documentation :), if you go to the pdef page, 
> you'll see.
> 
> Because you use the Lambert Conformal projection, the high latitude 
> side 
> is stretched far away (in a fan shape), GrADS have to do interpolation 
> to
> a lat-lon grid, here xdef and ydef are the definitions of this lat-lon
> grid (I call it view-grid), actually, if you prepare two GrADS 
> descriptor files,
> 
> one with
> 
> pdef 58 58 lcc 38.89 -77.00 29.00 29.00 30.00 60.00 -77.00 10000. 10000.
> xdef  166 linear  -80.65  0.0450
> ydef  126 linear   36.11  0.0450
> 
> The other with
> 
> pdef 58 58 lcc 38.89 -77.00 29.00 29.00 30.00 60.00 -77.00 10000. 10000.
> xdef  42  linear  -80.65 0.180
> ydef  32  linear   36.11 0.180
> 
> And if you focus on the corner of your plots, I am sure you will 
> understand what I said.
> 
> Regards,
> Xunqiang Bi
> 
> On Thu, 19 Nov 2009, Maurice.McHugh at noaa.gov wrote:
> 
> > Hello everyone,
> >
> > I'm trying to understand why in OUT_HEAD.CTL the pdef line states my 
> correct grid sizes (nx,ny), but the xdef and ydef lines which should 
> state the same grid size are different.
> >
> > For example,
> > <snip>
> > pdef   58   58 lcc   38.89  -77.00   29.00   29.00   30.00   60.00  
> -77.00  10000.  10000.
> > xdef  166 linear  -80.65  0.0450
> > ydef  126 linear   36.11  0.0450
> > </snip>
> >
> > In the above case the 58x58 nature of my grid is correctly specified 
> in pdef, but not in xdef nor ydef which give a 166x126 grid.
> >
> > In order to better understand the file I went ahead converted the 
> data to netCDF using CDO (climate data operators) and plotted markers 
> at the intersection of xlong and xlat from OUT_HEAD.CTL but when 
> plotted they produced a straight line.  I'm unsure whether this is a 
> CDO translation artifact, or if my understanding of the contents of 
> the OUT_HEAD.CTL file is woefully inadequate.
> >
> > Any insight to correctly extracting usable data from this file would 
> be much appreciated!
> >
> > Regards,
> >
> > Maurice
> >
> > _______________________________________________
> > RegCNET mailing list
> > RegCNET at lists.ictp.it
> > 
> >
> 
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>    Dr. Xunqiang Bi         email:bixq at ictp.it
>    Earth System Physics Group
>    The Abdus Salam ICTP
>    Strada Costiera, 11
>    P.O. BOX 586, 34100 Trieste, ITALY
>    Tel: +39-040-2240302  Fax: +39-040-2240449
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> _______________________________________________
> RegCNET mailing list
> RegCNET at lists.ictp.it
>