Thomas F. Gross
NOAA/NOS/OCS/Coast Survey Development Lab.
(301)713-2809 x139
&
Chesapeake Community Model Program,
Chesapeake Research Consortium
(410) 798-1283
attached mail follows:
Hi Sylvia and Dennis:
I am still using your cool NCL and NCO tools. But I think I
forgot a fundamental. I have downloaded a dozen Eta wind grib files.
Then I chew on them individually with some NCL to turn them into a dozen
netcdf files. Each has a scalar, single value for time (dim ncl2=1).
Then I concatenate them using
ncecat output*.nc total.nc
That creates an extra index called "record" I rename that to "time"
and squeeze the extra dimension from my time variable.
ncwa -a ncl2 total.nc total2.nc
ncrename -d record,time total2.nc ETA.nc
That's great. But I have a few variables which are not so nice.
The lon, lat variables which start as 2 dimensional become 3 dimensional
with a time dimension. The grid is not a time dependent thing so this
is rather annoying.
How can I squeeze the time dimension out of just the lon and lat?
Thanks, tom g
PS
If you'd like to try to run it do:
cc dateformat.c -o dateformat
Put it into your path. Its used for scripting date formats.
I am a Linux kinda guy, so I use wget and mktemp. Not sure if
they exist on SGI's etc.
Next run these (sh or bash shell)
etatime=`MostRecentEta.sh meso-eta_218`
NCLwebeta.sh "$etatime" meso-eta_218_
PPS the second NCL call does some rotations of the Eta lambert projected
wind directions to Eastward and Northward. I attempted a couple of
things in there to change the lon,lat dimensions and gave up.
Thomas F. Gross
NOAA/NOS/OCS/Coast Survey Development Lab.
(301)713-2809 x139
&
Chesapeake Community Model Program,
Chesapeake Research Consortium
(410) 798-1283
_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2b29 : Mon Apr 26 2004 - 23:24:02 MDT