Re: Save results to netcdf in NCL

Hi Dave,

Thanks. That solved my problem.

Regards,

Yuqiang

From: David Brown [mailto:dbrown@ucar.edu]
Sent: Friday, June 21, 2013 6:44 PM
To: Yuqiang Zhang
Cc: ncl-talk@ucar.edu
Subject: Re: Save results to netcdf in NCL

Hi Yuqiang,

What you did not do is assign only the data from your variables into the
file using the "(/" , "/)" syntax, as described in the last section of the
"Method 2" doc. Your variables probably do not have any assigned dimension
names. When you assign the variable (as opposed to just the data values in
the variable) it creates dimension names in the file (even though you are
assigning to a variable you have already created in the file).

If you replace lines like

outf->ene = var_regrid_ene

with

outf->ene = (/ var_regrid_ene/)

the extra dimensions should go away.

 -dave

On Jun 21, 2013, at 4:29 PM, "Yuqiang Zhang" <yuqiangzhang.thu@gmail.com>
wrote:

Hi Everyone,

I am a new guy with NCL. Recently I wrote a code to save the outputs into
netcdf with NCL. However when the code was done, I used the "ncdump -h" to
check the header information of the new netcdfs, it gave me more dimensions
than I defined (see the yellow lines). However in the code, I only defined
these dimensions as (time, south_north, west_east). In the new netcdfs, I
have saved 8 variables with the same dimension of (south_north, west_east).
Not sure why there are so many "ncl*" variables.

The red ones are my codes. It is also attached.

"dimensions:

        time = UNLIMITED ; // (1 currently)

        south_north = 128 ;

        west_east = 164 ;

        ncl3 = 1 ;

        ncl4 = 128 ;

        ncl5 = 164 ;

        ncl6 = 128 ;

        ncl7 = 164 ;

        ncl8 = 128 ;

        ncl9 = 164 ;

        ncl10 = 128 ;

        ncl11 = 164 ;

        ncl12 = 128 ;

        ncl13 = 164 ;

        ncl14 = 128 ;

        ncl15 = 164 ;

        ncl16 = 128 ;

        ncl17 = 164 ;

        ncl18 = 128 ;

        ncl19 = 164 ;"

  fAtt = True ; assign file attributes

 fAtt@title = "NCL Efficient Approach to netCDF Creation"

  fAtt@units = "kg/m2/s"

  fAtt@source_file = src_file

  fAtt@Conventions = "Conserve from RCP"

  fAtt@creation_date = systemfunc ("date")

  fileattdef(outf, fAtt ) ; copy file attributes

  ;Define dimensions:

  dim_names = (/"time", "south_north","west_east"/)

  dim_sizes = (/-1, nlat(0),nlon(1)/)

  dim_unlimited = (/True, False,False/)

  print("Defining dimension sizes to be " + dim_sizes)

  filedimdef(outf,dim_names,dim_sizes,dim_unlimited)

  ;Define file variables

  filevardef(outf,"ROW","float",(/"south_north","west_east"/))

  filevardef(outf,"COL","float",(/"south_north","west_east"/))

  filevardef(outf,"time","float","time")

  filevardef(outf,"ind","float",(/"south_north","west_east"/))

  filevardef(outf,"ene","float",(/"south_north","west_east"/))

  filevardef(outf,"dom","float",(/"south_north","west_east"/))

  filevardef(outf,"tra","float",(/"south_north","west_east"/))

  filevardef(outf,"wst","float",(/"south_north","west_east"/))

  filevardef(outf,"agr","float",(/"south_north","west_east"/))

  filevardef(outf,"slv","float",(/"south_north","west_east"/))

  filevardef(outf,"awb","float",(/"south_north","west_east"/))

  ;Define file attributes

  print("Defining file variable attributes")

  tatts = True

  tatts@long_name = "elapsed time"

  tatts@format = "YYYYMMDD"

  filevarattdef(outf,"time",tatts)

  ;create arrays

  ;filedimdef(outf,"time",ntime,True)

  outf->ROW = dst_lat

  outf->COL = dst_lon

  outf->time = times(0)

  outf->ene = var_regrid_ene

  outf->dom = var_regrid_dom

  outf->ind = var_regrid_ind

  outf->tra = var_regrid_tra

  outf->wst = var_regrid_wst

  outf->agr = var_regrid_agr

  outf->slv = var_regrid_slv

  outf->awb = var_regrid_awb

I used the example from NCL website:

"Method 2: netCDF output with file predefinition"

 <http://www.ncl.ucar.edu/Applications/method_2.shtml>
http://www.ncl.ucar.edu/Applications/method_2.shtml

-----------------------------------------------------------------

-----------------------------------------------------------------

Yuqiang Zhang

PhD student

Environmental Science and Engineering Department

Univerisyt of North Carolina at Chapel Hill

<zyq_ESMF_regrid_RCP_SMOKE.ncl>_____________________________________________
__
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
 <http://mailman.ucar.edu/mailman/listinfo/ncl-talk>
http://mailman.ucar.edu/mailman/listinfo/ncl-talk

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri Jun 21 21:06:38 2013