[RegCNET] Re: Clarification about 'postproc.in'

Sara Rauscher srausche at ictp.it
Tue Feb 28 18:27:51 CET 2006


Hi,

Regarding reading netcdf files produced by postproc in Grads --
("unfortunately, the Netcdf files, which you get from postproc, are not
readable in Grads") -- I have never had problems reading netCDF files
created with postproc with Grads. Svetlana's problem with the "no
discernable X coordinate" may be specific to her files -- Svetlana,
would you send me the netCDF file and/or additional information? Maybe I
can help.

Grads should read these netCDF files in, but the problem is that they
don't contain the correct lat/lon information. You can create a data
descriptor file that will supply the correct lat/lon information for
your netcdf file to Grads.

For example, I created the following file using postproc: ATM200208AVG.nc

If I use the ncdump utility (ncdump -h ATM200208AVG.nc) I can view the
file information:

netcdf ATM200208AVG {
dimensions:
        lon = 49 ;
        lat = 32 ;
        level = 18 ;
        time = UNLIMITED ; // (1 currently)
variables:
        float lon(lon) ;
                lon:long_name = "Longitude" ;
                lon:units = "degrees_east" ;
                lon:actual_range = -4.888423f, 32.60124f ;
        float lat(lat) ;
                lat:long_name = "Latitude" ;
                lat:units = "degrees_north" ;
                lat:actual_range = 37.03973f, 54.29553f ;
        float level(level) ;
                level:long_name = "Height_Index" ;
                level:units = "level" ;
                level:actual_range = 1050.f, 0.f ;
        double time(time) ;
                time:long_name = "Time" ;
                time:units = "hours since 1900-1-1 00:00:0.0" ;
                time:actual_range = 899544., 899544. ;
        float U(time, level, lat, lon) ;
                U:long_name = "Zonal Wind" ;
                U:units = "m/s" ;
                U:missing_value = -1.e+30f ;
        float V(time, level, lat, lon) ;
                V:long_name = "Meridional Wind" ;
                V:units = "m/s" ;
                V:missing_value = -1.e+30f ;
        float TK(time, level, lat, lon) ;
                TK:long_name = "Temperature" ;
                TK:units = "K" ;
                TK:missing_value = -1.e+30f ;
        float QD(time, level, lat, lon) ;
                QD:long_name = "Mixing Ratio" ;
                QD:units = "kg/kg" ;
                QD:missing_value = -1.e+30f ;
        float QC(time, level, lat, lon) ;
                QC:long_name = "Cloud Mixing Ratio" ;
                QC:units = "kg/kg" ;
                QC:missing_value = -1.e+30f ;
        float MSE(time, level, lat, lon) ;
                MSE:long_name = "Moist Static Energy" ;
                MSE:units = "m2/s2" ;
                MSE:missing_value = -1.e+30f ;
        float RH(time, level, lat, lon) ;
                RH:long_name = "Relative Humidity" ;
                RH:units = "fraction" ;
                RH:missing_value = -1.e+30f ;
        float HGT(time, level, lat, lon) ;
                HGT:long_name = "Geopotential Height" ;
                HGT:units = "m" ;
                HGT:missing_value = -1.e+30f ;
        float PS(time, lat, lon) ;
                PS:long_name = "Surface Pressure" ;
                PS:units = "hPa" ;
                PS:missing_value = -1.e+30f ;
        float RT(time, lat, lon) ;
                RT:long_name = "Total Precip" ;
                RT:units = "mm/day" ;
                RT:missing_value = -1.e+30f ;
        float TGRND(time, lat, lon) ;
                TGRND:long_name = "Ground Temperature" ;
                TGRND:units = "K" ;
                TGRND:missing_value = -1.e+30f ;
        float SMT(time, lat, lon) ;
                SMT:long_name = "Total Soil Water" ;
                SMT:units = "mm" ;
                SMT:missing_value = -1.e+30f ;
        float RB(time, lat, lon) ;
                RB:long_name = "Base Flow" ;
                RB:units = "mm/day" ;
                RB:missing_value = -1.e+30f ;

// global attributes:
                :domxmin = -4.888423f ;
                :domxmax = 32.60124f ;
                :domymin = 37.03973f ;
                :domymax = 54.29553f ;
                :domzmin = 1050.f ;
                :domzmax = 0.f ;

With information about the model grid (from OUT_HEAD.CTL), I can take
this information and make a data descriptor file that I can give any
name I like (say, atmavg). Once I write the descriptor file, I can open
the file in grads using "open atmavg", as indicated below:

DSET ^ATM200208AVG.nc
dtype netcdf
undef -1.e+30
pdef   49   32 lcc   45.39   13.48   24.50   16.00   30.00   60.00   13.48  60000.  60000.
xdef  175 linear   -9.61  0.2703
ydef   78 linear   34.70  0.2703
zdef 18 levels 74.08  122.24 175.22 237.83 310.08 387.14 464.20 541.26 618.32 695.38 767.62 830.23 883.21 926.56 960.27 984.35 998.80 1008.43
1008.43
tdef  1 linear  0z01aug2002  30dy
vars 13
U=>u          18  t,z,y,x westerly wind 
V=>v          18  t,z,y,x southerly wind 
TK=>tk        18  t,z,y,x air temperature 
QD=>qd        18  t,z,y,x air specific humidity
QC=>qc        18  t,z,y,x cloud water mixing ratio
MSE=>mse      18  t,z,y,x moist static energy
RH=>rh        18  t,z,y,x relative humidity
HGT=>hgt      18  t,z,y,x geopotential height
PSA=>psa      0   t,y,x   surface pressure 
RT=>rt        0   t,y,x   total precipitation
TGRND=>tgrnd  0   t,y,x   groud temperature in BATS
SMT=>smt      0   t,y,x   total soil water in mm H2O
RB=>rb        0   t,y,x   accumulated infiltration
endvars


Note that descriptor files (for netcdf!)  work with Grads version 1.9
and above ONLY - you can see the documentation at
http://grads.iges.org/grads/gadoc/descriptorfile.html.

I hope this is helpful, please let me know if you encounter problems or
it doesn't work.

cheers,
sara


Moetasim wrote:

>Hi,
>As far as I know...unfortunately, the Netcdf files, which you get from 
>postproc, are not readable in Grads. You have two options:
>1) Use ferret/NCL instead of Grads
>2) Write CTL file for files which you get after postprocessing in Grads format.
>Writing CTL file should not be very tricky. YOu can use ncdump to check the 
>contents/list of variables of Netcdf postproc files..The sequence of variables 
>in Grads formatted files will be same. You should use any ATM???.CTL file as 
>template for ATM postproc file and similarly RAD and SRF CTL files as template 
>for writing CTL files for your postprocessed Grads formatted files. Remember 
>that you just need to change variable info and time info in CTL files...Rest 
>should almost remain same...
>Try writing one.. If you are successful then do post your CTL files at 
>RegCNET. 
>And if you are not successful then................. you can ask again..or may 
>be someone else have already tried for such CTL files so he can post for you...
>I hope you will give a try to any of the options I have mentioned above.
>
>
>Quoting ?????? <dolgikh at meteo.kz>:
>
>  
>
>>Dear Moet.
>>
>>Thanks a lot for clarification about cru.param.
>>
>>But we face one more problem and can not go further.
>>
>>We have a problem with opening NetCDF files (ATM199703AVG.nc, etc) in GRads.
>>
>>The error is:
>>ga-> sdfopen ATM199703AVG.nc
>>Scanning self-describing file: ATM199703AVG.nc
>>SDF file has no discernable X coordinate.
>>
>>But when we try to open PRE199703.nc everything is ok.
>>May be it is possible to change parameters in postproc.in to get coordinates
>>
>>for output .nc file?
>>Is there description of postproc.in file anywhere?
>>it is possible to get output files in GRADS format (.dat), but we can't 
>>recieve .ctl file with .dat file.
>>
>>I would be very grateful if you agree to help us again.
>>Thank you in advance
>>Best regards
>>Svetlana
>>
>>
>>    
>>
>
>_______________________________________________
>RegCNET mailing list
>RegCNET at lists.ictp.it
>https://lists.ictp.it/mailman/listinfo/regcnet
>  
>

-- 

Sara A. Rauscher
Physics of Weather and Climate Section 
The Abdus Salam International Centre for Theoretical Physics
Strada Costiera 11 
34014 Trieste ITALY
ph: +39 040 2240 225 (please note change) fax: +39 040 2240 449
email: srausche at ictp.it





More information about the RegCNET mailing list