[netCDF #AZT-127023]: Can't ncdump a file...

Hi Tom,

> We have a NetCDF file which ncdump fails to recognize as a
> valid file.  I can open the file with the java-netcdf library using
> toolsUI-4.0.jar.  I'm not sure if this is a v3/4 or 32/64bit
> compatibility problem.  We found that neither NetCDF-3 or 4
> will ncdump this file:
> 
> ftp.ssec.wisc.edu/pub/incoming/MSG_SEVIRI_2008157_1200_AWG_SST.nc
> TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION.nc

You've provided 2 file names, so I'm not sur which one is causing
ncdump problems.

I downloaded the first one and ncdump seems to recognize it as a
netCDF-4 problem and read the metadata just fine:

 $ ls -l MSG_SEVIRI_2008157_1200_AWG_SST.nc
 -rw-r--r--   1 russ     ustaff   220481910 Jul  1 09:16 
MSG_SEVIRI_2008157_1200_AWG_SST.nc
 $ ncdump -k MSG_SEVIRI_2008157_1200_AWG_SST.nc
 netCDF-4
 $ ncdump -h MSG_SEVIRI_2008157_1200_AWG_SST.nc
 netcdf MSG_SEVIRI_2008157_1200_AWG_SST {
 dimensions:
         Columns = 3712 ;
         Rows = 3712 ;
 variables:
         float Latitude(Columns, Rows) ;
                 Latitude:long_name = "Latitude" ;
                 Latitude:units = "degrees" ;
                 Latitude:comments = "Pixel latitude (degree)" ;
         float Longitude(Columns, Rows) ;
                 Longitude:long_name = "Longitude" ;
                 Longitude:units = "degrees" ;
                 Longitude:comments = "Pixel Longitude (degree)" ;
         float SST_Regression(Columns, Rows) ;
                 SST_Regression:Units = "K" ;
                 SST_Regression:Comments = "Sea Surface Temperature (SST) 
derived from SEVIRI BTs using regression technique" ;
                 SST_Regression:long_name = "sst regression" ;
                 SST_Regression:_FillValue = -999.999f ;
                 SST_Regression:missing_value = -999.999f ;
         float SST_Physical(Columns, Rows) ;
                 SST_Physical:Units = "K" ;
                 SST_Physical:Comments = "Sea Surface Temperature (SST) derived 
from SEVIRI BTs using inversion (physical) method" ;
                 SST_Physical:long_name = "sst physical" ;
                 SST_Physical:_FillValue = -999.999f ;
                 SST_Physical:missing_value = -999.999f ;
                 SST_Physical:valid_range = 0.f, 400.f ;

 // global attributes:
                 :Conventions = "CF-1.4" ;
                 :Title = "AIT SST Product" ;
                 :Source = "AIT_Framework" ;
 }

Similarly for the second file, I can't duplicate the problem you are
seeing:

 $ ls -l TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION.nc 
 -rw-r--r--   1 russ     ustaff   27503920 Jul  1 09:25 
TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION.nc
 you have mail in /var/mail/russ
 $ ncdump -k TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION.nc
 netCDF-4
 $ ncdump -h TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION.nc
 netcdf TERRA_MODIS_2005155_3257_AWG_AEROSOL_DETECTION {
 dimensions:
         Columns = 1354 ;
         Rows = 2030 ;
 variables:
         float Latitude(Columns, Rows) ;
                 Latitude:long_name = "Latitude" ;
                 Latitude:units = "degrees" ;
                 Latitude:comments = "Pixel latitude (degree)" ;
         float Longitude(Columns, Rows) ;
                 Longitude:long_name = "Longitude" ;
                 Longitude:units = "degrees" ;
                 Longitude:comments = "Pixel Longitude (degree)" ;
         byte Smoke(Columns, Rows) ;
         byte Dust(Columns, Rows) ;
 }

I'm using the latest ncdump from our snapshot release, but I don't
think that's relevant here, unless you've identified a bug in an
older version of ncdump we don't know about.

What's the ncdump version (check the last line of the output from
ncdump --version)?

--Russ

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu



Ticket Details
===================
Ticket ID: AZT-127023
Department: Support netCDF
Priority: Normal
Status: Closed