problem with the RegCM4.1 output data
Added by Jinghua Chen over 12 years ago
when i use cdo to handle the RegCM4.1 output NC format files, it omited some variables in the chemistry output files.
i check up the file and found the possible reason maybe is some variables in the NC files have the same name.
how can i deal with this problem.
thanks in advance!!
Chen Jinghua
NUIST, Nanjing, China
chenjh1213@163.com
Replies (3)
RE: problem with the RegCM4.1 output data - Added by Jaison-Thomas Ambadan over 12 years ago
Hi,
Can you please upload a sample of your NetCDF file (must be less than 50MB). It will be very helpful for the CDO guys here to identify the problem a bit faster.
Cheers,
J.
PS: If the file size is large, then you may strip the data to a a region (using "sellonlatbox"), and with fewer time-step.
RE: problem with the RegCM4.1 output data - Added by Jinghua Chen over 12 years ago
i'm sorry that i reply this so late.
i upload a sample data that the RegCM4.1 is output.
when i operate the file the variables(such as 'trac', 'colb' and so on ) in the file can't well deal with.
thank you!
Kuo_CHE.1997110100.nc (44.4 MB) Kuo_CHE.1997110100.nc | chemistry output data of one day |
RE: problem with the RegCM4.1 output data - Added by Jaison-Thomas Ambadan over 12 years ago
Hi Jinghua,
As far as I understand, there are couple of issues with your input file: 1) your NetCDF file is more generic, and the "time" co-ordinate/time-step is not properly (step is 0) defined in the way CDO recognize, 2) the variable "trac" is a 5-dimensional variable, which CDO cannot handle at the moment. see https://code.zmaw.de/embedded/cdo/1.5.4/cdo.html#x1-20001 for more info.
BUT there is a cheat-sheet: you can change the attributes of the "time" co-ordinate variable using the NCO tools http://nco.sourceforge.net/nco.html#Introduction, and then use CDO for your analysis:
1. ncecat -O -h Kuo_CHE.1997110100.nc outfile.nc
2. ncpdq -O -h -a time,record outfile.nc outfile.nc
3. ncwa -O -h -a record outfile.nc outfile.nc
Step 1-3 will create a record dimension, which helps CDO to recognize the time attributes associated with the variables:
Now "cdo sinfov outfile.nc" will show all the variables EXCEPT "trac", which is a 5-dimensional variable
Hope this helps!
Cheers,
J.
PS: There is a nice documentation/examples of NCO tools by Todd Mitchell, where you can find more info: http://jisao.washington.edu/data/nco