Interpolation Problem using remapcon2
Added by SUMAN MAITY over 5 years ago
Dear All
I faced a new problem while interpolating emission data (0.25deg x 0.25deg) to T63grid (~1.8deg x 1.8deg). I interpolated using cdo (second order conservation remapping) as follows:
cdo remapcon2,t63grid dom.nc dom_remapcon.nc
After interpolation, while exploring the interpolated files (cdo infov infile), I found that negative emission value is introduced at some grid points and therefore minimum value becomes negative. I doubly checked my original file where minimum value was '0' and there was no -ve values. You can check it from the attached files.
I couldn't understand how to get rid of this issue. from where the -ve values are introduced? Am I doing some wrong steps?
It is also non-physical to interpret that emission is negative.
Please help/suggest me what to do. Any sort of suggestion will be highly appreciated.
Best
Suman
| dom_remapcon.nc (870 KB) dom_remapcon.nc | After using cdo remapcon2 | ||
| dom.nc (11.9 MB) dom.nc | Original file |
Replies (1)
RE: Interpolation Problem using remapcon2 - Added by Uwe Schulzweida over 5 years ago
The second order remapping with remapcon2 is using the gradients of the field. If a field to be remapped has strong gradients, numerical approximations to the gradient may be too steep. In such cases, the expected value of the field may overshoot.
Use a more stable method like the first order conservative remapping with remapcon to get rid of this issue.
Cheers,
Uwe