remapping from fine to coarse grid
Added by Tom Akkermans almost 13 years ago
Dear all,
We want to compare model output (0.25°) with its ERA-interim boundary conditions (1.50°). It is always told me that one should remap from fine to coarse resolution, that means that we have to remap the model output to 1.50° resolution. Now in CDO there are some options like bilinear (remapbil) and bicubic (remapbic) interpolation. Is it true that they only use respectively 4 and 9 surrounding 0.25° grid points to calculate the new value of the 1.50° grid points? In that case, these methods are not sufficient since the resolution is 12 times bigger. Hence, some pixels of the 0.25° output are not used to calculate the 1.50° pixels. And hence, I am looking for a proper method which takes into account all pixels of the original model output in the remapping procedure. Any ideas? thanks very much!
.
Replies (2)
RE: remapping from fine to coarse grid - Added by Uwe Schulzweida almost 13 years ago
Hi Tom,
Yes, thats true, the bilinar interpolation uses only 4 surrounding grid points. For the bicubic interpolation 16 surrounding points are used.
The conservative remapping with remapcon takes all source grid points into account!
Best wishes,
Uwe