Remapcon gives weird pattern
Added by Lina Teckentrup almost 5 years ago
Hi,
I'm trying to regrid JSBACH output from a coarse resolution to a finer half degree grid. For some reason I get weird lines (attached a figure) that don't appear for other models when I use the same grid file and command. The lines also disappear for other remap commands. I attached two files (original resolution and regridded as well as the grid file). I don't get an error or anything, does anyone understand why this is happening? My command was
cdo remapcon,fine_grid.txt ifile ofile
Thanks a lot!
JSBACH_oz_2018.nc (4.4 KB) JSBACH_oz_2018.nc | |||
JSBACH_oz_2018_halfdegree.nc (29.3 KB) JSBACH_oz_2018_halfdegree.nc | |||
fine_grid.txt (290 Bytes) fine_grid.txt | |||
grid_jsbach.jpeg (64.3 KB) grid_jsbach.jpeg |
Replies (5)
RE: Remapcon gives weird pattern - Added by Karin Meier-Fleischer almost 5 years ago
Hi Lina,
which CDO version are you using? I can't reproduce the problem. The higher parts between blue and orange are the results of the interpolation between values from coarse 1.875/1.865252 degrees input file to the much finer 0.5 degrees output grid.
-Karin
gpp_in_oo.png (74.8 KB) gpp_in_oo.png |
RE: Remapcon gives weird pattern - Added by Lina Teckentrup almost 5 years ago
That's weird. I'm using CDO version 1.8.0
RE: Remapcon gives weird pattern - Added by Lina Teckentrup almost 5 years ago
Sorry so your figure is the plot of data you regridded right? Because for me it looks like this when I read in the JSBACH_oz_2018_halfdegree.nc file with panoply (interpolation off)
gpp_halfdegree.png (59.7 KB) gpp_halfdegree.png |
RE: Remapcon gives weird pattern - Added by Karin Meier-Fleischer almost 5 years ago
It is the remapcon output file 0.5x0.5 degrees displayed with panoply.
The same for ncview
No lines. So I would guess that you have to update your CDO to the current release.
ncview_oo.png (14.2 KB) ncview_oo.png |
RE: Remapcon gives weird pattern - Added by Lina Teckentrup almost 5 years ago
So odd, I just tried with an older version and it seems to be fine then. This effect also only occurred for two models out of four? Anyway, thanks for your help!