Project

General

Profile

Conservative remapping from fine to coarse resolution

Added by François Roberge over 6 years ago

Hi all,

I am trying to remap a global 0.08° global sand and clay percentage dataset to a global 0.15° grid using conservative remapping.

By doing this :

cdo -remapcon,global_0.15 -div ifile.nc landmask.nc landdata.nc

land mask has been created by taking the datasets and by dividing it by itselft (values were 0 over water)

However, when I get the following error messages when interpolating (I only copied here a few lines..) :

cdo remapcon: Map weight < 0! grid1idx=2642978 grid2idx=2062321 nlink=3715256 wts=-4
cdo remapcon: Map weight < 0! grid1idx=2642979 grid2idx=2062321 nlink=3715259 wts=-4
cdo remapcon: Map weight < 0! grid1idx=2642980 grid2idx=2062321 nlink=3715261 wts=-0.5
cdo remapcon: Map weight > 1! grid1idx=6061476 grid2idx=1007086 nlink=6942599 wts=1.33333
cdo remapcon: Error: sum of wts for map1 1007086 1.5 1
cdo remapcon: Error: sum of wts for map1 2062322 2 1
cdo remapcon: Error: sum of wts for map1 2099703 1.075 1
cdo remapcon: Error: sum of wts for map1 2099770 1.125 1
cdo remapcon: Error: sum of wts for map1 2110348 1.25 1

What do you think could cause that?

I get similar error messages when using remapcon2.

Thank you,

François Roberge


Replies (2)

RE: Conservative remapping from fine to coarse resolution - Added by Karin Meier-Fleischer over 6 years ago

Hi Francoise,

it's hard to say what's going wrong. Is it possible to upload the data (1 timestep, 1 variable)?

-Karin

RE: Conservative remapping from fine to coarse resolution - Added by Uwe Schulzweida over 6 years ago

It is well know that the SCRIP implementation of the conservative remapping (remapcon) doesn't work correctly for some grid combinations. Please try the new implementation with remapycon.

    (1-2/2)