Project

General

Profile

cdo gencon aborts with error num_subseg exceeded limit

Added by Torsten Albrecht about 7 years ago

I am converting 1km Bedmap2 Antarctic ice geometry data (2GB) to a 15km grid. Conservative mapping works for coarse resolution (20, 30, 50km) but not for the finer ones (15,12,10,7,...km)...

cdo gencon,targetgrid_15km.nc bedmap2_1km_input.nc bedmap_15km_weights.nc

_cdo gencon: SCRIP first order conservative weights from curvilinear (6667x6667) to curvilinear (400x400) grid
_cdo gencon: 6%
_cdo gencon (Abort): Integration stalled: num_subseg exceeded limit (grid16136960: lon1=3.14677 lon2=3.14677 lat1=-1.18782 lat2=-1.18767)!

What could be the reason?
Thanks,
Torsten

cdo sinfo:

targetgrid_15km.nc
1 : curvilinear : points=160000 (400x400)
lon : -179.898 to 179.795 degrees_east
lat : -89.9349 to -50.1077 degrees_north
available : cellbounds
data:
x = -2795000, -2780000, -2765000 .... 3160000, 3175000, 3190000 ;
y = -2795000, -2780000, -2765000, .... 3160000, 3175000, 3190000 ;

bedmap_1km.nc
Grid coordinates :
1 : curvilinear : points=44448889 (6667x6667)
lon : -179.983 to 180 degrees_east
lat : -90 to -48.4644 degrees_north
available : cellbounds
data:
x = -3333000, -3332000 .... 3332000, 3333000 ;
y = -3333000, -3332000 .... 3332000, 3333000 ;


Replies (2)

RE: cdo gencon aborts with error num_subseg exceeded limit - Added by Uwe Schulzweida about 7 years ago

Hi Torsten,

We have reimplemented the conservative remapping in CDO because it doesn't work correctly for some grid combinations
Please try the new operators genycon and remapycon.

Cheers,
Uwe

RE: cdo gencon aborts with error num_subseg exceeded limit - Added by Torsten Albrecht about 7 years ago

Thanks Uwe,
this seems to work...
Cheers,
Torsten

    (1-2/2)