CDO

Hi CDO team,

I've been using CDO successfully to convert Lambert Conformal Conic NetCDF files to GRIB2 like this:

cdo -f grb2 copy in.nc out.grb

However, I now encountered strange behavior where for a 200-m resolution NetCDF file, the resulting GRIB2 grid seems incorrect. Here's the grid description for the source NetCDF file:

$ cdo griddes in.nc
#
# gridID 1
#
gridtype  = projection
gridsize  = 50000
xsize     = 250
ysize     = 200
xname     = x
xunits    = "m" 
yname     = y
yunits    = "m" 
xvals     = -3.72529029846191e-09 199.587681240402 399.945036566816 599.024267312139 
            799.385158726946 998.979852784425 1198.83229321986 1397.66247153655 1598.03042823169 
            ... many rows omitted for brevity
yvals     = 1.62981450557709e-09 199.014222515048 398.70365596097 598.730049135862 
            797.740483938949 1000.11186205386 1199.45798648917 1399.1410636201 1600.15574228903 
            ... many rows omitted for brevity
scanningMode = 64
grid_mapping = Lambert_Conformal
grid_mapping_name = lambert_conformal_conic
standard_parallel = 35.31035f
longitude_of_central_meridian = 75.f
latitude_of_projection_origin = 35.31035f
earth_radius = 6367470.
longitudeOfFirstGridPointInDegrees = 139.215f
latitudeOfFirstGridPointInDegrees = 35.30643f
cdo    griddes: Processed 5 variables [0.00s 51MB].

And here's the grid description for the target GRIB file:

t$ cdo griddes out.grb
#
# gridID 1
#
gridtype  = projection
gridsize  = 50000
xsize     = 250
ysize     = 200
xname     = x
xunits    = "m" 
yname     = y
yunits    = "m" 
xfirst    = 0
xinc      = 1.44
yfirst    = -89.55
yinc      = 0.9
scanningMode = 64
grid_mapping = Lambert_Conformal
grid_mapping_name = lambert_conformal_conic
standard_parallel = 35.310352
longitude_of_central_meridian = 75.
latitude_of_projection_origin = 35.310352
earth_radius = 6367470.
false_easting = -5424972.821748
false_northing = -1820872.47506838
longitudeOfFirstGridPointInDegrees = 139.215027
latitudeOfFirstGridPointInDegrees = 35.306431
cdo    griddes: Processed 5 variables [0.29s 81MB].

Based on xfirst, xinc, yfirst, and yinc, it seems to me that the target grid has x axis linearly spaced from approx. 0 to 360, and y axis from approx. -90 to 90. Instead, I expected xfirst and yfirst to be approx. 0 and xinc and yinc to be approx. 200.

I attach my in.nc and out.grb files. Any help or suggestions would be much appreciated.

Thank you!
Milan