Project

General

Profile

cdo remapbil on unstructured grid

Added by Aurora Lagos over 3 years ago

Hello,
I have an unstructured Netcdf file. I want to convert it to a regular grid and then interpolate with remapbil. I need to calculate the average of different models and have a common grid of 2 ° x2 °, so I must do this.

Any recommendation, I would appreciate it.

This is the description of the file.

$ ncdump siconc_SImon_AWI-ESM-1-1-LR_lgm_r1i1p1f1_gn_390101-391012.nc | more
netcdf siconc_SImon_AWI-ESM-1-1-LR_lgm_r1i1p1f1_gn_390101-391012 {
dimensions:
time = UNLIMITED ; // (120 currently)
bnds = 2 ;
ncells = 104431 ;
vertices = 18 ;
variables:
double time(time) ;
time:standard_name = "time" ;
time:long_name = "time" ;
time:bounds = "time_bnds" ;
time:units = "days since 3901-1-1 00:00:00" ;
time:calendar = "standard" ;
time:axis = "T" ;
double time_bnds(time, bnds) ;
float siconc(time, ncells) ;
siconc:units = "%" ;
siconc:CDI_grid_type = "unstructured" ;
siconc:_FillValue = 1.e+30f ;
siconc:missing_value = 1.e+30f ;
siconc:description = "Percentage of grid cell covered by sea ice" ;
siconc:coordinates = "lat lon" ;
siconc:standard_name = "sea_ice_area_fraction" ;
siconc:cell_methods = "area: mean where sea time: mean" ;
siconc:cell_measures = "area: areacello" ;
double lat(ncells) ;
lat:standard_name = "latitude" ;
lat:units = "degrees_north" ;
lat:bounds = "lat_bnds" ;
double lat_bnds(ncells, vertices) ;
double lon(ncells) ;
lon:standard_name = "longitude" ;
lon:units = "degrees_east" ;
lon:bounds = "lon_bnds" ;
double lon_bnds(ncells, vertices) ;


Replies (3)

RE: cdo remapbil on unstructured grid - Added by Ralf Mueller over 3 years ago

hi Aurora!

the netcdf dump looks CF-conform. so CDO should be ready to go. Depending on the input grid, you have to choose the best interpolation.

I don't understand, why you distinguish between 'convert' and 'interpolate'. I could just use conservative remapping (operator: remapcon)

 cdo -P 4 -remapcon,global_2 siconc_SImon_AWI-ESM-1-1-LR_lgm_r1i1p1f1_gn_390101-391012.nc lonlat_siconc_SImon_AWI-ESM-1-1-LR_lgm_r1i1p1f1_gn_390101-391012.nc

hth
ralf

RE: cdo remapbil on unstructured grid - Added by Aurora Lagos over 3 years ago

When I try to use remapbil this reported :

cdo remapbil (Abort): Bilinear/bicubic interpolation doesn't support unstructured source grids!

And when I use your recommendation the message is :

cdo remapcon: SCRIP first order conservative weights from unstructured (104431) to lonlat (180x91) grid
cdo remapcon: Target grid area error: 1345 -0.0172689
cdo remapcon: Target grid area error: 1347 -0.0296989
cdo remapcon: Target grid area error: 1352 -0.0280743
cdo remapcon: Target grid area error: 1361 -0.0174829
...

anyway I did change the variable from siconc to siconca. The difference is that the first is from atmospheric grid and the other oceanic. The oceanic grid is F32z grid type and don't have problem to interpolate to regular grid.

Now my problem is about what is the best way, if remapcon or remapbil. I noted that remapcon it's so slowly ...

RE: cdo remapbil on unstructured grid - Added by Ralf Mueller over 3 years ago

you choose: conservative versus bilinear interpolation. check the docu for details. it all depends on what you want to do with the data. conservative remapping is meant to keep the intergral of fields after the interpolation, bilinear doesn't. but you have to decide yourself

    (1-3/3)