CDO

hi!

you input is VERY far from being nc-conform. So CDO is unlikely to be useful with it. I recommend to use gdal

cheers
ralf

Thank you Ralph!
-Bob

Hi Bob,

I've tested to remap the variable ACHA_mode_8_Tc_Lower_Cloud from the file input.nc to a global 0.05x0.05 grid and it works quite well, but you have to do some steps before.

1. add coordinates attribute to the variable

ncatted -a coordinates,ACHA_mode_8_Tc_Lower_Cloud,c,c,"pixel_longitude pixel_latitude" input.nc lower_cloud.nc

2. extract the variable ACHA_mode_8_Tc_Lower_Cloud from input.nc

cdo -selname,ACHA_mode_8_Tc_Lower_Cloud lower_cloud.nc lower_cloud_with_coordinates.nc

3. remap to global 0.05x0.05 degree lonlat grid

cdo -remapbil,global_0.05 lower_cloud_with_coordinates.nc lower_cloud_remap.nc

Karin,

I would like to remap another type of netCDF file. It appears that this file may need to be modified differently than the previous file that you we able to remap.

In this case, I would like to remap the variable "PRES" found in the attached file input2.nc. As before, mygrid describes the latitude-longitude grid file.

Thank you for your help!

-Bob Rabin

Sorry, but I wasn't able to get any correct result.

Thank you for checking.
I think the problem may be that the netCDF doesn't contain arrays of latitude and longitude locations. Instead they are arrays of x,y positions relative to a satellite grid.

I guess, you can remap the data to a lonlat grid with CDO but you need to know the correct projection. The following gridfile unfortunately is not fully correct but maybe you can start with it.

ncatted -O -a coordinates,PRES,o,c,"y x" input2.nc input2_coords.nc

gridfile.txt:

gridtype  = projection
gridsize  = 29419776
xsize     = 5424
ysize     = 5424
xname     = x
xlongname = "GOES fixed grid projection x-coordinate" 
xunits    = "radians" 
yname     = y
ylongname = "GOES fixed grid projection y-coordinate" 
yunits    = "radians" 
xfirst    = -0.151844
xinc      = 5.6e-05
yfirst    = 0.151844
yinc      = -5.6e-05
grid_mapping = goes_imager_projection
grid_mapping_name = geostationary
long_name = "GOES-R ABI fixed grid projection" 
perspective_point_height = 35786023.
semi_major_axis = 6378137.
semi_minor_axis = 6356752.31414
inverse_flattening = 298.2572221
latitude_of_projection_origin = 0.
longitude_of_projection_origin = -75.
sweep_angle_axis = "x" 
proj_params = "+proj=geos +h=35786023.0 +lon_0=-75.0 +sweep=x" 

targetgrid.txt:

gridtype  = lonlat
gridsize  = 240000
xsize     = 600
ysize     = 400
xname     = lon
xunits    = "degrees_east" 
yname     = lat
yunits    = "degrees_north" 
xfirst    = -120.
xinc      = 0.1
yfirst    = -55.0
yinc      = 0.1

cdo -b F32 -P 8 -remapnn,targetgrid.txt -setgrid,gridfile.txt -selname,PRES input2_coords.nc outfile.nc

NOAA provides a Python function that uses the GOES-R ABI fixed grid projection to calculate the latitude and longitude values.

https://www.star.nesdis.noaa.gov/atmospheric-composition-training/python_abi_lat_lon.php

So, maybe you better should go with Python.

And there is a NOAA web page that points to datasets containing the appropriate latitude longitude grid.

https://www.star.nesdis.noaa.gov/atmospheric-composition-training/satellite_data_goes_imager_projection.php

Thank you! This is very helpful. I will try these suggestions.
-Bob