For GRIB:

On Linux you can use the grib_copy tool. This tools comes with ECMWF's ecCodes:

grib_copy grib_file_1 grib_file_2 ... output_grib_file


On Linux you can simply concatenate the GRIB files:

cat grib_file_1 grib_file_2 ... grib_file_N > output_grib_file

For NetCDF (unsupported by ECMWF)

A way to merge netcdf files is using CDO merge:

cdo merge infile1 infile2 infile3 outfile

or

cdo mergetime infile1 infile2 infile3 infile4 outfile


Another option could be NCO (NetCDF Operators) ncrcat. Be sure that the files are unpacked before merging them. Here an example of merging for ERA5 nc files:

Download the data:

import cdsapi
c = cdsapi.Client()
c.retrieve(    
	'reanalysis-era5-single-levels',    
	{
		'product_type': 'reanalysis',        
		'variable': ['surface_latent_heat_flux', 'surface_sensible_heat_flux',],        
		'year': ['2018'],        
		'month': ['01'],        
		'day': ['31'],        
		'time': ['19:00','20:00','21:00','22:00','23:00'],        
		'format': 'netcdf',    
},'ERA5_slhf_sshf_01.nc')
    
c.retrieve(   
	'reanalysis-era5-single-levels',    
	{
		'product_type': 'reanalysis',        
		'variable': ['surface_latent_heat_flux', 'surface_sensible_heat_flux',],        
		'year': ['2018'],        
		'month': ['02'],        
		'day': ['01'],        
		'time': ['00:00', '01:00', '02:00', '03:00', '04:00', '05:00',],        
		'format': 'netcdf',    
	},'ERA5_slhf_sshf_02.nc')

Set record dimension in each file (this is the dimension to merge along):


module load nco ncks --mk_rec_dmn time ERA5_slhf_sshf_01.nc ERA5_slhf_sshf_01_time_rec.nc ncks --mk_rec_dmn time ERA5_slhf_sshf_02.nc ERA5_slhf_sshf_02_time_rec.nc

Data are still packed - need to be unpacked before merge:


ncpdq -U ERA5_slhf_sshf_01_time_rec.nc ERA5_slhf_sshf_01_time_rec_unpack.nc ncpdq -U ERA5_slhf_sshf_02_time_rec.nc ERA5_slhf_sshf_02_time_rec_unpack.nc

Then they can be merged:


ncrcat ERA5_slhf_sshf_01_time_rec_unpack.nc ERA5_slhf_sshf_02_time_rec_unpack.nc ERA5_slhf_sshf_0102_time_rec_unpack.nc


Related articles