generating initial files in 2 separate stages

[jpen]

How to use ERA5 Data From Copernicus Database
Following from the above the procedure for separately handing surface and pressure level ERA5 data, may I know whether is it possible to include both pressure level intermediate file FILE:YYYY-MM-DD and and surface level intermediate file SFILE:YYYY-MM-DD for config_met_prefix in the namelist.init_atmosphere? If not, how can I process separately FILE:YYYY-MM-DD, followed with SFILE:YYYY-MM_DD? and what changes are required in the namelist file? Can the resulting outputs be combined later with cat command?
I tried with the FILE:YYYY-MM-DD and got the following error:

$ tail -f log.init_atmosphere.0000.err
----------------------------------------------------------------------
Beginning MPAS-init_atmosphere Error Log File for task 0 of 48
Opened at 2025/01/07 13:12:22
----------------------------------------------------------------------

ERROR: ********************************************************************************
ERROR: LANDSEA field not found in meteorological data file FILE:2017-06-01_12
CRITICAL ERROR: ********************************************************************************
Logging complete. Closing file at 2025/01/07 13:12:22

 

[gdicker]

I think this other thread may help with initializing MPAS from ERA5 data: Initialize MPAS with ERA5 data

Though I haven't tried creating initial conditions with ERA5 data, you seem to be on the correct path to combine ERA5 data with a static file to create initial conditions for MPAS-A.

From that other thread, I think you actually want to use the cat command to combine your FILE with the SFILE. The resulting file should have all the expected meteorological fields that MPAS needs.

 

[jpen]

Thanks for the response. It is not possible to generate init.nc file separately for the pressure level and surface level data. There seems to be some level of interdependence, for instance, I found that landsea mask is a necessary parameter for processing pressure level data. I am not aware if any changes in the namelist file (e.g. &preproc_stages) are necessary for processing them separately?

 

[gdicker]

I also am not aware of any way to process them separately. You need to combine the ungribbed ERA5 files to generate init.nc, as I understand from the linked thread.

 

[jpen]

I also am not aware of any way to process them separately. You need to combine the ungribbed ERA5 files to generate init.nc, as I understand from the linked thread.

Understood. Thanks again.