Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Questions around upscaling cam6 restart files

sebastien

Sébastien Barthélémy
New Member
Hi,

For some application with an ESM, I want to upscale restart files from a 2 degrees version of cam-6 to a 0.9 degree version of cam-6. I am using cdo to perform the upscaling of the fields that are on the usual lat,lon grid such as for example U or CO2. But, I do not know what to do with the other numerical variables, such as lcl_tod_start, or hyai. Should I copy them from the source file to the destination file, or should I leave them unchanged?

Another question is related to the files cam.rs. In the cam.rs file I have 2 dimensions x2a_nx and a2x_nx that are the size of the product lat*lon. How can I perform interpolation between the higher and the lower resolution versions of cam.rs with these dimensions?

I want to emphasize that I want to use the upscaled restart file as initial condition for a run with the lower resolution model.
 

aherring

Adam
Member
It would be a lot easier to just upscale an initial condition file (the 'ncdata' namelist) than to upscale restarts. Restarts are intended for bfb runs, which you're not doing because your regridding and using it for a different resolution. Do you happen to have a *.cam.i.* file from the 2deg run that you could upscale instead?
 

sebastien

Sébastien Barthélémy
New Member
Hi Adam,

Thanks for your answer. It brings a few other questions:
  1. What is a bfb run?
  2. Do I need to upscale only a *.cam.i.* file from the 2deg run for what I want to do? I do not need to upscale the *.cam.r.* *.cam.rs.* *.cam.h0.* files?
 

aherring

Adam
Member
bfb refers to 'bit-fot-bit' ... the cam restart files ensure you can exactly reproduce the solution upon restarting, and therefore a lot of data such as all the fluxes to the coupler need to be read in from the cam restart files. This ensures for example that the model solution doesn't depend on the numbers of times you restart the model.

For your purposes you are not looking to restart the model exactly; you are going to change the resolution and therefore by design cannot reproduce the run you are restarting from bfb. Therefore you probably want to do a RUN_TYPE=hybrid, in which you pick and choose which components are initialized by the restart files and which are initialized using a initial condition file. In your example you want to omit the rpointer.atm file (pointing to these cam.r and cam.rs files) and instead set ncdata in user_nl_cam to the remapped initial condition file. This refers to the *.cam.i.* file.

Note that the clm grid is (usually) on the atm grid, and so you will need to omit the rpointer.lnd file and instead set finidat in user_nl_clm to the *.clm2.r.* restart file (in contrast to CAM, you can use the restart file as the inic in CLM). Even though it's a different resolutoin it will actually map this inic to the target grid at run time using its own internal algorithms, and so you don't have to worry about upscaling it yourself.

Note that depending on how cdo handles the lat/lon/area variables, you may need to copy these in a from an existing 0.9deg inic file after you've remapped the rest of the file.
 

sebastien

Sébastien Barthélémy
New Member
Thanks Adam for that thorough answer. It brings a much needed clarification on how I am supposed to proceed, it will definitely help and make things much easier.
 

sebastien

Sébastien Barthélémy
New Member
Hi, I have a basic question: how to set the parameter files so that the model writes out the cam.i files together with the restart files cam.r., cam.rs. etc. Is it in env_run.xml or env_archive.xml?
 
Top