CLM5 compatibility with spared grid cells

[liliyao]

Hi,
I am wondering if the CLM5 from release-cesm2.2.0 and recent develop version (i.e., ctsm5.1.dev118) can run with spared grid cells. I learned that the CLM5 Perturbed Parameter Ensembles version can run with the spared cells using these two lines in user_nl_clm file:
hist_type1d_pertape(1) = 'GRID'
hist_dov2xy(1) = .false.

But when I added these two lines to the user_nl_clm in release-cesm2.2.0 or dev118, I got the following error:
ERROR: Command /my_cesm_sandbox_2_2_0/components/clm/bld/build-namelist failed rc=255
out=
err=ERROR(Build::Namelist::_parse_next): expect a equal '=' sign or '+=', instead got: = 'GRID'

So, I am uncertain whether release-cesm2.2.0 and ctsm5.1.dev118 have the capability to run with spared grid cells or different command should be added to the user name list.
Thank you very much!

 

[oleson]

Those changes in user_nl_clm will simply output the history variables in 1d format, i.e., (time,gridcell), not (time,lat,lon).
I don't see anything obviously wrong with your two lines you've added, but check to make sure you are using single quotes around GRID and not backquotes.

I assume by spared you mean sparse. The sparse grid for the PPE means that a limited number of gridcells are actually modeled in a global simulation setup, as specified by the domain file. For the CLM PPE this is 400 gridcells. Specifying those two lines is a way to output just those 400 grid cells in a history file in vector format.

 

[liliyao]

Thank you very much, Dr. Oleson, and sorry for the delayed follow-up. I was tied up with other problems. Yes, I mean the sparse grid. Can I also model a limited number of grid cells in a global simulation setup using the ctsm5.1.dev118, like what can be done in CLM PPE version? For example, in this 6x8 domain, only the 7 grey cells are active, and I would like to model and output only the 7 grey cells. If ctsm5.1.dev118 also has this capability, how do I set up the domain and mesh files? Thank you!

 

[slevis]

Hello @liliyao,

I will not go into the details, because you can find a very detailed explanation here:

Setting up (high-res sparse) regional-grid CTSM simulations · ESCOMP CTSM · Discussion #1919

Context This started as issue #1773 from which I linked a google doc (now out of date and obsolete) with step-by-step instructions intended initially for a case that @XiulinGao needed to run. To th...

github.com

Sam Levis

 

[liliyao]

Hello @liliyao,

I will not go into the details, because you can find a very detailed explanation here:

Setting up (high-res sparse) regional-grid CTSM simulations · ESCOMP CTSM · Discussion #1919

Context This started as issue #1773 from which I linked a google doc (now out of date and obsolete) with step-by-step instructions intended initially for a case that @XiulinGao needed to run. To th...

github.com

Sam Levis

Hi Sam, thank you for pointing me to the detailed steps. I followed these steps and tried to set up a very simple case with three sparse grids within a 3x3 regional grid. I generated a "landmask.nc" file and created "lnd_mesh.nc" from a 3x3 "fsurdat" file. However, when I attempted to generate the mask mesh by executing "./mesh_mask_modifier" from CTSM/tools/modify_input_files/, I encountered an error: "ERROR: lon_in needs to be in the range 0 to 360."

But the longitude in my "landmask.nc" file is already within the range of 0-360 degrees. Therefore, I have no idea how to resolve this issue. Any help or suggestions would be appreciated! I have also attached my "landmask","lnd_mesh", ".cfg", and "fsurdat" files. Additionally, I do my simulations on NERSC and don't have access to Cheyenne.

Thanks!
Lili

 

[slevis]

Hi @liliyao,

Just a few minutes of investigation tells me that the problem may be with your lnd_mesh_3x3.nc file. Use ncdump to look at the contents of the file. You will see longitude values of 607 instead of 247.

@liliyao I would like to take this opportunity to make a kind request of everyone on the CESM forum:
Please spend time troubleshooting things yourself before turning to the forum for help. I say this for two reasons:
1) All model users and model developers (including the most experienced) run into problems when they try new things. The way you become truly experienced is by troubleshooting things that don't work. This forum is good for posting questions if you have spent time troubleshooting and are still stuck.
2) The number of model users is constantly increasing (this is good) and @oleson and I already answer dozens of questions per week. Ideally we should only need to answer questions for users who cannot make progress in their work after all their troubleshooting.

@liliyao I apologize if my comment does not apply to you and you already did significant troubleshooting before posting the last question, and please understand that I am trying to make this point as tactfully as possible.

Sincerely,
Sam Levis

 

[liliyao]

Hi Sam, thank you for your prompt response and taking the time to help me troubleshoot the issue. I apologize if I did not spend enough time troubleshooting before posting my question on the forum. I will be sure to keep your advice in mind in the future.