CTSM resolution problem

[maruf2812]

Thank you @oleson for the reply. Does this script work (CTSM/getco2_historical.ncl at master · ESCOMP/CTSM)? And do I need to edit the script for 1 deg resolution?

 

[oleson]

I think that script just creates global or latitude-band CO2 files for use by datm using as input CAM ghg files.
To my knowledge all of the files have been created and are at:

/glade/p/cesmdata/cseg/inputdata/atm/datm7/CO2/

 

[maruf2812]

Hi @oleson, I have used this line to create surface data for 1 deg run (./getregional_datasets.pl -sw 25,235 -ne 52.9375,292). In this case what will be my input file?

 

[oleson]

I haven't done this before but I imagine your input files would need to be global 1deg domain and surface datasets. You could get these filenames by creating a 1deg global case, running preview_namelists, and then looking at lnd_in. @erik may be able to provide further guidance here. If you are using CTSM master/main then the subset_data.py might be an alternative script to use, although I think there is further development of this script in an upcoming tag.

 

[maruf2812]

Hi @oleson, Thank you for your kind help. I have followed your instruction: create a case for 1 deg, get the surface and atm dataset from ./getregional_datasets.pl. I put the atm and surface dataset in my user_nl_clm but in case.build step is failing and showed the following error. I have attached the screenshots.

 

[oleson]

I think you need to remove fatmlndfrac from your user_nl_clm and instead specify the path and filename in env_run.xml using LND_DOMAIN_PATH and LND_DOMAIN_FILE.

 

[maruf2812]

Hi @oleson thank you! The case.build is successful but the case.submit shows the following error. I have attached the screenshot.

 

[oleson]

You are requesting more processors than the number of land grid cells. Your env_mach_pes.xml is probably still setup for a global case. Try reducing the number of tasks (NTASKS in env_mach_pes.xml).

 

[maruf2812]

Hi @oleson, I reduced my NTASK to 20, it was 50 before. Now I encountered the following error:

 

[oleson]

I think your ATM_DOMAIN_FILE needs to be set to the NLDAS2 resolution if that is what your forcing data is.

 

[maruf2812]

Hi @oleson thank you again. Can I use my 0.125 deg resolution run ATM_DOMAIN_FILE (domain.lnd.0.125nldas2_0.125nldas2.190410.nc) for this 1 deg run? In that case do I need to only change the surface dataset that I got from ./getregional?

 

[oleson]

I believe so.

 

[maruf2812]

Hi @oleson, tried with this but got the following error. I think my surface data and atm file are different and it causes the error.

 

[maruf2812]

I was wondering if I can make NLDAS2 1deg resolution atm file so that both the file will be same size.

 

[oleson]

You could create an NLDAS2 1deg forcing file. But you should be able to force your 1deg surface with the NLDAS2 atm data, I must be missing something. I'll take a closer look at this in a day or two.

 

[maruf2812]

Hi @oleson, thank you! I have attached the log file here.

 

[oleson]

Ok, I think I have something that is working now. A 1deg surface forced by NLDAS atm.
Rather than go into all of the details, here is a case you can look at:

/glade/work/oleson/ctsm_runs/CTSM_CLM5.0_1deg_V1

I started by cloning this case:

/glade/u/home/maruf/CTSM_CLM5.0_SP_V1

So just look at the differences in files between yours and mine (e.g., user_nl_clm, env_run.xml, user_nl_datm, env_mach_pes.xml, etc.).

Your 1deg surface data domain is larger than the NLDAS forcing domain so you'll get some strange results at the boundaries of your domain.

 

[maruf2812]

It is working now! Thank you @oleson for your relentless support and help. Without your support, our research would be difficult.