Running CESM2 BSSP585 compset

[demetray]

I am using CESM2.1.3 to run a BSSP585 compset beginning in 2015. I have tried a variety of restart files (generally those available on NCAR Derecho here: /glade/campaign/cesm/cesmdata/inputdata/cesm2_init) along with BSSP585 and BSSP585cmip6, but I keep getting a variety of errors related to the land interpolation scheme.

I have gone through various attempts to debug this, involving setting CLM namelist variables use_init_interp = 'true' and init_interp_method = 'general', and yet I still get some kind of error.

I assume this could also be resolved by choosing certain restart files and compsets. How do I run a BSSP585 (or BSSP585cmip6) compset in CESM2.1.3? Exactly what restart files are compatible with either of these compsets?

Generally, I am concerned with the ocean and atmosphere more. Biogeochemistry and land are not important for my simulation. Please advise!

 

[strandwg]

The restarts for the SSP5.85 runs that were done for CMIP6 are on derecho at

/glade/campaign/collections/cmip/CMIP6/restarts/CESM2_f09_g17_ssp_runs/b.e21.BSSP585cmip6.f09_g17.CMIP6-SSP5-8.5.(101,102,103)

 

[demetray]

Okay, I didn't know this directory existed. Thank you! Although, I am still a bit confused: I need to start my simulation in the year 2015, which I think means I need restart files based on the end of a historical simulation (so I can pick up in the year 2015), whereas the directory you mentioned has restarts several years into the SSP585 simulation (2025 onwards).

Do you know which restarts can be used to initialize an SSP585 simulation in the year 2015? Which restart files can I use to start a simulation in 2015 (which I will then subject to forcing according to either the BSSP585 or BSSP585cmip6 compset)?

 

[strandwg]

The restarts (including 2015) for predecessor historical runs for the SSP5.85 simulations are located at

/glade/campaign/collections/cmip/CMIP6/restarts/CESM2_FV1_historical_runs

The SSP5.85 runs branched from the following historical runs:
run 101 branched from b.e21.BHIST.f09_g17.CMIP6-historical.010
run 102 branched from b.e21.BHIST.f09_g17.CMIP6-historical.011
run 103 branched from b.e21.BHIST.f09_g17.CMIP6-historical.004

 

[demetray]

Thank you so much! I was able to run a CMIP6 SSP585 simulation beginning in 2015 with the following specifications:

Restart files (on Derecho):
/glade/campaign/cesm/cesmdata/inputdata/cesm2_init/b.e21.BHIST.f09_g17.CMIP6-historical.001/2015-01-01
--> To clarify: is this just a different ensemble member for a run that is comparable to the ones you mentioned available at /glade/campaign/collections/cmip/CMIP6/restarts/CESM2_FV1_historical_runs ?

Component set:
SSP585_CAM60_CLM50%BGC-CROP-CMIP6DECK_CICE%CMIP6_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD

CLM namelist variables:
use_init_interp = .true.
init_interp_method = 'general'

 

[strandwg]

The two sets are identical. However, b.e21.BHIST.f09_g17.CMIP6-historical.001 was not used as the initial state of any of the CMIP6 SSP5.85 runs; 004, 010 and 011 were used.