CESM LGM runs: inconsistency between land fraction and sea ice fraction

[s_l_bradley-1@tudelft_nl]

Hello, I hope you can help. I am running cesm2.1 on cheyenne with the following compset: HIST_DATM%GSWP3v1_CLM50%BGC_CICE_POP2_RTM_CISM2%NOEVOLVE_SWAV
It is for an LGM case, where we have  produce new mapping and domain files, using the gen_mapping/gen_domain within the cesm2.1 package The error is 255: (seq_domain_check)  inconsistency between land fraction and sea ice fraction 
The create case is: /glade/u/home/mpetrini/LGM_simulations/script_create_cases/create_LGM_J-PDRTM
I understand this is caused by the routine: seq_domain_mct.F90  in CIME src/drivers/mct/main and it is due to a difference between fracl (land fraction) and fraci (ice fraction)
However, the sea ice fraction variable fraci[/b] seems to have CAM gridsize rather than POP gridsize; so I wonder how this variable is calculated; my guess was that it was taken from domain files and the n mapped from POP grid to CAM grid.If I change domain files I get the same error.
3) new mapping files /glade/work/mpetrini/LGM_initial/CIME/gen_mapping_files/ 4) new domain files /glade/work/mpetrini/LGM_initial/CIME/gen_domain/domain.lnd.fv0.9x1.25_gx1v7.190315.nc/glade/work/mpetrini/LGM_initial/CIME/gen_domain/domain.ocn.gx1v7.190315.nc
I have attached all the logs:  note there are some additional outputs flag, so I have attached the minor revision made to the seq_domain_mct.F90
Thanks
Sarah 

 

[duvivier]

This seems like an issue with how the new domain files were produced and there is at least one point that the land thinks is land but the ice/ocean thinks is ice/ocean. Do you have a grid point(s) it's pinpointing in the error files? It may be worth contacting the paleo climate working group (http://www.cesm.ucar.edu/working_groups/Paleo/) to ask how they generally verify the grids are generated properly in these types of runs.

 

[duvivier]

This seems like an issue with how the new domain files were produced and there is at least one point that the land thinks is land but the ice/ocean thinks is ice/ocean. Do you have a grid point(s) it's pinpointing in the error files? It may be worth contacting the paleo climate working group (http://www.cesm.ucar.edu/working_groups/Paleo/) to ask how they generally verify the grids are generated properly in these types of runs.

 

[s_l_bradley-1@tudelft_nl]

Hello Thank you for the reply. Unfortunately the paleo group are busy at the moment.  We followed the default gen_domain code as within the cesm2.1 default code and the suggestions within the paleo toolkit.Sarah 

 

[s_l_bradley-1@tudelft_nl]

Hello Thank you for the reply. Unfortunately the paleo group are busy at the moment.  We followed the default gen_domain code as within the cesm2.1 default code and the suggestions within the paleo toolkit.Sarah 

 

[dbailey]

I have moved this to the Paleoclimate forum. Hopefully mvr@ucar.edu can chime in here.Dave

 

[dbailey]

I have moved this to the Paleoclimate forum. Hopefully mvr@ucar.edu can chime in here.Dave