Running F2010 with a SOM / Error with input files

[pgoddard08]

Hello,

I was able to successfully build a compset I defined as F2010_SOM (long name 2010_EAM%CMIP6_ELM%SPBC_MPASSI%PRES_DOCN%SOM_MOSART_SGLC_SWAV)

Though before the script can run, I believe I need input data for docn, namely the domain1 and file1. The error message is as follows:

hecking that inputdata is available as part of case submission
Setting resource.RLIMIT_STACK to -1 from (-1, -1)
Loading input file list: 'Buildconf/elm.input_data_list'
Model elm no file specified for finidat
Loading input file list: 'Buildconf/eam.input_data_list'
Loading input file list: 'Buildconf/mpassi.input_data_list'
Loading input file list: 'Buildconf/docn.input_data_list'
Model docn missing file domain1 = '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET'
Model docn missing file file1 = '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET'
Loading input file list: 'Buildconf/mosart.input_data_list'
Loading input file list: 'Buildconf/cpl.input_data_list'
Using protocol wget with user  and passwd
Client protocol svn not enabled
Client protocol None not enabled
Checking server https://web.lcrc.anl.gov/public/e3sm/inputdata/ with protocol wget
Setting resource.RLIMIT_STACK to -1 from (-1, -1)
Using protocol wget with user  and passwd
Loading input file list: 'Buildconf/elm.input_data_list'
Model elm no file specified for finidat
Loading input file list: 'Buildconf/eam.input_data_list'
Loading input file list: 'Buildconf/mpassi.input_data_list'
Loading input file list: 'Buildconf/docn.input_data_list'
Model docn missing file domain1 = '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET'
Trying to download file: 'ocn/docn7/SOM/UNSET' to path '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET' using WGET protocol.
wget failed with output:  and errput --2026-02-23 11:18:28--  https://web.lcrc.anl.gov/public/e3sm/inputdata/ocn/docn7/SOM/UNSET
Resolving web.lcrc.anl.gov (web.lcrc.anl.gov)... 140.221.25.4
Connecting to web.lcrc.anl.gov (web.lcrc.anl.gov)|140.221.25.4|:443... connected.
HTTP request sent, awaiting response... 404 Not Found
2026-02-23 11:18:28 ERROR 404: Not Found.

Model docn missing file file1 = '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET'
Trying to download file: 'ocn/docn7/SOM/UNSET' to path '/N/project/obrienta_startup/datasets/CESMinputfiles/ocn/docn7/SOM/UNSET' using WGET protocol.
wget failed with output:  and errput --2026-02-23 11:18:28--  https://web.lcrc.anl.gov/public/e3sm/inputdata/ocn/docn7/SOM/UNSET
Resolving web.lcrc.anl.gov (web.lcrc.anl.gov)... 140.221.25.4
Connecting to web.lcrc.anl.gov (web.lcrc.anl.gov)|140.221.25.4|:443... connected.
HTTP request sent, awaiting response... 404 Not Found
2026-02-23 11:18:28 ERROR 404: Not Found.

Loading input file list: 'Buildconf/mosart.input_data_list'
Loading input file list: 'Buildconf/cpl.input_data_list'
Checking server https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata with protocol svn
Setting resource.RLIMIT_STACK to -1 from (-1, -1)
Client protocol svn not enabled
Checking server None with protocol None
Setting resource.RLIMIT_STACK to -1 from (-1, -1)
Client protocol None not enabled
ERROR: Could not find all inputdata on any server

Are there available data (some SOM forcing dataset for file1 and a compatible domain file for domain1) that I can use? I'm not sure whether I'm close to getting this to run and just need input files, or if I'm missing something bigger here.

Thank you for your help.
Paul

[jonbob]

Hi @pgoddard08 -- what resolution are you using?

[pgoddard08]

I'm using ne30pg2_EC30to60E2r2 in E3SMv2.0

[whannah1]

@pgoddard08 @jonbob I'm pretty sure no one has ever created new SOM data for E3SM. I think Oluwayemi Garuba at PNNL was working on this years ago, but I don't think a final dataset was ever created.

Alternatively - a year or so ago I implemented a "relaxed slab ocean" (RSO) capability. You can find some information about it here in the new documentation:
https://docs.e3sm.org/E3SM/Data-Models/user-guide/data-ocean/

Here is the PR where it was implemented => #6798

This mode will relax the SST back to a prescribed SST dataset using either a uniform or varyiable mixed layer depth. I don't think we ever saved an SST dataset with the spatially or temporally varying mixed layer depth - so it'll take some work to use that. The uniform mixed layer depth is easy to use with existing SST datasets.

The caveat is that the relaxation is not perfect - and it will introduce long-term regional biases. The reasons for this were never clear to me, but it seems that the relaxation is never quite adequate for regions with large influence from ocean dynamics. One spot I remember in particular is that a temperature bias develops along the path of the gulf stream. In short-term simulation the RSO mode will produce cold wakes behind tropical cyclones and things like that - so for studying individual storms or sub-annual variability I think it's still appropriate to use. I think there's an easy paper to write that explains the trade-offs of this capability - but I never got around to writing it.

Let me know if you have any questions about that.

[whannah1]

also - this might not be obvious - if you want to use RSO you will likely need to switch to v3.

[jonbob]

Thanks @whannah1 -- I was just going to say pretty much the same thing. SOM forcing is sort of left to the user at this point

[pgoddard08]

Great, thank you all for your responses. I have some things to consider.

One item that would be helpful - where (what path) is the SOM code being compiled from? I'm curious to compare E3SM SOM to the one used in CESM. Many thanks.

[whannah1]

I think this is what you're looking for:

E3SM/components/data_comps/docn/src/docn_comp_mod.F90

Line 785 in c4b884d

case('SOM')

Something that I don't understand about that part of the code is where the "Q-flux" tendencies come into play... I would need to study it more to understand it.

[pgoddard08]

Great - thanks!

[rljacob]

FYI: the only SOM cases we test are with standalone mpas-seaice tests:
2000_DATM%JRA-1p5_SLND_MPASSI_DOCN%SOM_DROF%JRA-1p5_SGLC_SWAV_TEST (TL319_IcoswISC30E3r5)
2000_DATM%NYF_SLND_MPASSI_DOCN%SOM_DROF%NYF_SGLC_SWAV_TEST (T62_oQU240, T62_oQU240wLI)
There's an RSO case: 2010_EAM%CMIP6_ELM%CNPRDCTCBCTOP_MPASSI%PRES_DOCN%RSO_MOSART_SGLC_SWAV but we don't test it.
@whannah1 do you remember what resolution you ran that RSO case at?

[ogaruba]

I just saw this conversation. I have tested the SOM in E3SMv2. The code change is on my fork (https://github.com/ogaruba/E3SM/tree/ogaruba/E3SMv2_slab), and the paper is here: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2023MS003910. @pgoddard08 I also made the forcing files. I can share them if you're still interested. The SOM is not stable in other versions of E3SM I have tested, including v3, unless the qflux derived from another coupled simulation with robust AMOC is used. I have recently fixed this problem, and I can merge the code changes when the work is done, if there is interest.

[pgoddard08]

Hi @ogaruba
That's great! I will be using E3SMv2, so hopefully I can follow what you have done to success. I would be very grateful for the forcing files.

[ogaruba]

@pgoddard08 Sounds good. Are you working on Perlmutter? I can send the forcing file to you on there.

[pgoddard08]

@ogaruba Yes, I have access to Perlmutter -- are you able to put the file somewhere with permissions that I can copy it to one of my directories? If it helps, my username on PM is pgoddard

[ogaruba]

@pgoddard08 I shared the file. Let me know if you received it or if you have any questions while setting up the simulation