New BLOM/iHAMOCC spin-up for NorESM2.5/NorESM3 test simulations?

[TomasTorsvik]

@matsbn , @JorgSchwinger , @jmaerz
For recent testing of NorESM2.5/NorESM3, we have been using restart files from a NOIIAJRAOC20TR run

/cluster/shared/noresm/inputdata/restart/NOIIAJRAOC20TR_TL319_tn14_ppm_20240816/rest/1775-01-01-00000/

With the v1.7.1 BLOM/iHAMOCC version we now have a fairly substantial shift in some of the
ocean BGC fields, see for instance:
https://ns2345k.web.sigma2.no/datalake/diagnostics/noresm/matsbn/n1850.ne30_tn14.hybrid_wavparam_vfcor_fatessp.20250205/HAMOCC_DIAG/yrs32to41-n1850.ne30_tn14.hybrid_wavparam_fatessp.20241226/set1/set1_ann_avg_2models.png

I'm not sure we will learn much from future coupled simulations if we only see the initial adjustment.

Would it make sense to create a new set of restart files for BLOM/iHAMOCC and CICE for use in NorESM2.5/NorESM3 coupled simulations based on the v1.7.1 model version? Would it be sufficient to just run the JRA forcing for a few cycles, or should we try also to spin up the sediments? It probably doesn't make sense to have a long spin-up run now as we have yet to decide on the basic configuration that we want to end up with, but maybe a short spin-up, potentially starting from existing simulations, could be useful.

My initial approach would be to just run NOIIAJRAOC20TR_TL319_tn14 for a few cycles, but I would like to hear from you if there may be more sensible options.

[jmaerz]

@JorgSchwinger and @TomasTorsvik , related, but not necessarily exactly corresponding: I wonder, if it would be worth it to have a shared folder for restart files - particularly for the sediment - which then could spin-up further with every run that we do? (Ideally having some climatological sedimentation fluxes to bottom attached as well - for #502).

Related to your question: maybe worth starting from a spun-up tnx2 run and trying to employ the remapping, followed by some further run time in tnx1? - it would give some more insights into a real simulation scenario when remapping is carried out, while not being critical, since it still not a production run?! - so just as a test with the aim to reduce drifts (and needed spin-up time) in tnx1.

[TomasTorsvik]

I think it makes sense to have specific sets of restart files under /cluster/shared/noresm/inputdata/restart/<casename>/rest/<restartdate>, but we should not use this space for storing full sets of output files. For spinup runs, I suggest we use the space on Nird under /nird/datalake/NS9560K/noresm/HAMOCCSPINUP.

[TomasTorsvik]

@jmaerz , @JorgSchwinger , @matsbn , @tjiputra
I have started to put OMIP output on Nird under /nird/datalake/NS9560K/noresm/omip/cases/. At the moment I am extending the NOIIAJRAOC20TR.TL319_tn14.hybrid_spinup.3.0a02.20250409 case, hopefully this can provide a useful starting point for future fully coupled runs as well.

[JorgSchwinger]

Hi @TomasTorsvik , the NOIIAJRAOC20TR is a 20th century transient forcing compset. Most notably it has transient CO2 (at least it used to have in NorESM2), which is not suited for spinup. Also (but less important) it would be nice to use a repeat-year forcing (early on in the JRA cycle to avoid a too strong warming signal) for spin-up purposes.

[JorgSchwinger]

(allthough the CO2 forcing will be tied to the actual model years, so as long as the model years stays before 1850, it is probably ok)

[TomasTorsvik]

Hi @JorgSchwinger - so NOIIAJRAOC is the recommended compset for spinup? I will use that next time.

I know you have a setup for repeated 1961 forcing for NorESM2, and I was looking into how this can be done with NUOPC/CDEPS. I see there are already some options available in components/cdeps/datm/cime_config/config_component.xml, with labels JRA-RYF8485, JRA-RYF9091, and JRA-RYF0304. I started a test with JRA-RYF8485 now, but maybe we should define a new one for early 1960s?

[TomasTorsvik]

@jmaerz , @JorgSchwinger , @matsbn , @tjiputra
Based on discussion in NorESMhub/CDEPS#14, it seems that we may want to use the new JRA-RYF6161 forcing for the next pre-spinup run. I'm hoping to get this into a new noresm3_0_alpha03a tag and get a NOIIAJRARYF6162OC run started soon. It would be nice if someone could have a look at PR #545 so I can merge and tag BLOM/iHAMOCC for the new NorESM tag (makes documenting the run a bit easier).