search

NorESMhub / BLOM

Bergen Layered Ocean Model
GNU Lesser General Public License v3.0
17 stars 25 forks source link

Initialization of preformed tracers #356

Open jmaerz opened 5 months ago

jmaerz commented 5 months ago

Hi @JorgSchwinger and @TomasTorsvik , I was just checking, which data products would be available for initializing the preformed tracers. One reference that I found is: Carter et al. 2020: Preformed Properties for Marine Organic Matter and Carbonate Mineral Cycling Quantification. https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GB006623 - however, it is lacking preformed DIC. It seems, however, that one can back-calculate preformed DIC from the other available tracers via Eq. (5a) in Li et al. 2000: Remineralization Ratios in the Subtropical North Pacific Gyre. One thing in general is the potentially deviating stoichiometry used for deriving the preformed tracer fields. In case you agree, I would fill the mixed layer data with GLODAPv2 data (on which the product is based) which then could serve as initialization (maybe we also should consider moving to WOA18 or GLODAPv2 where possible for the CMIP7 cycle). Thoughts, opinions?

Once done, the restart part for preformed silicate can be cleaned up.

TomasTorsvik commented 5 months ago

Hi @jmaerz , I don't have any insights to offer, but it seems reasonable to use GLODAPv2 or WOA18. I see there is also WOA23 available now https://www.ncei.noaa.gov/news/World-Ocean-Atlas-2023-v2 , mainly extending the datasets to 2022. GLODAPv3 is expected end of 2025, so that is probably too late for CMIP7 fast track.

JorgSchwinger commented 5 months ago

Regarding WOA version, I would wait until input4MIPs is being updated. Currently we use the OMIP-recommended input-files based on WOA13. I suggest we update this when new input4MIPs data is published.

Regarding the preformed, I didn't have the chance to read through the articles (on a conference right now), but I share the concern about differing stociometric ratios.

jmaerz commented 5 months ago

I am not sure, if one can correct in some way for the stoichiometry (ongoing discussion). Another option would then be spinning up the preformed tracers in i.e. a 2 degree ocean and interpolating to a 1 degree ocean...