aradhakrishnanGFDL
commented
3 weeks ago @jtmims @wrongkindofdoctor Passing along these updated catalogs (this catalogbuilder PR has additional info on how to reproduce these runs. One can use fre-cli as well - open GitHub issues in the respective repo to take things for a spin or report issues. thanks!)
/home/a1r/github/noaa-gfdl/catalogs/CM4.5v01_om5b06_piC_noBLING.json
/home/a1r/github/noaa-gfdl/catalogs/ESM4.5v01_om5b04_piC.json
/home/a1r/github/noaa-gfdl/catalogs/CM4_piControl_C.json
What problem will this feature solve? Catalogs for new simulations
Here are pointers to data catalogs @Ciheim generated using the light-weight GFDL catalog builder, for two out of the three simulations @jkrasting wants MDTF to be run on.
(thanks @Ciheim!) `Alrighty, I've generated the first two catalogs. You can find them under ~/ESM-OM5-CM5_cat1.csv and ~/ESM-OM5-CM5_cat2.csv. The third gives an error at this time.
Known issues:
CF name is not populated for the variables, and it looks like the internal variable names in the look-up tables we use do not have those variables. If we use the cmip6 tables, we may be able to get many others populated. So, we will attempt to do that first and report back. But, please try out these catalogs and let us know if there is anything else so we can start using these earlier on in the process this time.
I see a sea ice POD that uses siconc. Perhaps we can also prioritize the CF names for the POD requested variables at this time.
Info: Except for the standard_name, everything else does not require opening the file and we are trying to bypass that by using the look up table, until there is a better way for workflow to inform these utilities. catalog builder has an option to open the dataset, but it is not "light".
thanks!