search

NCAR / CARMA_dev

Discussions on CARMA development and descriptions of model simulations
3 stars 0 forks source link

CAMchem CARMA v14 nudged 2deg 22 bins for ATom Period 2016-2018 #36

Open tilmes opened 1 year ago

tilmes commented 1 year ago

Case name and location: /glade/p/cesm/chwg_dev/tilmes/cases/carma/FCARMAnudged_f19.carma_trop_strat13.fixall.2016_2018.optics.fixmix3_22bins

Code directory: /glade/work/fvitt/camdev/carma_trop_strat13/ for new cases, we want to use /glade/work/fvitt/camdev/carma_trop_strat14 and the SourceMods other than carma_model_mod.F90 are not needed.

Run specifics: adding new bins requires changing the namelist (see user_nl_cam), carma_model_mod.F90 to change the number of bins, and env_build.xml. For 22 bins we have 415 advected tracers. For more bins, this number has to be increased. The model fails if the number is wrong, but states what number is needed:

Comparisons to ATom using this ncl code: /glade/u/home/tilmes/ncl/CARMA/NumberSizeDist-StratClim85hPa_ATom_flight_track_log_mam4_22bins.ncl for new runs, you have to change fname, first entry for CARMA, second is MAM, and Nbins, and the name of the plot (plotname). You will also have to change paths specific to my runs... Results show that CAMchem with 22bins shows a reduction in PRSULF and an increase in number for smallest mixed bins. This degrades the performance. 22bins 1deg vs 2deg show the same results, so it is fine to use the cheaper 2deg model. (Costs: ~7.7K Corehours per year. Results: Number for 20bins 1deg CAMchem: image Number for 22bins 1deg CAMchem: image Number for 22bins 2deg CAMchem image