search

NCAR / LMWG_dev

Repository to track LMWG development simulations
3 stars 0 forks source link

CRUJRA-SP (ctsm51d142_f19_CRUjra_[1850, hist]_flds) #39

Open wwieder opened 7 months ago

wwieder commented 7 months ago

Description:

Given results in #20 and #35 I wonder if we should try using the CRUJRA longwave?

Initially, a clone of #20 seems logical to see if we see a downward trend (or any other issues) with using the FLDS provided in CRUJRA?

Case directory: Spin-up: /glade/u/home/slevis/cases_LMWG_dev/ctsm51d142_f19_CRUjra_1850_flds Hist: /glade/u/home/slevis/cases_LMWG_dev/ctsm51d142_f19_CRUjra_hist_flds

Sandbox: /glade/work/slevis/git_ctsm_tags/ctsm5.1.dev142

usernl changes: As in #20 except: diff between user_nl_datm_streams

49c49
< CLMCRUNCEPv7.TPQW:datavars = QBOT Sa_shum,PSRF Sa_pbot,TBOT Sa_tbot,WIND Sa_wind,FLDS Faxa_lwdn
---
> CLMCRUNCEPv7.TPQW:datavars = QBOT Sa_shum,PSRF Sa_pbot,TBOT Sa_tbot,WIND Sa_wind

Also user_nl_clm in the hist case here will point to this case's corresponding finidat.

SourceMods: NONE

Diagnostics: With CRUjra FLDS vs. with clm-calculated FLDS: https://webext.cgd.ucar.edu/I20TR/ctsm51d142_f19_CRUjra_hist_flds/lnd/ctsm51d142_f19_CRUjra_hist_flds.1995_2014-ctsm51d142_f19_CRUjra_hist.1995_2014/setsIndex.html

With CRUjra FLDS vs. GSWP3, showing trends from 1901 to the present: https://webext.cgd.ucar.edu/I20TR/ctsm51d142_f19_CRUjra_hist_flds/lnd/ctsm51d142_f19_CRUjra_hist_flds.1995_2014-ctsm51d142_f19_GSWP3V1_hist.1995_2014/setsIndex.html

ILAMB: https://webext.cgd.ucar.edu/I20TR/ctsm51d142_f19_CRUjra_hist_flds/lnd/_build_ILAMB2_7_ForcingDatasets_flds/

Output: /glade/campaign/cgd/tss/people/slevis/i_cases/cases_LMWG_dev/ctsm51d142_f19_CRUjra_1850_flds/ /glade/campaign/cgd/tss/people/slevis/i_cases/cases_LMWG_dev/ctsm51d142_f19_CRUjra_hist_flds/

Contacts: @slevis-lmwg

Extra details: Same as #20

slevis-lmwg commented 7 months ago

@olyson @wwieder I decided to try premium and the run has started! (So I moved #38 to premium, as well.)

wwieder commented 7 months ago

awesome! any preliminary results (even as an ncview of ncdiff) of monthly snowdepth or TBOT that we can share at tomorrow's meeting could be helpful.

I'm working on slides here https://docs.google.com/presentation/d/1Y2uYNcWb1RcSY30I6E8J0AboTC0nG6HD9Jtlssu0_Bs/edit?usp=sharing .

slevis-lmwg commented 7 months ago

@wwieder I just posted diags above. I have not looked at them, yet. (The simulation finished last night.)

slevis-lmwg commented 7 months ago

I added a plot to the slides

slevis-lmwg commented 7 months ago

I have started new diagnostics to plot the full trends: 1901-2022

slevis-lmwg commented 7 months ago

Full trends should be up. See link above.

wwieder commented 7 months ago

Thanks. Global FLDS trends seem ok here. I don't see any glaring issues in the arctic regional plots either.

There seems to be a jump in global FSDS that I don't really understand, but suspect they're related to the transition from CRU to JRA forcing? I think JRA55 starts in 1958.

wwieder commented 7 months ago

Also less summer snow at the end of the run with CRU-JRA, but air temperature are colder, here for the whole polar region.

@olyson when you're back next week can you add this case to the ILAMB-SP diagnostic plots. Then, assuming the seem OK, it's maybe worth repeating #35 with the CRU-JRA_flds inputs included?

olyson commented 7 months ago

Sure!

olyson commented 7 months ago

Added this case to the ILAMB comparison here: https://webext.cgd.ucar.edu/I20TR/ctsm51d142_f19_CRUjra_hist_flds/lnd/_build_ILAMB2_7_ForcingDatasets_flds/ Snow water equivalent certainly looks better with flds.

wwieder commented 7 months ago

Thanks for posting, Keith.

So does the seasonal cycle of incoming longwave! image.