f.c6_3_41.FWscHIST.ne30_L58_exner.002

[cecilehannay]

Description: Same as https://github.com/NCAR/amwg_dev/issues/53 + second set of modifications to SE to alleviate numerical noise near topography

• Map PHIS from GLL grid to physics grid to alleviate noise in derived fields such as PSL and eddy Z500)

• Exner function formulation of pressure gradient force (PGF) + subtract reference profiles from T and dp before applying hyperviscosity (see complete description)

Case directory: Locally (if still available): /glade/p/cesmdata/cseg/runs/cesm2_0/f.c6_3_41.FWscHIST.ne30_L58_exner.002

On github: https://github.com/NCAR/amwg_dev/tree/f.c6_3_41.FWscHIST.ne30_L58_exner.002

Sandbox: Locally (if still available): /glade/work/hannay/cesm_tags/cam6_3_041

On github: https://github.com/ESCOMP/CAM/tree/cam6_3_041 hash: 963ebdc

Diagnostics: AMWG diags (if available) https://webext.cgd.ucar.edu/FWscHIST/f.c6_3_41.FWscHIST.ne30_L58_exner.002/atm/

Contacts: @PeterHjortLauritzen, @adamrher, @JulioTBacmeister, @cecilehannay

[cecilehannay]

@PeterHjortLauritzen and @adam: this run completed 10 years. I added the diags: https://webext.cgd.ucar.edu/FWscHIST/f.c6_3_41.FWscHIST.ne30_L58_exner.002/atm/

[adamrher]

See here 10 year means and variance of FT (lev=50) and OMEGA500:

A couple observations

• There's a back-buldge feature (geology term) on the back side of the Andes in the exner only run. It can be seen in means/variance of FT and the mean of OMEGA500. Not sure if this is a gibbs ringing type phenomena, but it is eliminated in exner+refp simulation, for the most part.
• There's a similar pos-neg-pos wave over the Himalayas in the exner only run, as seen in the mean OMEGA500 plot. The exner+refp eliminates this and creates a single local maxima aligned with the front range. This seems like a more physical solution to me.

I see that we PRECC/PRECL on the se-grid as well. I will add those plots to this thread later.

[adamrher]

10 year mean/variance of PRECT:

Note the reduction in spurious precip around the Peruvian bend in both the means and variance. I think it's clear the exner+refp is the better model. As peter says, refp is more effective at mitigating the spurious upslope transport that is likely related to the Peruvian precip problem.

[PeterHjortLauritzen]

I agree with Adam! PR is being prepared with the modifications for "exner+refp". Please use these source mods in any future simulations until the PR has been approved and merged with cam_development. Thank you!

[justin-richling]

@PeterHjortLauritzen, @adamrher, @JulioTBacmeister, @cecilehannay

ADF diags: https://webext.cgd.ucar.edu/FWscHIST/f.c6_3_41.FWscHIST.ne30_L58_exner.002/atm/ADF/