AMIP roadmap overview

[charleskawczynski]

This issue outlines the AMIP roadmap overview for the atmosphere. This will be updated periodically.

Next to each work delegation item is listed a Nearest (unfinished) Key Result (NKR), pointing to the job ID that the work is contributing towards. A Related Issue (RI) is listed next to the NKR that provides more specific information regarding the general task.

If any team member is working on a stepping stone experiment/job, please comment on this issue so that we can add it to the list. We have a finite set of stepping stones, and outlining all of them should not be overwhelming. If a key result is finished, please comment on this issue so that we can check off the list

Atmos / physics work

• Radiation [Zhaoyi] (NKR: None for now)
• Topography [Akshay] (NKR: held_suarez_topo) RI: CA#408
• Moisture [Jia, Anna] (NKR: equil_baroclinic_wave, RI: CA#363
• Surface Fluxes [Jia, Lenka, Ben, Yair] (NKR: equil_held_suarez)
• EDMF [Yair, Charlie] (NKR: equil_idealized_aqua_allskyrad_edmf), we may want to add a stepping stone to this
  • Implicit time stepping

Features/support needed in ClimaCore

• spatially-varying boundary conditions [Ben] (NKR: equil_held_suarez)
  • stretched grid with correct weights [Simon] (NKR: held_suarez)
  • upwinding [Simon] (NKR: held_suarez)
  • Spherical limiters [Valeria, Simon] (NKR: equil_baroclinic_wave)
  • performance [Sriharsha] (NKR: held_suarez)

Land / Coupler work

• Coupler [Lenka, Ben]
  • equil_baroclinic_wave_slab - add slab ocean with MPI CC#39
  • sea_breeze_topog - Cartesian topography + land / ocean mask in building on the rising bubble setup CC#40
  • equil_baroclinic_wave_bucket - stub for the full land model CC#41
  • equil_baroclinic_wave_bucket_seaice - sea ice + land /sea mask CC#42
  • equil_baroclinic_wave_bucket_seaice_topog CC#43
• Land [Katherine]
  • Implicit time stepping (non-bucket land model)

List of key results (via buildkite jobs)

[completion %] job_id:          configuration flags
[ ✓ ] baroclinic_wave:                          --config SP
[ ✓ ] held_suarez:                              --config SP --HSF true
[0.6] held_suarez_topo:                         --config SP --HSF true --topo true
[ ✓ ] equil_baroclinic_wave:                    --config SP --moist equil --micro 0M
[ ✓ ] equil_held_suarez:                        --config SP --moist equil --HSF true --SF true --VD true --micro 0M
[ ✓ ] equil_idealized_aqua_grayrad:             --config SP --moist equil --SF true --VD true --micro 0M --rad gray
[ ✓ ] equil_idealized_aqua_clearskyrad:         --config SP --moist equil --SF true --VD true --micro 0M --rad clear_sky
[   ] equil_idealized_aqua_clearskyrad_topo:    --config SP --moist equil --SF true --VD true --micro 0M --rad clear_sky --topo true
[   ] equil_idealized_aqua_allskyrad_edmf:      --config SP --moist equil --SF true --rad all_sky --turbconv edmf
[   ] equil_idealized_aqua_allskyrad_edmf_topo: --config SP --moist equil --SF true --rad all_sky --turbconv edmf --topo true
[ ✓ ] radiative_equilibrium_clearsky:           --config SC --DCE false --rad clear_sky
[ ✓ ] radiative_equilibrium_gray:               --config SC --DCE false --rad gray
[ ✓ ] radiative_equilibrium_allsky:             --config SC --DCE false --rad all_sky
[0.6] radiative_convective_equilibrium:         --config SC --DCE false --rad all_sky --turbconv edmf

Physics acronyms

• DCE = Dynamical core equations
• topo = Topography
• GW = Gravity wave
• turbconv = TurbulenceConvection
• rad = Radiation (i.e., RRTMGP.jl)
• 0M = 0-moment microphysics
• VD = Vertical diffusion
• HSF = Held Suarez forcing
• L = Land
• M = moisture (dry, equil, non-equil)

Configuration acronyms

• SP = Spherical configuration - A full 3D spherical configuration, at coarse or fine resolution (determined by cli arguments)
• SC = Single Column configuraiton - A "single column" configuration where we run a spherical case with 6 faces, 1 element per face, 0-poly order and horizontally decoupled. These configurations with horizontal coupling (i.e., horizontal advection) + higher resolution should yield an effectively 3D spherical case

[szy21]

For the radiation, it is done but we only added radiative_equilibrium_allsky in the buildkite (maybe a better name would be radiative_equilibrium_clearsky because there is no cloud). We'll add a regression test to it once we finalize the details (e.g. interpolation). I think Dennis would like to change something in the command line arguments and driver, and we can add radiative_equilibrium_gray after that.

I think held_suarez would be a better NKR for stretched grid with correct weights. It's better to make sure it works on the sphere configuration. held_suarez can also be an NKR for upwinding.

[bischtob]

@szy21 , can you update the numbers for radiation runs?

[szy21]

This is a bit outdated, and we have new documents tracking this, so I'm closing the issue.