ClimaAtmos.jl is a library for building atmospheric circulation models that is designed from the outset to leverage data assimilation and machine learning tools. We welcome contributions!
Integration of the EDMF hosting code in Clima atmos and changing the mean and updrafts equations. We are doing so by moving (copying) the src and driver from TC.jl to CA so that we can use the compressible host model in CA. The end goal is to retire TC.jl (not now though).
Cost/Benefits/Risks
Cost - time it takes for porting the code from TC.jl to CA; added dependencies in CA
Benefits - Ability to avoid working with releases in TC.jl, using CA compressible model (not building one in TC).
Risks - Compiling CA might be slower. Code duplication with TC.jl for a period of time until TC.jl is retired.
People and Personnel
(@ the personnel required/requested to implement the software solution)
Lead: @charleskawczynski
Collaborators: @yairchn , @trontrytel
Reviewers: @trontrytel , @yairchn
Components
Step 1: move TC src and driver to CA #583
Step 2: Add addition TC cases to maintain behavior #590
Step 3: Add update pressure from GM density in update_aux (prep for compressible) #599
Step 4: compute env w based on grid mean and updrafts in update_aux (prep for compressible no. 2) #641
Step 5: Run compressible GABLS (turbulence only) in CA (explicit solver - slow) - compare results #668
Step 6: add the EDMF filter into CA callbacks #678
Step 7: Unify compressible flag in TC and CA components #683
Step 8: Run compressible Bomex (convection and turbulence) in CA - we found that this requires ignoring the dycore w when backing out the environment w.
Step 9: Make Sure that the initial state is a discreetly balanced - PR #724.
Step 10: Change updraft variable to total energy (consistent only in a compressible model) - PR #718
Step 11: Add implicit solver for EDMF with energy as updraft variable see issue #789
[x] Child Issue #588
[x] Child Issue #598
[ ] Child Issue #789
Results and deliverables
Compressible Bomex simualtion
Compressible implicit Bomex simulation + test
Compressible implicit Bomex simulation with energy as updraft / environment variable + test
Purpose
Integration of the EDMF hosting code in Clima atmos and changing the mean and updrafts equations. We are doing so by moving (copying) the src and driver from TC.jl to CA so that we can use the compressible host model in CA. The end goal is to retire TC.jl (not now though).
Cost/Benefits/Risks
Cost - time it takes for porting the code from TC.jl to CA; added dependencies in CA Benefits - Ability to avoid working with releases in TC.jl, using CA compressible model (not building one in TC). Risks - Compiling CA might be slower. Code duplication with TC.jl for a period of time until TC.jl is retired.
People and Personnel
(@ the personnel required/requested to implement the software solution)
Components
Step 1: move TC src and driver to CA #583 Step 2: Add addition TC cases to maintain behavior #590 Step 3: Add update pressure from GM density in update_aux (prep for compressible) #599 Step 4: compute env w based on grid mean and updrafts in update_aux (prep for compressible no. 2) #641 Step 5: Run compressible GABLS (turbulence only) in CA (explicit solver - slow) - compare results #668 Step 6: add the EDMF filter into CA callbacks #678 Step 7: Unify compressible flag in TC and CA components #683 Step 8: Run compressible Bomex (convection and turbulence) in CA - we found that this requires ignoring the dycore w when backing out the environment w. Step 9: Make Sure that the initial state is a discreetly balanced - PR #724. Step 10: Change updraft variable to total energy (consistent only in a compressible model) - PR #718 Step 11: Add implicit solver for EDMF with energy as updraft variable see issue #789
Results and deliverables
Compressible Bomex simualtion Compressible implicit Bomex simulation + test Compressible implicit Bomex simulation with energy as updraft / environment variable + test
Proposed Delivery Date - Sept 15