Coupled Timestepping Interface

[jb-mackay]

Introduces the coupled timestepping interface in which component models provide step! methods used in the coupled system driver. Updates to coupler state interface will occur in a separate PR.

To do:

• Implement basic run! function that executes simulation timestepping
• Update unit tests
• Coupled atmos-ocean-land example implementation (Future PR)
• Update dependencies
• Update docs

[codecov]

Codecov Report

Merging #16 (e937d7d) into main (d548586) will increase coverage by 7.97%.
The diff coverage is 47.05%.

❗ Current head e937d7d differs from pull request most recent head 555713f. Consider uploading reports for the commit 555713f to get more accurate results

@@            Coverage Diff             @@
##             main      #16      +/-   ##
==========================================
+ Coverage   57.14%   65.11%   +7.97%     
==========================================
  Files           4        4              
  Lines          63       43      -20     
==========================================
- Hits           36       28       -8     
+ Misses         27       15      -12     

Impacted Files	Coverage Δ
src/CoupledSimulations/coupled_simulation.jl	0.00% <0.00%> (ø)
src/CouplerMachine.jl	100.00% <ø> (ø)
src/CouplerState/coupler_state.jl	72.72% <62.50%> (ø)
src/CoupledSimulations/clock.jl	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d548586...555713f. Read the comment docs.

[jb-mackay]

Documentation preview: https://clima.github.io/CouplerMachine/previews/PR16/couplerstate/

[LenkaNovak]

Looks good! Do we want to remove the ClimateMachine-compatible interface completely though? I wonder whether it's still useful to keep it somewhere until have a way to support the CM with the new interface. I guess we can always go back the commit history once we come to that bridge though...

[jb-mackay]

Looks good! Do we want to remove the ClimateMachine-compatible interface completely though? I wonder whether it's still useful to keep it somewhere until have a way to support the CM with the new interface. I guess we can always go back the commit history once we come to that bridge though...

@LenkaNovak I think it is straightforward to use the new interface with a ClimateMachine simulation via ODESolvers.general_dostep!.