O2.5.3: Run single column model coupled runs of ClimaAtmos + ClimaLand soil model

[juliasloan25]

The Climate Modeling Alliance

Software Design Issue 📜

Purpose

O2.5.3: Run single column model (SCM) coupled runs of ClimaAtmos + ClimaLand soil model with different component model timesteps and verify stability.

This works toward O2.5:

O2.5 Couple the land surface model with soil/canopy/snow to atmosphere model and update surface flux computations to enable fractional cover with land properties (e.g., snow and canopy). [Possibly move up if needed for good AMIP.]

Setting up and maintaining single column coupled runs of ClimaAtmos and the more complex ClimaLand model is a step towards global coupled simulations. Using the soil model rather than the full soil+canopy+snow model will allow us to break down the complexity further.

Cost/Benefits/Risks

costs: developer time
benefits: exercising coupler code in a wider range of cases will help point us towards rigidity in the code
risks: there have been some problems with ClimaCore PointSpace support in the past, and more may surface as we implement this SCM example

People and Personnel

• Lead: @juliasloan25
• Collaborators/Reviewers: @akshaysridhar @kmdeck @szy21

Components

• As a first step, run SCM example of coupled ClimaAtmos and bucket (this should be straightforward)
• New land model option in ClimaCoupler.jl
• Extend any flux exchange methods that weren't caught in O2.5.2: Generalize coupling and surface fluxes to accommodate full land model #1031
• Add new configuration using soil land model

Inputs

• Generalized coupler infrastructure from O2.5.2: Generalize coupling and surface fluxes to accommodate full land model #1031

Results and Deliverables

• Experiment in CI running SCM coupled ClimaAtmos and bucket
• Experiment in CI running SCM coupled ClimaAtmos and soil land model
  • Make sure simulations are short enough that they don't increase overall CI time

Tasks

Preview Give feedback

1. Add SCM atmos + bucket experiment #1042
   🍃 leaf enhancement +2
   
2. Support multiple land model options in ClimaCoupler.jl
3. Extend any flux exchange methods that weren't caught in O2.5.2: Generalize coupling and surface fluxes to accommodate full land model #1031
4. Add experiment using soil land model

SDI Revision Log

• opened 17 October 2024 by @juliasloan25

CC

@tapios @sriharshakandala @charleskawczynski @cmbengue

[tapios]

This looks good to me.