new tutorial ideas

[wwieder]

Here are some ideas for some mini tutorials to develop:

• [ ] Anomaly forcing for future scenarios

• [ ] Annotated table of contents of tutorials in README

• [ ] Additional FATES examples

• [ ] PPE-like work, maybe around medlyn slope or other topic?

• [x] CLM-SP example for hydrology questions

• [ ] Code modifications

• [ ] Hillslope model

• [ ] ... multilayer canopy [eventually :)] ---- Additional tutorial ideas ----

• [ ] Needs from education committee

• [ ] BFB testing (maybe focused more on developers).

• [ ] Regional tests

• [ ] Bug fixes

• [ ] ILAMB or ModelEvaluation.org

• [x] Namelist change: switch from PHS to BTRAN,

• [x] Changing parameters

  • looking at existing output may be more helpful than doing new runs.
  • One at a time experiments for modifying parameters

• [x] FATES-SP case (or just CLM-SP).

  • Adrianna suggested running initialized FATES data from NEON AOP information for FATES footprint.
  • NorESM group could also have some work to bring in here.

• [x] Experiments, and spinup

• [x] Additional simple visualization examples GPP with different frequencies (including writing out a .csv of the data for use in other analyses)

• [x] changing history file output (variables, frequency, etc) -> name_list modifications

• [x] modifying surface datasets (e.g. C3 vs. C4 grasses)

• [x] Updating streams (PRISM)

[wwieder]

Expectations for these tutorials:

• Simple, focused on a single topic.
• Tested, this is likely easiest to develop on Cheyenne, and should be simple to translate to the cloud.
• Acknowledged, make sure we give credit to the developers
• Potentially organized more sequentially (especially for education-inspired needs)?

[adrifoster]

This is all great.

How about also

• [ ] using NEON (or other) data (flux/AOP/obs) to initialize FATES and/or CTSM
• [ ] using NEON (or other) data (flux/AOP/obs) to compare to FATES and/or CTSM

I'm thinking more hands-on than what was done previously (so full stack of download data from NEON to running the model, etc.)

[katiedagon]

What about demonstrating tools to post-process single column output? Either PFT-level or CLM output indexed by 'lndgrid', for example when looking at coupled runs with CAM-SE. Maybe this doesn't apply in a single point case, but I do think it is a common use case that comes up on the forums, zulip, etc.

[wwieder]

This would be helpful @katiedagon. I also wonder what the overlap between tutorial material and the CTSM_PY gallery should be? It seems like @samsrabin is starting to have more activity on this repo? It would be good to have a design for where we'd like these contributions to come in from the LWMG.

[rosiealice]

What about running a (fully dynamic) FATES case with modified PFT parameters? The tools charlie developed for chopping and changing the PFT files should make this relatively straightforward? I'm sure we have enough existing scripts that we can piece such a thing together, given a general 'tutorial' template.

[wwieder]

@rosiealice this would be excellent! I'm happy to help with this, especially if you have tools and a workflow that we can build off of.