one line bug fix to address screen level temperatures #313

[har917]

CABLE

Thank you for submitting a pull request to the CABLE Project.

Description

A single line bug fix to address cases where the diagnosed screen level temperature does not deviate from the reference level air temperature. This only impacts some sites with tall vegetation.

Note that this change addresses the immediate issue - other aspects involving code structure and when in the calling sequence the screen level diagnostics should occur have been left untouched.

This is intended as a bug fix impacting only the screen level temperature and humidity - all other variables should remain as before. However this change is - by design -not bitwise comparable with the MAIN so may fail Benchcab testing.

Fixes #313

Type of change

Please delete options that are not relevant.

• [x] Bug fix

Checklist

• [x] The new content is accessible and located in the appropriate section.
• [ ] I have checked that links are valid and point to the intended content.
• [ ] I have checked my code/text and corrected any misspellings

Testing results

benchcab/me.org results: https://modelevaluation.org/analyses/anywhere/32ykm3nn4EfJCro8q/s6k22L3WajmiS9uGv/fHu3wwScyjcNpjApy/all (added after approval as the analyses needed some debugging)

---

📚 Documentation preview 📚: https://cable--316.org.readthedocs.build/en/316/

[har917]

benchcab log and example failure text attached - as expected diffs were detected in the screen temperature and humidity variables. benchmark_cable_qsub.sh.o118801918.txt AU-Ctr_2010-2017_OzFlux_Met_S0_R0_R1.txt

[har917]

modelevaluation.org failed for some reason. error message:

Command failed: R --no-save < /pals/executions/RG6rfN9WSQzoHZi7a/rRunner.r Error in BenchmarkInfo[[m]] : subscript out of bounds Calls: source ... withVisible -> eval -> eval -> Load_PropertiesAllMO Execution halted

See me.org

I also note that one of the benchcab builds failed - it seems to have failed to set up a container (at which point there's no ContainerID for later parts of the build).

Sending to review for advice on how to proceed.

[har917]

Tried rerunning me.org but with the 'benchmark added' - different error message has been generated

Command failed: R --no-save < /pals/executions/ofP9FdSfunJoCaXnR/rRunner.r Error in if (applyQC & (SiteData[[v]]$qcexists & forceqc$filterexists)) { : argument is of length zero Calls: source ... withVisible -> eval -> eval -> lapply -> FUN -> Load_Site Execution halted

[har917]

Note that this bug fix needs to make it's way to CABLE3 as used in ESM1.6 and ACCESS-3

[ccarouge]

It turns out the analysis script in me.org is currently broken. @gabsun says it shouldn't take too long to fix but he is just recovering from being sick so the fix won't come in right away.

It would be good to have the results of the analysis for this bug fix as it only affects some sites.

[ccarouge]

The issue with the automated checks is now solved @har917 and that check has been successfully run.

[ccarouge]

Since the analysis script in modelevaluation.org is currently not working, this will be reviewed without the test results.