Recipe and diagnostics to plot climate change hotspots: Cos et al., ESD 2022

[pepcos]

Description

This pull request aims to close the issue #2478 opened by @sloosvel

It contains the recipe and diagnostics to produce figures 2, 3, S1, S2 and S4 from Cos et al. 2022

Figure 2: Future Mediterranean climate change hotspot projected by CMIP5 and CMIP6 for a 8.5 Wm-2 radiative forcing
Figure 3: Scatter plot for the Mediterranean vs Global warming as projected by CMIP5 and 6 for three scenarios.
Figure S1: Same as Fig.2 but for 2.6 Wm-2 radiative forcing
Figure S2: Same as Fig.3 but for Mediterranean pr vs Global pr
Figure S4: Same as Fig.3 but for Mediterranean pr vs Global tas

Closes #2478

---

Before you get started

• ☝ Create an issue to discuss what you are going to do

Checklist

It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.

• 🛠 This pull request has a descriptive title
• 🛠 Code is written according to the code quality guidelines
• 🛠 Documentation is available
• 🛠 Tests run successfully
• 🛠 The list of authors is up to date
  - [ ] 🛠 Any changed dependencies have been added or removed correctly
• 🛠 All checks below this pull request were successful

New or updated recipe/diagnostic

• 🧪 Recipe runs successfully
• 🧪 Recipe is well documented
• 🧪 Figure(s) and data look as expected from literature
• 🛠 Provenance information has been added

### New or updated data reformatting script

- [ ] 🛠 Documentation is available
- [ ] 🛠 The dataset has been added to the CMOR check recipe
- [ ] 🧪 Numbers and units of the data look physically meaningful

---

To help with the number of pull requests:

• 🙏 We kindly ask you to review two other open pull requests in this repository

[sloosvel]

Very nice work @pepcos ! I am kind of jealous that you don't get to fight Codacy, not sure why the check does not appear... I will start reviewing but first, can you try to set a nice descriptive title for the pull request?

[sloosvel]

Ok here we have the checks: https://app.codacy.com/gh/ESMValGroup/ESMValTool/pullRequest?prid=9121004

Try to address the easy ones first like thing related to missing docstrings, long lines, blank lines, and if you don't know how to solve them let me know commenting in the line.

Then we can move on to the more "complicated" ones.

[sloosvel]

The run got stuck in jasmin, but I am getting these warnings when collecting the provenance for the first diag:

WARNING Ancestor file(s) /work/scratch-nopw/sloosvel/recipe_climate_change_hotspot_20220505_071053/preproc/pr_cmip6_26/pr/CMIP6_ACCESS-CM2_Amon_historical-ssp126_r1i1p1f1_pr_gn_1980-2100.nc specified for recording provenance of /work/scratch-nopw/sloosvel/recipe_climate_change_hotspot_20220505_071053/work/pr_cmip6_26/pr_cmip6_26/rolling_mean_regional_jja.nc, created by diagnostic script /home/users/sloosvel/ESMValTool/esmvaltool/diag_scripts/cos22esd/climate_change_hotspot.py in task pr_cmip6_26/pr_cmip6_26 is a string but should be a list of strings

Nevertheless, how would it look if the datasets that are not available/workin in Jasmin are commented out?

I personally think it does not look very clean but go for it if you think it's better.

[sloosvel]

Ok @pepcos , the recipe finished successfully by removing the dataset that is stored in a strange way. I have also added a fix for another dataset in , because it was crashing at the regrid step.

As discussed offline, I removed the ensemble_statistics step as it was not impacting the results and it just added to the run time.

Could you please try to take care of the warning in the provenance?

[sloosvel]

Thanks @pepcos , the provenance warnings do not appear anymore.

[sloosvel]

@pepcos I would say that from the technical point of view this looks fine to me, everything runs smoothly and other, than one codacy error that I think can be overlooked, the code looks fine. I will wait for the scientific results to be checked before giving the approve, in case something needs to be changed.

[rmarcos-ub]

Hi everyone! Here is my take on the scientific review.

From a scientific standpoint, the results obtained from the recipe seem fully consistent among them and in accordance with the current community knowledge as, for instance, regarding the warmer results obtained in CMIP6 when compared to CMIP5. On the other hand, the scenario combination plots offer a readily available tool to identify sound hotspot behaviours as well as variations within CMIP experiments, as it can be seen in the Central Australia domain. In summary, it seems the science parts are doing a good job on their expected function. Regarding the documentation, it is written in a clear and straightforward language, with examples that are easy to follow for any potentially interested user.

[zklaus]

Thanks, @rmarcos-ub. Excellent review!

[bouweandela]

@pepcos Could you try merging the main branch into this branch, to see if the tests run then?

[pepcos]

It seems only the documentation tests and codacy are run @bouweandela

[sloosvel]

I think it may be because you did not link your account to CircleCI, in any case all tests are passing now. You can try to push again some small change (add a dot in the title?) or something and then remove it just for the sake of checking if you trigger the tests now. Or just leave it for another time, whatever you prefer.