Log the merge of the task load merge branch

[treerink]

In order to be able to handle the preferences of certain table - variable combinations depending on the used EC-Earth3 model configuration required some deeper changes in ece2cmor3. With the preference handling we address the "double counting" issue #224. Instead of the discussed preferences file a preferences python routine has been implemented (resources/prefs.py) in order ot be more flexible. Furthermore a rewrite of the taskloader was needed (which needs a fix by now, see #409).

[treerink]

Because the developments of the task load prefs branch caused many (in first instance) unexpected differences in the whole chain of identification and genecec, we added a dictionary order on the checkvars.py output (the xlsx files). This caused order differences in the standard genecec output, the control output files. But with this order the results of checkvars and genecec could be checked much better between the master and the branch (and will be helpful in future developments as well).

[treerink]

Finally the output of checkvars.py was the same in the task load prefs branch with the master, except for one improvement: The comment containing the expression for the "double counting" variables which are thus subject to the preference issue, contains now both options, thus the expression for both cases is listed, which is an improvement.

[treerink]

There are differences in the control output files. Before tasks (preferences) depended on the realms, which was not working so well, now this has been removed and that causes that certain table - variable combinations are not longer ignored based on their realm, and thus now pop up in the control output files. Indeed there are additional grib codes in the new version. This concerns the following grib codes: 131 (ua), 132 (va) 135 (wap) and 185 (cltc). They are usually part of MFP3DFS MFPPHY,.

[treerink]

Further in the new version the variable vegFrac is more often part of the LPJ-GUESS instruction file, although we did not trace back completely the reason for this difference, both the final instruction files and the log of genecec seem to show that this is an improvement in the end.

[treerink]

Finally one difference in the ppt files concerns the model levels specified in RFP3P in the following diffs:

diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-long-eq/ppt0000000000 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-long-eq/ppt0000000000
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-long-eq/pptdddddd0600 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-long-eq/pptdddddd0600
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-full/ppt0000000000 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-full/ppt0000000000
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-full/pptdddddd0600 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-full/pptdddddd0600
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-strat/ppt0000000000 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-strat/ppt0000000000
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-strat/pptdddddd0600 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-strat/pptdddddd0600
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-surf/ppt0000000000 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-surf/ppt0000000000
diff -r cmip6-output-control-files-2019-v17/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-surf/pptdddddd0600 cmip6-output-control-files-2019-v13/VolMIP/cmip6-experiment-VolMIP-volc-pinatubo-surf/pptdddddd0600

Where actually the new version seems correct, i.e. the same as usual:

<             30000.0, 25000.0, 20000.0, 17000.0, 15000.0, 13000.0, 11500.0,
<             10000.0, 9000.0, 8000.0, 7000.0, 5000.0, 3000.0, 2000.0, 1500.0,
<             1000.0, 700.0, 500.0, 300.0, 200.0, 150.0, 100.0, 70.0, 50.0,
<             40.0, 30.0, 20.0, 15.0, 10.0, 7.0, 5.0, 3.0, 
---
>             30000.0, 25000.0, 20000.0, 15000.0, 10000.0, 7000.0, 5000.0,
>             3000.0, 2000.0, 1000.0, 500.0, 100.0, 

While the bottom in the diff is from the previous version.

[treerink]

Hi Gijs,

Here are the logs of genecec for the VolMIP-volc-long-eq experiment, accompanying the previous post here.

The log from the task load prefs branch: sub-log-15.txt
The log from the previous master branch: sub-log-13.txt

[treerink]

Anyway closing the issue.