-
Notifications
You must be signed in to change notification settings - Fork 6
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Understanding which values are missing from varlist #435
Comments
Hi Paolo, thanks for your remarks. You are correct, the whole point of having central varlists is to have a benchmark that everybody should be able to cmorize without errors. The 3 atmospheric variables you mention are zonal means, which are not (yet) supported in ece2cmor3. The thing is... we are planning to support zonal means and climatological variables, so I am not sure whether we should exclude them now... |
In the |
For completeness, I report all the error/warnings that I get with the 0.9.9 and historical AOGCM varlist. We can see if these are known or unlisted yet.
It takes about 15min for NEMO with 1 core and 50 min for IFS with 12 cores on CCA |
Thanks! |
Hi Paolo, can you check whether there exist u and v fields in the grib file with level type 109 (model levels) and level value 100? In that case EC-Earth is writing height-levels as model-levels, I think this may be a bug in the IFS... |
Uhm, it seems this is the case. Couldn't be that this is a default IFS feature? I am wondering this since I don't know if we are extracting any other variable on a height level...
|
Concerning the performance: at the KNMI HPC for IFS with 1 node (28 cores) it takes about 23 minutes. Seems in agreement with 50 min using 12 cores by @oloapinivad, who's performance is 7% better then ours at KNMI taking this rough numbers. |
Could we agree on how many variables can be obtained with ece2cmor3 1.0.0 and in the different experiments, at least in the most standard ones? for historical Has anybody else tried the same setup? |
So most of the skipped variables under point 4 are PISCES variables (see #486), except from
I guess these are due to an older version (set)? |
I am closing this issue after #486 is solved. Issue #490 will take over for the current version, note that as long we don't have a test-all-ifs test, we have to report for IFS variables separately there for piControl, historical and scenario etc. Currently piControl has been reported in #490 and historical and scenario have to be added. |
Hi all,
I am currently able to cmorize with success (i.e. ending the ece2cmor command without error) all variables from
cmip6-data-request-varlist-CMIP-historical-EC-EARTH-AOGCM.json
However, when I check I can see some of them are missing. I have this bash script that compares variables from the varlist.json and obtained output
Of course, this take into account everything. I am aware that Oclim and Oyr are not available and that we are voluntary dismissing model levels. However, there are a few other errors/warnings (that does not affect the correct workflow of ece2cmor) that I am not 100% sure they should be ok.
For instance
or
Another example are the biogeochemical variables included in NEMO varlist, which I am not sure should be there.
So my question: having the varlist.json for each experiment is great, but is there a way to have also the subset which is correctly cmorized by ece2cmor3? I am asking this - even if I am aware is quite tough thing - because I wonder how each research center will proceed. For example, do KNMI, BSC and SMHI all get the same number of cmorized variables for the DECK runs?
Thanks for any hint - and sorry if this is an already known issue.
The text was updated successfully, but these errors were encountered: