Missing section volume transports and global average figures

[YanchunHe]

(A. Moree)

I noticed some changes in the output of the diag package. Before, I would successfully generate section transports and global averages, but this does not work at the moment:

For example compare

http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs3950to4000-obs/indexnew.html

to

http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs2950to3000-obs/indexnew.html

This NOINYOC_tn21_20A run has a different grid, causing some of the information to already be missing in the 2950-3000 diag (which is OK), but in the 3950-4000 diag I do not get any section transport anymore.

The logs are here: /projects/NS2345K/noresm_diagnostics/out/anne89/MICOM_DIAG/config/NOINYOC_tn21_20A/logs

It says several variables are not present, but that would be surprising because I did not change the output variables during the run. The output files are here: /projects/NS2980K/NORSTORE_OSL_DISK/NS2980K/moree/NOINYOC_tn21_20A/ocn/hist

The CICE and HAMOCC output is the same as before as far as I see now.

Do you see what happened here?

[YanchunHe]

I updated the code of transport part recently (about a month ago), and it then checked whether the transport sections are exactly 15 or 17, otherwise, will not plot any of the transports. This has not been checked in previous code, so it will anyhow show some figures there. However, you can find the figures under:http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs3950to4000-obs/set1/

The reason that they are not shown is to avoid any missing leading (incorrect) information of the name of sections. But I do need to improve the flexibility of this part to have arbitrary number of sections with names defined in the netcdfile (from coordinate 'section'). I will note down and this and update later. Since I am quite busy with urgent CMOR work recently (and some others). Please wait for a short while, sorry for this..

For the global temperature and salinity. I also did update the code, if the global averages, tempga/salnga already exist, they will be used, instead of calculate from temp/saln,dz.

I see in your log file, the timeseries files are already existed, so in the log files, it is skipped to calculated again. Therefore, I can't really why it can't find temp/saln/dz and get the global averages. Could you diagnose NOINYOC_tn21_20A for a new period for test? for example, diag_run -m micom NOINYOC_tn21_20A -s 3950 -e 3955 -i ... and then send me the log file again.

[annemoree]

OK, I understand and will use the 'set' folders to find the missing figures for now.

About the test you suggested: The new period has the same problem of missing section transports and global averages due to missing variables (even though the variables are there): http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs3950to3955-obs/indexnew.html
The logs are in /projects/NS2345K/noresm_diagnostics/out/anne89/MICOM_DIAG/config/NOINYOC_tn21_20A/logs/out_190911_155809.log

Thanks for your help