An error is reported when BHIST outputs ocean daily data

[ZhangZJ]

Hi, when I output the daily data of the ocean velocity in POP (VVEL and WVEL), the BHIST model is interrupted after half an hour, and the error is: , how can I solve it?

Thanks,

Zhang

 

[klindsay]

Please describe/show us what you modified to get daily POP ocean velocity.

Also, please provide the output from running ./pelayout from your case directory,
so that we can be sure about which component is generating the error message.

 

[ZhangZJ]

First I checked ocn.base.tavg.csh, where tavg_freq_opt = 'nmonth' 'nday', so I changed the coefficients in front of VVEL and WVEL from 0 to 1 in gx1v7_tavg_contents, corresponding to tavg_freq_opt = 'nday'. And write history_freq_opt='nday' , history_freq=1 in user_nl_pop, and modify STOP_OPTION=nyears , STOP_N=3 in env_run.xml.

(Using the same modification method, after changing the coefficients of UVEL and TEMP from 0 to 1 in gx1v7_tavg_contents, the model can run normally, and successfully output the daily data of UVEL and TEMP, but after modifying VVEL and WVEL, it reports an error after running for half an hour. )

Running ./pelayout in the case directory results in:


Thanks,

Zhang

 

[klindsay]

Zhang,

I have reproduced on the NCAR computer the error that you are getting.
I have a fix for the error that works here. The attached tar file has code mods for a fix.
The mods are based on CESM 2.1.3.
If you are using that version, you can untar the tar file contents
by running the command "tar xf tavg_fix.tar" while in the subdirectory
SourceMods/src.pop of your case root.
If you are using a different version of CESM, you will need to modify the fix for your version.
The tar file has a diff of the mods versus the CESM 2.1.3 so you can see what was changed.
There are just a few lines changed, so I'm optimistic that you can do this.

Please let us know if these mods fix your problem.

Please note that because of the way that POP computes the MOC diagnostic,
that changing the frequency of just the velocity fields causes the MOC diagnostic to not be generated.
I wanted to be sure that you know this.

Also, adding the history_freq_opt change to user_nl_pop is not needed. This setting is related to
generating instantaneous snapshot output from POP.

I will move this question to the POP section of the forum, which is a more natural location for it.

Keith

 

[ZhangZJ]

Hi Keith,

Thank you so much for your detailed response! My problem is solved.

Best Wishes,
Zhang

 

[klindsay]

Zhang,
Thanks for confirming that the mods fix the problem for you.
I've filed an issue ticket in the POP github repository to track the problem.
We should be able to get the fix included in future releases.
That said, I'm not certain about future release timelines.
Keith

 

[ZhangZJ]

There is an additional question:

The grid resolution I run is f19_g17. In the outputs, there are missing values in the variables ULAT, ULONG, TLAT and TLONG that define the latitude and longitude. What could be the cause?

Zhang