Main menu


How to couple the co2 cycle between ocean and atmosphere

7 posts / 0 new
Last post
How to couple the co2 cycle between ocean and atmosphere

Hi, all

I made a fake abiotic co2 cycle in the CESM, that is coupled between ocean and atmosphere model. Specifically, I added abiotic co2 mixing ratio in cam_export subroutine, and abiotic co2 flux to cam_in. Correspondingly, abiotic co2 are added in ocn_in and the the flux in ocn_out. Also, the  Sa_co2abio and Faoo_fco2abio_ocn are created in seq_flds_mod (driver). Everything is following the co2 cycle (more specifically, co2diag). But, when I run the model, the ocean cannot get abiotic co2 from coupler.

The error occurs in abio_dic_dic14_init_sflux, “POP aborting... 43918 644: abio_dic_dic14_init: abio_dic_dic14 module requires atmopsheric CO2 from the flux coupler and it is not present”.

Note: abio_dic_dic14_init_sflux is a module for radiocarbon tracer. 

If I give ocean the diagnostic co2, which is buildt in CESM, it can receive it. Everything is following diagnostic co2, but the ocean model can not receive my own abiotic co2. quite werid.


Any suggestions?



It looks like you are running into problems in your case c14_cycle.e13.Bi1850C5.f19_g16.PI.coupling.test on yellowstone. I've taken a look at it. I think a principle problem you're running into is that some of the files in SourceMods/src.pop have the extensions of their filenames changed from the original .F90 to .f90. POP's build script is not using these .f90 files as replacements for the original .F90 files, so I don't think your mods are being used. I think some of your problems will be resolved if you rename your .f90 SourceMods files to .F90, the filename extension on the original file.

I'm not sure if the renamed filename extension in SourceMods/src.drv/seq_flds_mod.f90 is a problem or not. Either way, I advise not changing the file extension from that of the original file.

Hope that helps, Keith


Woo, It works. Thanks! this is a problem I wont find if no one tells me. 

A further question, How do you know my case on yellowstone? That's magic. 



Glad to hear that this helped out.

Thinking that seeing your case would make it easier to diagnose what went wrong, I looked for your case. I got your name from your forum profile and saw that you have an account on yellowstone. The default location for run directories on yellowstone is /glade/scratch/$USER, so I looked at the most recently written to directory there. This led to c14_cycle.e13.Bi1850C5.f19_g16.PI.coupling.test. I looked at the log files in that case's run directory and saw the error message you reported, so I figured I was on the right track. The first line of $EXEROOT/bld/atm/obj/Filepath points to $CASEROOT, where I looked to try to diagnose the problem. Looking in SourceMods, I noticed the unusual .f90 file extensions. This only worked because you are running on yellowstone, with the default run directory, and your directories are world readable.



That's a incredible! Can I ask a specfic question about CO2 cycle? From the raw code, CO2 shall be the combination of CO2FFF + CO2LND + CO2OCN. 

I tested two different cases for this in CESM. The first one is using compost: Bi1850C5, which includes cam5 and clm4. The co2 surface distribution is the same with CO2_OCN, but CO2_FFF and CO2_LNd has nonzero value and they are equal. Another case is using compost: B1850C5L45BGC, which includes CAM5 and CLM4.5. In this case, the CO2 surface distribution is not the combination of the three components. None combination of these components can lead to CO2. So, what’s the correct thoughts on earth?


NOTE: all the cases are just coupled the all CESM components, not using the CO2 boundary flux file. The setting in cpl namelist is flds_co2c=true(the one shall be the most interactive co2 cycle)


Unless you are specifically using the developmental water isotope branch of CESM, you should not be using the Bi1850C5 compset.  This is part of an new upcoming feature of CESM and hooks are just starting to be placed in the CESM code for it.  The WISO in the longname refers to water isotopes.   Most of the isotope features are still under development and are not part of the alpha/beta releases of CESM.


Hi, I'm also confused by the values of CO2_FFF, CO2_LND and CO2_OCN, in the output files of CAM modules, absolutely CO2_FFF+CO2_LND+CO2_OCN≠CO2. Have you figure this out?


Log in or register to post comments

Who's new

  • Nicholas.Davis@...
  • numarsanifa@...
  • bingdian_46@...
  • mxy2832029@...
  • nthg2000@...