Hi,
Please feel free to move this post to where it may need to be.
I am running CESMv2.1.3 on Plasma at the University of New Hampshire with G1850ECOIAF, no carbon isotopes, by3 grid resolution. I have attempted to add a new interior boundary forcing for hydrothermal vent input of dissolved organic carbon (DOC). Currently, the model has Fe from vents, so I followed the fesedflux code and duplicated it for my docventflux code, which gets added to interior tendency forcing doc_ind in marbl_interior_tendency_mod.F90 as
My edits follow (as best as I can) the FESEDFLUXIN (fesedflux) interior forcing, where I read in DOCSEDFLUXIN (docventflux) variable from a nc file that works. I checked it by renaming the variable to FESEDFLUXIN and running with out of the box model with several resubmits before making code edits. DOC (mmol/cm^3/s) is inputted into the ocean at the kmt where seafloor is 0-1 Myr.
The nc file is read in with user_nl_pop edit
My code edits are in 9 files spread out between namelist, yaml, pop, and marbl:
After uploading the code changes and running with DEBUG=T, my model builds and crashes in 50 seconds on day 00010102. In my cesm.log, I get a floating point divide by 0 error message with ww3 model error below (w3iogomd.f90:508).
I made no code changes in the wav model. Also, prior out-of-the-box runs worked before I edited the code.
I guess the question is what is the feed through in marbl and pop that causes the ww3 to crash? Is there any documentation on adding interior boundary forcings in pop model that gets fed into marbl?
I also note I ran with compset G1850ECO, and got the cesm.log error below after day 00010817. I could modify the Jint_Ctot_thres to log_noerror, but it would likely crash with the same ww3 error above. I got this same result when I duplicated this code but with DOC removal by vents. In the future, I will combine the source and sink of DOC from vents into one model.
Thank you,
James
Please feel free to move this post to where it may need to be.
I am running CESMv2.1.3 on Plasma at the University of New Hampshire with G1850ECOIAF, no carbon isotopes, by3 grid resolution. I have attempted to add a new interior boundary forcing for hydrothermal vent input of dissolved organic carbon (DOC). Currently, the model has Fe from vents, so I followed the fesedflux code and duplicated it for my docventflux code, which gets added to interior tendency forcing doc_ind in marbl_interior_tendency_mod.F90 as
My edits follow (as best as I can) the FESEDFLUXIN (fesedflux) interior forcing, where I read in DOCSEDFLUXIN (docventflux) variable from a nc file that works. I checked it by renaming the variable to FESEDFLUXIN and running with out of the box model with several resubmits before making code edits. DOC (mmol/cm^3/s) is inputted into the ocean at the kmt where seafloor is 0-1 Myr.
The nc file is read in with user_nl_pop edit
My code edits are in 9 files spread out between namelist, yaml, pop, and marbl:
- build-namelist, namelist_defaults_pop.xml, namelist_definition_pop.xml,
- diagnostics_latest.yaml,
- ecosys_forcing_mod.F90,
- marbl_diagnostics_mod.F90, marbl_init_mod.F90, marbl_interface_private_types.F90, and marbl_interior_tendency_mod.F90.
After uploading the code changes and running with DEBUG=T, my model builds and crashes in 50 seconds on day 00010102. In my cesm.log, I get a floating point divide by 0 error message with ww3 model error below (w3iogomd.f90:508).
I made no code changes in the wav model. Also, prior out-of-the-box runs worked before I edited the code.
I guess the question is what is the feed through in marbl and pop that causes the ww3 to crash? Is there any documentation on adding interior boundary forcings in pop model that gets fed into marbl?
I also note I ran with compset G1850ECO, and got the cesm.log error below after day 00010817. I could modify the Jint_Ctot_thres to log_noerror, but it would likely crash with the same ww3 error above. I got this same result when I duplicated this code but with DOC removal by vents. In the future, I will combine the source and sink of DOC from vents into one model.
Thank you,
James