Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

initialize cesm1.2.2 hybrid case with cesm1.0 binary restar files: 'POP aborting... could not find field in binary header file'

Dear all,

I am using cesm1.2.2 with intel compiler version 14 on a Cray XC30 machine with the B_1850-2000_BGC_BPRP compset and resolution 0.9x1.25_gx1v6.

I'm trying to initialize a cesm1.2.2 hybrid-start case with cesm1.0 binary restart files. Configuration and compilation works perfectly fine, but when running, the POP module crashes with the following error message:

"POP aborting... Could not find field in binary header file".

According to the POP2 FAQ (http://www.cesm.ucar.edu/models/cesm1.2/pop2/doc/faq/#init_hybrid_binary) this should generally work. The restart file itself, the header file as well as the rpointer.ocn.restart file are in the execution directory and the rpointer file has the specification "RESTART_FMT=bin". Furthermore, I considered the suggestion in this previous post (https://bb.cgd.ucar.edu/pop-aborting-could-not-find-field-binary-header-file), but this didn't work neither.

When I set up a hybrid run with restart files from cesm1.2 everything works perfectly fine. As these restart files are in a netcdf format, I suppose there error is somehow due to the binary restart file for pop. Has anyone got a suggestion on this? 

Best wishes
Sandro
 

njn01

Member
When pop2 writes a restart file in binary format, it also writes an ascii companion file, the restart header file, which contains a list of all restart variables that are contained in the binary restart file.  Your error message is telling you that according to your restart header file, the restart field that you are trying to read is not available on the binary restart file.This would occur, for example,  if you set up your new hybrid case with an OCN_TRACER_MODULES option that was not active in the original run that created the restart file.  For example, the original run might have had OCN_TRACER_MODULES = 'iage' and your new hybrid case has OCN_TRACER_MODULES = 'iage cfc' or OCN_TRACER_MODULES = 'iage ecosys'.  
 

klindsay

CSEG and Liaisons
Staff member
njn01 is correct that this error message is indicating that POP is not finding a field in its restart file that it expects to find.I suspect that the reason for this is that you are trying to run with the ocean ecosystem, which B_1850-2000_BGC_BPRP implies, in CESM 1.2.2, while using restart files from CESM 1.0. The ocean ecosystem model changed considerably in CESM 1.2 and is not backward compatible with the ocean ecosystem model in CESM 1.0. In particular, you cannot use ocean ecosystem simulations from CESM 1.0 (or 1.1) to initialize ocean ecosystem simulations in 1.2 (or later). (We investigated keeping the model backwards compatible, but this would have led to a maintenance burden that we could not meet.) The options present for you to get around this depend on what you are trying to accomplish:If you want to use the ocean ecosystem model from CESM 1.0 (or 1.1), you need to stick with that version of the model.If you want to use the ocean ecosystem modem from CESM 1.2, please be aware that the only supported option for initializing the ocean ecosystem in CESM 1.2 is to specify init_ecosys_option=ccsm_startup. (This will happen automatically if you set RUN_TYPE=startup.) We do not have spunup ocean ecosystem initial conditions for CESM 1.2. If you want to initialize a transient simulation, which B_1850-2000_BGC_BPRP is, you will need to setup and run your own control simulation. Please note that there are no coupled scientifically supported configurations in CESM 1.2 that include the ocean ecosystem. This version of the ocean ecosystem model has not been vetted in coupled simulations.Hope this helps, Keith
 
Thank you very much for the advise!The hybrid simulation runs now properly after considering both hints. Besides the init_ecosys_option='ccsm_startup' option, I furthermore had to add an ecosystem initial file: init_ecosys_init_file = '/scratch/daint/blumer/inputdata_cesm1_2//ocn/pop/gx1v6/ic/ecosys_jan_IC_gx1v6_corrO2_Feb2013.nc'.
Best wishes
 
Top