ggslc / bisicles-uob Goto Github PK

UKESM has GUIs that aid configuration. BISICLES namelists need user-facing metadata to explain what each option does. Current attempt needs improving.

UKESM simulations run alongside systems that can monitor output and produce overview plots. Ice sheet outputs from BISICLES need to be compatible with these systems as well.

Likely involves some scripts (python?) that fit with the monitoring system and know how to read plot*.hdf5 files and return the required fields/quantities.

running gmake all from the $BISICLES_HOME/bisicles-camip/examples/Greenland directory produces the following error message:

GNUmakefile:17: *** target file `clean' has both : and :: entries. Stop.

removing lines 17 - 21 (the clean and realclean options) from GNUmakefile bypasses this error but produces the following error:

Fatal Error: Cannot open module file ‘netcdf.mod’ for reading at (1): No such file or directory
compilation terminated.

Basal melt function definitions (e.g. in MISMIP+ example) using the Python interface (in shortshallow.py) only provide 5 arguments, whereas the interface now expects 6 arguments to be provided (including the ice fraction).

A temporary fix is to add *etc as a 6th argument to the basal melt function definitions but will need to update these examples eventually.

BISICLES interacts with two components of the UKESM1 models.

1. Atmosphere / Land surface: supplies surface elevation s(x,y), receives surface mass balance a(x,y,z=s) and heat flux or temperature.
2. Ocean: supplies ice shelf draft d(x,y), receives melt rates m(x,y,z=d)

At the moment, 1 requires BISICLES wrapped into a program called Unicicles, which lives in a separate svn repo and includes a legacy version of the Glimmer-CISM fortran ice sheet model. Legacy glimmer-CISM does not compile under recent gfortran. It was always difficult to set up the entire stack. We use only a small part (GLINT) in practice. We might replace this with new code and remove the need for unicicles.

Projects in which BISICLES is involved need outputs in netCDF files on a user-friendly grid and with comprehensive metadata.

This can be done in a post-processing step, but we need a system or scripts that any user can apply easily to format output as required.

ESM projects that BISICLES is a part of require time-mean outputs (as ESMs provide as standard), rather than output snapshots at specific time points.

Need to develop these capabilities within the model output options, rather than just averaging between snapshot plot*.hdf5 files.