search

COSIMA / cice5

Clone of The Los Alamos sea ice model (CICE) with ACCESS drivers. See https://github.com/CICE-Consortium/CICE-svn-trunk/tree/cice-5.1.2
4 stars 13 forks source link

Wombat #47

Closed russfiedler closed 4 years ago

russfiedler commented 4 years ago

Added 10m wind forcing for use with WOMBAT (ACCESS-OM-BGC) and enable enhanced langmuir mixing in OM2

aekiss commented 4 years ago

@aidanheerdegen, @nichannah just wondering if you're happy with this PR

aidanheerdegen commented 4 years ago

Haven't look TBH. Will take a gander, though not sure I have any useful domain/code knowledge.

aekiss commented 4 years ago

great, thanks @aidanheerdegen. @nichannah if you could look this over too that would be good.

aidanheerdegen commented 4 years ago

Seems ok to me.

aekiss commented 4 years ago

Merged - @nichannah was also happy with it, but curious about what the change in drivers/auscom/cpl_netcdf_setup.F90 does - @russfiedler can you elucidate?

russfiedler commented 4 years ago

@aekiss Good point, I'm not sure who first introduced that piece of code nor why it was dropped. I remember seeing it somewhere and added it. It might be in the ACCESS-CM code. Anyway, what it does is provide space in the header of the file for adding new variables and attributes and aligns data. TBH I think a better call would be to name the arguments to make things clearer

call ncheck(nf__enddef(ncid,h_minfree=1028,v_align=512,v_minfree=512,r_align=512), 'create_ncfile: nf__enddef')

The h_minfree option seems pretty necessary as there looks to be quite a few additions of variables along with writing fields. This should avoid copying the file unnecessarily. I use it quite a bit. I'm not sure on the performance gains/hits with the other options.

aekiss commented 4 years ago

ok thanks Russ