search

CABLE-LSM / CABLE

Home to the CABLE land surface model and its documentation
https://cable.readthedocs.io/en/latest/
Other
8 stars 3 forks source link

Reformat BIOS met forcing input to netcdf #322

Open ccarouge opened 1 week ago

ccarouge commented 1 week ago

We want a version of the BIOS configuration that will use the met forcing in netcdf format instead of the current binary format.

This will be worked on from the CABLE-POP_TRENDY branch, using the configuration act9test-serial-no-luc.

ccarouge commented 1 week ago

Outdated (To start working on this, @AlisonBennett needs to copy the met forcing data for 1900-1902 in rp23.)

See update below.

SeanBryan51 commented 1 week ago

Reply to @har917

Sean,

Can I check what you need from us regarding BIOS meteorology files you would like access to? I got a bit confused as to where we landed yesterday.

At the moment I/we could easily

  • copy some of the flt/hdr files from our archive across to rp23 (this is an rsync activity) – which could then be converted to netcdf. Which years would you like?
  • copy our archive of 1985-2014 in netcdf format across to rp23

Both are on the same 0.05 degree grid.

Personally I suspect that (for the purposes of testing) you could use the 1985-2014 netcdf and change the cable_user%startyear and %endyear accordingly – instead of 1900-1902. It is slightly less hassle to copy the netcdf archive than the flt/hdr – but happy to go with what you need. The netcdfs are larger if that’s a concern.

cheers, Ian

I don't see any issues with copying the archive of 1985-2014 in netcdf format across to rp23. Even better that it is available in NetCDF format.

@Whyborn has volunteered to look into the BIOS forcing - do you have anything to add?

Whyborn commented 1 week ago

I think we can start with just the NetCDF archive of 1985-2014 and check that it's in a format compatible with the Met forcing routines associated with POP_TRENDY.

har917 commented 1 week ago

Post-processed AGCD (plus McVicar wind) data are now in the BIOS3_forcing/netcdf/ directory in the appropriate directory in rp23.

I would not be surprised if there needs to be some tweaks to the metadata, dimensions, variable names, time axis to make these consistent with CRUJRA/TRENDY. Note that BIOS also runs with a slightly different set of met forcing to TRENDY.