|
This is a summary of what was agreed to put in POP 1.4 for
the first CCSM coupled run at the Ocean Model Working Group meeting held on January 22:
- Variable-thickness surface layer - freshwater flux.
The issue still to be decided is whether to have this the
default or an option. This feature will affect freshwater coupling with
all other component models. In the initial CCSM run melting and
freezing ice as well as river runoff will be assumed fresh, so all salt
will remain in the ocean and be conserved there.
- New anisotropic viscosity routine.
Done - we still have to decide on parameters and the form
of viscous coefficients, but all the needed options are in the
code.
- New GM routine with capability for different
coefficients and tapering for bolus and diffusion terms.
We (Wooyoung Choi, Phil Jones, and Rick Smith) will start with the code from
Gokhan Danabasoglu to try and optimize it. One thing we will try to take
advantage of in optimizing the performance is that the different coefficients and
tapering will only be needed in a few places, so the skew-flux
cancellation will still take place almost everywhere.
- New time manager.
Phil Jones has this and will install it. One issue is that
the restart header files will contain different information, so Nancy Norton needs
to write some conversion routines to make it possible to restart
from previous restart dumps from the old code.
- netCDF output capability.
For POP 1.4 this will just have the routines that Nancy Norton has
put together which - I think - produce netcdf time-average and history
files. By summer Phil Jones will have more general routines that will go
into POP 2.0.
- Various minor bug fixes.
Done.
- Jackett and McDougal EOS.
Done - but how well has this been tested? Will we use
it in the CCSM run? Will it significantly effect performance?
- Marginal sea balancing.
What is the status of this? Can we get it done and into POP
1.4 in the next 2 weeks?
|
