[cam-users] Release of CAM3.1

[Phil Rasch]

Dear CAM users

This email is to announce the release of CAM-3.1.

YOU DO NOT NEED TO UPGRADE UNLESS YOU WANT:

1) to do model development
2) to exploit some of the new functionality mentioned in the following
   paragraphs.

This release is primarily a software engineering release. It reproduces the
CAM-3.0 climate in configurations for which control simulations were
provided. In those configurations, each code change between versions
3.0 and 3.1 either produces identical results, or introduces roundoff
level changes as validated by the perturbation growth test. It is
important to note however that CAM-3.1 will not pass the perturbation
growth test compared to CAM-3.0. That is because perturbation growth
tests are done using the aqua-planet configuration, and there were
non-roundoff level changes in the aqua-planet configuration.

The main reason for the release is to provide parameterization
developers with code that contains updated physics interfaces in the
context of a model with a validated climate. The 
physics interface is important for a few reasons:

1) it will allow developers to add new parameterizations, or replace
   current parameterization much more easily than in previous versions of
   the model.

2) it will isolate parameterzations much more strongly from other
   components of the model. This means that people working on a particular
   parameterization should be able to stay abreast of future CAM evolution
   much more easily.

There are also a number of features that provide additional science
capabilities, although we stress that the features do not change the
model climate. They only allow you to do more things with the model.

Among the new new Science Features are:

 
. Add flexibility to where state is written to initial files.

. Allow radiation calculations to be done every timestep for
  fixed number of steps at startup.

. Add column radiation model (CRM) and improvements to the single
  column model (SCAM).

. Allow CAM to use 30 second topographic variance field.

. Add FV offline driver.

. Add diagnostic fields required for energy and water budgets.


There are a few extra features on the software engineering side also:

. SPMD optimizations and extensions for CRAY/NEC.

. Add compiler directives for Cray X1 optimization.

. Changes to support IA64 cluster.

 
All the bug fixes from the CAM-3.0.p1 release are included in this
release. In addition the following problems were resolved:  

. Removed equivalence statements.  This was causing problems for the intel
  compiler (really a compiler bug, but equivalence should be removed anyway). 

. Remove addfld call for TAUVIS which is no longer used in model.

. Remove date checking in history.F90 which was only valid for 365 day
  calendar.

. Write correct value of calendar attribute to output netCDF files.

. Fixes to allow FV w/ 2D decomp to run under lf95/debug.

. Fix bug in date comparison in prescribed_aerosols.F90 (provided by Jim
  Boyle). 

 

Build Changes

 
. Default optimization for pgf90 was reduced from -O2 to -O.


-- 
Phil Rasch, Climate Modeling Section, National Center for Atmospheric Research
Mail     --> P.O. Box 3000, Boulder CO 80307  
Shipping --> 1850 Table Mesa Dr, Boulder, CO 80305
email: pjr at ucar.edu, Web: http://www.cgd.ucar.edu/cms/pjr Phone: 303-497-1368, FAX: 303-497-1324