Mission Statement:

Provide support to create a vector friendly version of WACCM to run on the
Earth Simulator.

Goal:

WACCM vectorizes well enough to be run on the Earth Simulator.
  
Motivation:

Access to Earth Simulator resources.

Authority: 
   Boville

Project Lead:
   None

Scientific Lead: 
   None

Implementation Plan (from Boville):

We are just beginning to test vectorization in waccm. (We mostly means
CRIEPI people). What we know now is
1. the fv dycore needs some work (Art Mirin has started on this for the X1)
     1-2 pm
2. the cam physics is in decent shape (as far as I know)
3. The waccm physics will probably need some work, but not much
     <1 pm
4. the waccm chemistry, as comitted now, will probably not vectorize
particularly well but Stacy Walters can generate a vector version fairly
easily. A vector version is running well on the SX6s in Hamburg.
     1 pm

What we need is some time from Brian Eaton and/or George Carr and access
to a vector machine (X1 and SX6) to test the existing code and assess
where work is needed. We can get some time from NEC people (Parks and
Snyder) for porting to the ES, including vectorization. It would really
help to do our own assessment of the bottlenecks though. Parks and Snyder
should be able to arrange SX6 access for us.

After that we need to divide up the tasks and coordinate the resulting
work getting brought back into our repository. I think that 3-4 person
months should be plenty. There is some urgency, the time on the ES is
available now.