---------------------------------------------
Report on the Dec. 5, 1993 GEM Mini-Workshops 
---------------------------------------------

                 Working Group 5:  GGCM                                      
                      George Siscoe

The pre-AGU WG 5 session began with a brief report on the 2-day October 
meeting of the working group held at Boston College.  The WG 5 report 
on the Snowmass Workshop in an earlier GEM Messenger described the 
substance of that meeting.  It clarified the WG's near-term goal and 
delegated tasks to reach it.  Its near-term goal (i.e., 2 to 3 year time 
scale) is to construct a prototype of a modularized GGCM.  'Prototype' 
means a GGCM that has scientific applications (e.g., model testing) and 
commercial applications (e.g., predictions).  The WG divided the GGCM 
into three subunits: a core unit, a magnetopause-boundary-layer unit, 
and a substorm-and-tail unit.  The core unit comprises a magnetosheath 
module, a polar-cap-electric-field-mapping module, an inner-magnetosphere-
convection module, and an ionosphere-specification module.  Highly 
serviceable versions of these modules already exist.  Combining them is a 
feasible "two-year" task producing a valuable prediction-quality GGCM, 
satisfying the requirement of achieving at least advanced development 
leading to commercial applications.  The magnetopause-boundary-layer and 
substorm-and-tail units are intended to test boundary layer and substorm 
models.  The idea is to adapt existing alternative models for integration 
into the core unit.  Runs of the thus-augmented core unit can be compared 
with observations.  This tests the models more thoroughly than possible 
for models in isolation, satisfying the requirement of achieving 
scientific applications.  

Carrying out the Boston College agenda is explicitly a matter of solving 
interface problems.  The task of marshaling community effort to develop 
interfaceable codes for the various models and for defining interface 
variables and standards was assigned to Dick Wolf (core unit), Bill Lotko 
(boundary unit), and Michael Hesse (Goddard Space Flight Center) (substorm 
unit).  The subgroup leaders were invited to describe their plans at the 
December 5 meeting.  Basically, the elements of the subunits were described 
and some progress toward defining how to treat the interface problems was 
reported.  Michael Hesse has received positive responses from several of 
the key substorm modelers for the idea of developing versions of their 
models that can be interfaced with the core unit.  The idea that the 
boundary unit and substorm unit might be treated as subroutines of the 
core program appeared to be a promising approach.  

The WG heard invited presentations from David Stern and Kolya Tsyganenko 
on the modeling program at Goddard.  The Goddard program emphasizes magnetic 
field models.  It uses a variety of theoretical and empirical approaches.  
This gives it flexibility in being able to tailor models to particular 
applications, such as Tom Sotirelis demonstrated at the Snowmass Workshop.  
After many years of dedicated effort, it has reached a state of high 
maturity and sophistication.  It has generated a veritable library of field 
models, many of which, or derivatives of them, are used by different groups 
in the world.  Stern demonstrated the power of some of the latest products.  
They are now able to simulate many of the features of the regions 1 and 2 
currents.  Tsyganenko described the improvements he is incorporating in a 
new version of his semi-empirical model, which has achieved the unofficial 
status of the standard field model for the discipline.  It was agreed that 
the Goddard and the GEM modeling efforts are complementary, and that more 
interactions between the programs should take place.  Robert Winglee 
requested time to describe the program he is conducting at the University 
of Washington on global magnetospheric modeling using full particle 
simulations.  Global modeling is traditionally the prerogative of MHD, but 
Winglee demonstrated that full particle simulations can reproduce global 
features and can reveal some surprises.  The demonstration made clear that 
full particle global simulation is an approach to magnetospheric modeling 
whose development can yield valuable, perhaps unique, additions to our 
knowledge of magnetospheric physics and behavior.

Returning to the near-term goal defined at the Boston College workshop, 
the WG listed its set of GEM Milestones:
  1. Module Specification
	(Magnetosheath Module, Electric Field Mapping Module, Augmented RCM 
	Module, and AMIE; These are stand-alone modules that integrate into 
	the core program)
  2. Interface Standards Defined
  3. Interfacing Modules of the Core Unit
  4. Interfacing Boundary Layer Modules to the Core Unit
  5. Interfacing Substorm Modules to the Core Unit.
Each of these items is important enough to label as a GEM Milestone 
with full press coverage in Eos and in some cases in Science.

Items that still need to be defined were also listed:
  1. Integration of WG Efforts
  2. Communication between WG 5 Subgroups and Pacing of their Efforts
  3. Marketing and Public Relations for the GEM Program
These items need to be addressed at the next Snowmass Workshop.

As agenda items for the next Snowmass Workshop, the WG listed the following:
  1. Lecture Topic: Space Weather: The Big Picture
  2: Reports from Subworking Group Leaders
  3. Reports on Other Projects
  4. Round Table Discussion: Relation of Space Physics to a National Space 
     Weather Service (NSWS).  Needs of a NSWS in terms of Model Outputs
  5. Subgroup Working Sessions