***************************
			 **   THE GEM MESSENGER   **
			 ***************************
						     Volume 5, Number 18
						     September 6, 1995

---------------------------------------------------------
Report on GEM Snowmass Meeting, June 26-30, 1995 - Part I
---------------------------------------------------------

GGCM Campaign 
George Siscoe and Joel Fedder

The GGCM Campaign comprises three Working Groups:
SWG 1: The Core Unit--Dick Wolf, leader
SWG 2: Radiation belts and storms--Mary Hudson, leader
SWG 3: The Tail/Substorm Unit--Michael Hesse, leader

This year's GGCM Campaign (formerly Working Group 5) sessions met on
Wednesday, June 28.  The themes were storms, radiation belts, evolving
the capabilities of a modularized GGCM, and substorms as GGCM entities.
They were devoted to four tasks: 1. establishing GEM's participation
in an inter-disciplinary, inter-agency magnetic storm project; 2. 
establishing a GEM program that identifies and addresses scientific
issues pertaining to radiation belts; 3. continuing the effort to 
increase the capabilities of a prototype, modularized GGCM--as 
represented by the RCM--and its couplings to the solar wind and the
tail; and 4 continuing the effort to treat the substorm as an element
in a GGCM.  The storm-radiation belt theme is a new subject of Working 
Group 2.  Mary Hudson took over the leadership of WG 2 from Bill Lotko,
who led the group during the boundary-layer phase of GEM activities.

To give an overview of the Wednesday's proceedings, we start with a 
review of the agenda items: 
Opening remarks...G. Siscoe
Tutorial--Outer zone electrons: Observations...D. Baker
Tutorial--Radiation belts: Theory and modeling...M. Hudson
Sunrise, CEDAR, GEM Storm Study...D. Knipp
Working Group 1, Session 1 (R. Wolf), Topic: Global MHD 
    models; Presentations by  J. Raeder, J. Lyon, and W. White.
Working Group 2, Session 1 (M. Hudson), Topic: Magnetic storm
    modeling and the interdisciplinary storm study; Presentations by
    J. Lyon, and J. Kozyra.  Session 2 (M. Hudson and 
    D. Baker), Topic: Radiation belts; Presentations by B. Blake, 
    T. Clayton, J. Wygant, and J. Alpert.
Working Group 3, Session 1 (M. Hesse); Topics: Substorm 
    mechanisms and substorm model testing; Presentations by 
    K. Quest and G. Siscoe.
Working Groups 1 & 3, Joint session (R. Wolf and M. Hesse), Topics: 
    magnetic field modeling and uniting magnetosphere and tail 
    models; Presentations by K. Tsyganenko, F. Toffoletto, and F. Cheng.
GGCM wrap up plenary (J. Fedder)

George Siscoe opened the proceedings by reminding the audience of the
goals of the GGCM Campaign and of its progress toward those goals, 
especially its division of the goals into applications for scientific
purposes and applications for operational purposes.  He noted the 
change in subject and leadership of Working Group 2.  He called 
attention to the introduction this year of a joint GEM-GGS project,
the aim of which is to strengthen and advance both programs through
joint campaigns that unite common interests.

Dan Baker's tutorial on outer zone electron observations stressed the
significant increases in our knowledge of outer belt phenomenology made 
possible through the data bases provided by the CRRES and SAMPEX missions.
The outer belt fluxes of electrons with energies within a half order of
magnitude of 1 MeV are strongly correlated with Kp and the solar wind 
speed and, through these, to coronal holes, the 27 day solar rotation,
and the 11 year solar cycle.  The fluxes are as dynamic as Kp and solar
wind speed.  Seen on a yearly time scale, they increase and decrease 
more or less simultaneously throughout the belt, though an inward 
diffusion across the slot region into the inner belt is clearly 
discernible.  In contrasting this behavior with the standard, static
models of electron fluxes, e.g., AE-8, Baker remarked that `models 
must be much more dynamic than they presently are to provide a 
reasonable description of what actually happens.'  He noted that the
spectra are very hard, more like Jovian electrons than solar flare 
electrons.  On short time scales, like minutes, the fluxes show 
intense isotropic spikes which, being isotropic, dump massive fluxes
of penetrating electrons into the atmosphere.  Baker suggested that
substorms inject copious amounts of 10 to 100 KeV electrons into about
L = 5, which act as a seed population for the MeV outer zone electrons. 
The acceleration process might entail a combination of radial diffusion 
and recycling, e.g., a la Nishida's recycling mechanism.  Whatever 
the mechanism, it operates fast; the belts can be repopulated within
one day.  Baker concluded by noting the possible transportability of
this knowledge to astrophysics.

Mary Hudson's tutorial on modeling the inner radiation belt took as 
its prime example the storm of March 24, 1991, in which in minutes 
a second inner radiation belt appeared in CRRES data.  She and her 
Dartmouth colleagues have modeled the event so well, the discoverer
of the new belt, Bernie Blake, mistook their results for his data.  
In their model, the new belt happened when a seed population of 
medium energy ions and electrons in the outer magnetosphere drifted
rapidly inward riding the strong electric field of a shock wave that
transited the magnetosphere.  Only a fraction of the seed population
had the right of mix of starting conditions to ride the wave 
coherently, that is, to surf; still that fraction sufficed to make 
the new belt.  The model owes its remarkable verisimilitude in part 
to John Lyon's numerical computation of the electric wave.  For 
this, he used a global MHD model to simulate the interaction between
a heliospheric shock wave and the magnetosphere.  The detailed, 
quantitative success of the model implies that the operative physics
for 2nd belt creation has been identified, and that the means for
calculating second belt parameters are available.  The potential for
using this capability for operational applications did not go united.

Delores Knipp introduced the GEM audience to the storm she is proposing
to be the focus of an interdisciplinary study--sun, solar wind, 
magnetosphere, ionosphere, thermosphere--involving segments of the
CEDAR, GEM, and solar/heliosphere communities.  The storm commenced
on November 3, 1993 and covers 11 days of activity.  It was selected
after considering the merits of many candidates and choosing the one
with the best balance of data serving all disciplines including an
optimum configuration of satellite and ground stations.  It is a
certified `major' storm, with Dst peaking at -110.  Though it is one
of a series of recurring storms, it seems to be associated with a CME.
The solar wind/IMF data are spotty, especially during the recovery 
phase, but in general the wealth of data is remarkable.  Some examples:
IMP 8 solar wind and IMF data (albeit spotty); Mauna Loa and Yohkoh
solar images; radar data from Millstone Hill and Superdarn; DMSP F11 in
the northern hemisphere and F8 in the southern hemisphere; Freja
magnetometer data; ground magnetometer data (e.g., IMAGE, MACCS, CANOPUS,
and the Greenland chain); F region polar patches and sun-aligned arcs;
Geotail particle and field data; and much, much more.  A full account 
describing the event and listing who is contributing what data may be 
found on a Space Weather Event Home Page assembled by Chris Russell: 
http://www-ssc.igpp.ucla.edu/gem/event_nov93.html.  

Dick Wolf opened the first session of Working Group 1 (Core Models) by 
stating the Group's objective is `to provide the basic computational 
framework for an accurate, comprehensive, predictive of global 
magnetospheric dynamics.'  He noted that the group has `not quite 
achieved' its objective.  `Nevertheless, the business of global 
magnetospheric modeling has changed dramatically in the last two 
years--partly because of GEM.'  Examples: 1. no longer are the 
same groups doing the same thing year after year, as seen by MHD 
codes now attacking the inner magnetosphere;  2. global magnetospheric 
models are addressing regional differences in a constructive, 
collaborative way; and 3. they are making real progress toward a 
comprehensive model.  

Jim Raeder discussed results from some new applications of his 
parallelized global MHD code: He used the code's magnetic and electric
fields to trace particles in connection with a storm simulation; he 
showed that northward IMF turnings trigger substorms but that increasing
ionospheric conductivity suppresses them; and he obtained a good fit 
between a model run and Geotail data taken at 200 Re downtail.  

Bill White gave the first `public' showing of a hybrid MHD-RCM code 
that is being developed for space weather applications, sponsored by 
the DNA.  Besides a major integration effort to merge MHD and RCM 
physics, the code features a seamless transition from the magnetosphere
to the ionosphere (and, in principle, through the ionosphere) using
two fluid (plasma and neutrals) equations.  The MHD-RCM merger is 
still in progress.  White showed results of a model run for a 
northward IMF using only the MHD part of the code.  

John Lyon presented his MHD simulation of an interplanetary shock 
wave impacting the magnetosphere, showing compression and rebound of 
the magnetopause.  This is the simulation that let Hudson and coworkers 
simulate details of the March 91 creation of a second inner radiation 
belt (see next).

Working Group 2 Report (Radiation Belts and Magnetic Storms) 
(From Mary Hudson)
Storm events, radiation belt and ring current particle effects: 
John Lyon presented details from his simulation of the March 91 
storm event.  Electron and proton test particle populations have now 
been pushed with output from the MHD code and show formation of new
radiation belts as reported by Mary Hudson in the overview talk.
The inner boundary of the MHD code has been moved into L=1.8 since
last reported at the June 94 GEM meeting, and necessary for the 
radiation belt modeling. Both the formation of a second proton 
belt and extension of the new electron belt to lower L values than
protons are consistent with CRRES observations, as are ULF 
oscillations in the 1-2 minute period range, also seen by the 
Japanese 210 degree magnetometer chain (Yumoto, private communication).

Anthony Chan reported on results from the MSFM model for the 
November 3-4, 1993 event described by Delores Knipp in her overview 
talk.  The MSFM model is available and continuously updated over the
World Wide Web using Kp as input.  For this same event, two data sets
were presented by Tom Cayton, 1) from the LANL geosynchronous 
satellite energetic particle measurements and 2) from four GPS 
satellites making energetic particle measurements in a 12 hour orbit
at 4.2 Re with 55 degree inclination.  Xinlin Li finished the morning
with a discussion of Sampex energetic electron measurements available
for the November 93 event.

Janet Kozyra led off the afternoon with a comparison of ring current 
evolution for strong and moderate magnetic storms, suggesting that 
loss to the magnetopause may be more important for strong storms such
as the March 91 event, when the magnetopause was displaced inside
geosynchronous.  Data from the November 93 event was presented by Joe
Borovsky and Geoff Reeves, again from the LANL geosynchronous data set.
Borovsky addressed the MPA instrument data, 1 eV - 4 keV ions and
electrons, available on WWW at http://sst.lanl.gov.  Excursions into 
and out of the boundary layer are evident for this event.  Reeves
described the availability of pre-89 CPA (>30 keV) and post -89 SOPA
(>50 keV) data for up to tens of MeV particles which provide a long
time period data base at geosynchronous.  SOPA data, described earlier
by Tom Cayton for the November 1993 event, shows a rise in particle
fluxes at geosynchronous delayed from the SSC by days, very different 
from the order of magnitude increase in fluxes of MeV protons in 
less than a drift time scale, for example, seen in the March 91 event.

John Wygant showed CRRES electric field data, with quasistatic 5 mV/m 
fields penetrating inside L=4.  Substorm fields can be as large as 
20-60 mV/m for minutes and 100 mV/m for shorter time periods, providing
transport by radial diffusion into L=4.  Radial diffusion coefficients
larger by a factor of 10^3 than shown in Schulz and Lanzerotti's 
monograph follow from these electric field measurements. The electric 
field measured by CRRES during the March 91 event at L=2.5 on the 
nightside implies even larger, e.g. 200-300 mV/m, dayside fields for 
this transient event, which produce the nondiffusive acceleration 
described by Hudson in her overview talk.

Jay Albert discussed the inadequacy of the AE8 empirical radiation belt
model as well as the product of radial diffusion calculations when 
compared with CRRES energetic particle measurements both before and
after the March 91 event.  The suggestion was made during the 
discussion that nondiffusive processes may be important, and that 
average empirical models may not adequately describe either very quiet
or very active periods.

Harlan Spence briefly described the availability of recently 
declassified MCP data from 63.5 degree inclination orbits which cut
through the radiation belts providing radial information not available
from the geosynchronous spacecraft.

In the first session of Working Group 3 (substorms as an element in 
GGCMs) Michael Hesse introduced a talk by Kevin Quest on substorm onset
mechanisms.  Quest reviewed the tortuous history of attempts to make 
collisionless tearing work as a reconnection onset mechanism.  At 
present, it still does not work, despite the overwhelming observational
evidence supporting reconnection in the tail (though not necessarily
at onset).  Next he considered crossfield current instabilities, as
invoked in the current disruption scenario.  Here candidates for an
operative micro-instability are lower hybrid and Weibel instabilities, 
the latter of which he thought might be equivalent to tearing.  The 
theory in this case is not as developed as in collisionless tearing, 
partly because it does not treat the self-consistent co-evolution of 
the macrophysical parameters on which it depends.  Another type of 
onset mechanism is external forcing, or driven reconnection.  The 
models here are inconclusive because of uncertainties related to the 
effect of realistic mass ratios and boundary conditions.  Finally he 
looked at the possibility of suddenly destabilizing a firehose 
equilibrium.  The problem is demonstrating the existence of the 
equilibrium and identifying the destabilzing mechanism.  He concluded 
that the onset mechanism, which has eluded capture for 30 years, 
remains at large.

The theme of slow progress in substorm research was voiced as a 
concern by George Siscoe.  He contrasted substorm research with 
research in other fields that also got their conceptual starts in the 
early 1960's: the genetic code; the standard model of elementary 
particles; the environmental movement; sea floor spreading and 
continental drift; deterministic chaos; lasers; quasars; and the 3 
degree cosmic background radiation.  The point being that other 
fields have had greater impact, though they may be no more interesting
scientifically. Unlike these other cases, the substorm concept 
has not resulted in a new, powerful scientific tool, which, in this 
case, could be transported to solar physics, planetary physics, or 
astrophysics, or that could be encoded as part of a physics-based 
space weather algorithm.  Nor, without having captured for good 
the substorm mechanism, can the field give a satisfactorily complete, 
coherent description of its own central paradigm, magnetospheric 
convection.  Siscoe thought that a way to break the stalemate might
be to put more emphasis on model testing.  For this he solicited 
ideas of tests from the audience.  The following ideas were offered:
triggering (test model's triggering ability against statistics); 
convection bays (model must be consistent with their existence); 
wave characteristics (the current disruption model needs waves with
certain characteristics; are they there?); MHD tests (non-reconnection
onset models initiate reconnection with a rarefaction wave propagating
outward after onset--can MHD models simulate rarefaction initiated
reconnection?); pseudobreakups (model must be able to explain these); 
width of the current wedge (we have some data on this--what do the 
models predict?); pre-breakup electron heat flux (if reconnection is
the onset mechanism, it should start before auroral breakup; thus
reconnection heat flux carried by electrons should be a precursor). 
Siscoe concluded by encouraging individuals to take up testing as a
research topic.

Harlan Spence reported on a new source of data which can be of use
in substorm studies.  The source is satellites in Molnya orbits, 
highly elliptical with periods at the 2nd, 3rd or 4th harmonics of 
Earth's spin period, meant to give long hovering times over specific 
geographical points, usually at high latitudes which are unfavorable 
for servicing by geosynchronous satellites (e.g., Moscow).  Some of 
these satellites are well instrumented for gathering particle and field 
data, and some of the data are becoming available.  Spence showed 
an example in which a substorm onset signature was seen at a 
geosynchronous satellite considerably before it was seen at a Molnya 
orbit satellite in the same local time sector but at a higher L shell.  
This kind of comparison can help locate the site of substorm onset.

Howard Singer gave a brief progress report on a GEM-inspired 
electrojet specification project being carried out by D. Vasiliadis, N. 
Maynard, D. Baker and himself.  It is based on a blending of the 
forecasting power of the input-state space technique pioneered in 
the magnetospheric context by Vasiliadis and Baker and the spacial 
specificity of the Heppner-Maynard convection patterns.  It is being 
developed for potential operational deployment at SEL.  There will 
be a more substantial progress report on it at the joint SEL/GEM-GGCM 
Campaign workshop in January 1996.

The second sessions of Working Groups 1 and 3 met together.  Kolya 
Tsyganenko gave an update on the project he heads at Goddard to 
develop ever more powerful and flexible semi-empirical magnetic field 
models.  New features include a fully shielded magnetopause; 
continuous parametric variation of dynamic pressure, Dst, AE, and 
dipole tilt.  He showed field line maps that illustrated the effect 
of varying one parameter while holding the others fixed.  One sees
field line stretching in the tail as AE goes up.  On the other hand,
one sees little change around the magnetopause as Dst goes down.  
Future improvements will include the fields of parallel currents and
IMF effects.  The Goddard models can be accessed by e-mail at 
ys2nt@epvax.gsfc.nasa.gov.

Frank Toffoletto presented a solution to the basic problem of finding
equilibrium solutions of particles and fields in the magnetosphere.
He extended to the whole magnetosphere the magneto-frictional relaxation
technique that Hesse and Birn applied to the tail.  Toffoletto started 
with the Tsyganenko 89 magnetic field model and an informed guess as
to the particle pressure distribution that is in equilibrium with it, 
but that inevitably is not.  Then he brought the field and pressure 
into equilibrium by evolving them with the MHD equations modified by 
the inclusion of a magneto-frictional relaxation term.  He found that 
a good first guess is crucial for ultimate convergence, and that even 
so, it takes 1000 interactions to get the departure from equilibrium 
under 2%.  Though immensely important to the project of quantitative 
magnetospheric modeling, the change that the approach to equilibrium 
brings about is not visually startling.  The pressure goes from axial 
symmetry--the initial guess--to being reduced on the dayside and 
increased on the nightside.  The field changes very little.  The 
demonstrated success of the technique means that there is a way to 
merge the outputs of MHD and RCM runs on a step-by-step basis.  This 
is the intended application.

Frank Cheng also treated the equilibrium quasi-static magnetosphere 
problem.  He presented the first analytical solution to the three 
dimensional problem.  He had worked out the two-dimensional, axially 
symmetric case earlier, employing the advanced iterative metric 
technique.  In this case he introduces three dimensionality by making
the outer bounding flux surface conform to the general shape of the 
magnetopause with dayside compression and nightside stretching.  
The qualitative new feature that results is a toroidal component to 
the magnetic field, which gives rise to field aligned currents in the
region 2 sense.  The code he developed is fast, converging in about 50
iterations, and it can be generalized to open field lines and 
anisotropic pressures.

Dick Wolf offers these concluding comments:  `Dynamic global 
magnetospheric modeling is making some really rapid progress, 
particularly coupling MHD codes to the inner magnetosphere.  No, 
the large-scale modelers haven't solved the substorm problem yet.  
However, really exciting progress can be expected in the next couple of
years.  On a longer time scale, advanced techniques--e.g., adaptive 
grids, hybrid plus full particle codes--will provide increasing insights
into global dynamics.'

The winter meeting of the GGCM Campaign will be held January 11 and 
12, 1995, at the Space Environment Laboratory.  An agenda will be 
circulated soon.


+-------------------------------------------------------------------------+
|To add name to the mailing list, send a message to: editor@igpp.ucla.edu |
|For message to whole GEM mailing list, send to:   gem@igpp.ucla.edu      |
|For message to a specific working group (BL Campaign), send to:          |
| gem_field@igpp.ucla.edu (WG1);    gem_boundary@igpp.ucla.edu (WG2);     |
| gem_current@igpp.ucla.edu (WG3);  gem_data@igpp.ucla.edu (WG4)          |
| gem_ggcm@igpp.ucla.edu (WG5)                                            |
| gem_chair@igpp.ucla.edu (WG chairs, Odile and W. J. Hughes)             |
|                                                                         |
|URL of GEM Home Page: http://www-ssc.igpp.ucla.edu/gem/Welcome.html              |
|Please update your e-mail address.                                       |
|CAUTION: Do not send messages to gem@igpp.ucla.edu unless you want       |
|         your message to go to everyone in the GEM mailing list!         |
+-------------------------------------------------------------------------+