***************************
			 **   THE GEM MESSENGER   **
			 ***************************
						     Volume 12, Number 31
						     August 20, 2002

-----------------------------------------------------
GEM 2002 SUMMER WORKSHOP REPORT FROM THE IMS CAMPAIGN
-----------------------------------------------------
From: Anthony Chan <aac at rice.edu>

The Inner Magnetosphere-Storms (IMS) Campaign held a very active three day
program at the 2002 GEM Summer Wokshop. Two tutorials, a poster session, and
14 breakout sessions were held, including two sessions joint with the GGCM
and MI-Coupling campaigns on "The Magnetosphere and Ionosphere under Extreme
conditions: Points of Contact between Global MHD Simulations, IMS Modeling,
and Data". In addition, the GEM Student-Sponsored Tutorial "Radiation Belt
Ups and Downs: Acceleration, Transport, and Loss of Relativistic Electrons"
was given by Geoff Reeves of Los Alamos National Laboratory.

The joint GGCM-IMS-MIC sessions have been reviewed by George Siscoe (see The
GEM Messenger, Volume 12, Number 24, July 29, 2002). The remaining IMS
activities are summarized below.

Anthony Chan (aac at rice.edu), IMS Campaign Convener

****************************************************************************

IMS WG1 (PLASMASPHERE AND RING CURRENT) REPORT

WG1 Co-chairs: Mike Liemohn (liemohn at engin.umich.edu)
               Dennis Gallagher (Dennis.Gallagher at msfc.nasa.gov)

The Inner Magnetosphere Storms Working Group 1 (IMS/WG1) held six
independent and joint sessions during the June 2002 GEM, as well as an
invited tutorial presentation.  The sessions focused on new work regarding
inner magnetospheric electric and magnetic fields, on the new techniques
that have become available for deriving the distributions of thermal plasma
and results from their application, and on the observations and modeling of
ionospheric, plasmaspheric, and ring current interactions in the inner
magnetosphere.  The tutorial, given by Prof. Dick Wolf of Rice University,
gave a lucid description of why inner magnetospheric dynamics are important
to the global stormtime scenario: low and midlatitude ionospheric effects,
primarily driven by the asymmetric ring current, are presently one of the
biggest space weather concerns.  Changes in upper atmospheric flow patterns
and total electron content impact GPS and communication systems, resulting
in adverse societal impacts.  Prof. Wolf detailed this connection and the
history of our understanding of it.


Inner Magnetospheric Electric and Magnetic Fields

Presentations on inner magnetospheric magnetic and electric fields discussed
newly observed properties and modeling.  Several empirical/physical B-field
models exist for the inner magnetosphere.  Some use upstream conditions for
the parameterization, while others use magnetospheric parameters.  These
models are indicating that, for intense storms, most of the inner
magnetospheric B-field perturbation (and also Dst) is from the ring current.
Determinations of partial-versus-symmetric ring current contributions are
also becoming incorporated into these models.  In small storms, the tail
current plays a much larger role.  It was discussed that modular,
event-oriented B-field models are more flexible than statistical models and
may be the best option for the strongest storms.  However, with the
collection of more data, statistical models are becoming better at very
disturbed times.  B-fields from field-aligned currents (FACs) are the hardest
to include because the FACs that flow along B also modify B.  Subauroral FACs
are often carried by soft electrons, so they close in the F region, with
little or no ground magnetometer signature.

The biggest E-fields in the inner magnetosphere occur inside of L=5
(evening/nightside), opposite of the standard picture.  Over shielding
requires an inward penetration of the plasma sheet followed by a northward
turning of the IMF Bz.  There is a 30-minute time delay between changes in
the IMF and the reconfiguration of inner magnetospheric electric fields,
probably because the inner magnetospheric electric fields depend on the build
up and decay of ring current pressure.  The inner magnetospheric convection
electric field extends at least through the early storm recovery phase.
Comparisons between various electric field descriptions shows that the
E-field is quite complicated and the simple field models used up until now
are probably inadequate to truly describe the stormtime features.  Fresh
injections appear common throughout storms, including the recovery phase.
New online DMSP derived products will soon become in support of these studies
e.g. integrated precipitating energy flux, ion drifts, ion composition, and
satellite tracks.  A new SIMEF electric field model is being developed as
spherical harmonic fits to the RCM.  It is intended to be both easy to use
and capable of reproducing physically interesting features, e.g. SAPS.  For
this meeting, MSM was used to explore the range of storm-time magnetotail
depolarization effects.  The analysis found distortions reaching into L=5 at
midnight and found quiet-time geosynchronous field lines stretching to L=25
in the tail during storm-times.


New Density-Deriving Techniques for the Inner Magnetosphere

The past few years have seen the emergence of new techniques which are
capable of delivering much more global descriptions of ionospheric and
plasmaspheric density distributions.  Ground magnetometer observations of ULF
waves are used to measure Eigen modes of field oscillations.  Frequencies are
related to magnetic field strength, field line length, and mass loading.  The
analyses of observed resonances yield field line mass densities.  The
technique assumes a smooth, monotonic variation of Alfven velocity versus
location along an L-shell.  Resonances occurring directly over an observing
station are found to be difficult to interpret.  The technique also assumes a
purely toroidal mode and that sufficient broadband noise exists to excite all
modes and latitudes of interest.  Resonance analysis is simple and easy to
perform and can provide mass density estimates from L<2 to L<12. It can also
be used to provide continuous measurements over most of the dayside with 30
min resolution time resolution.  Magnetoseismology makes use of propagation
times for subsolar origin ULF waves that propagate to the ground.  By
analyzing the time of arrival and wave phase, mass densities can be inferred
over large regions of the magnetosphere.  More extensive placement of ground
stations and improved inversion techniques are needed.  GPS TEC is becoming
mainstream for obtaining total electron densities.  The technique uses
Faraday rotation of trans-ionospheric signals, which depend on the frequency
and line-of-sight integrated electron column density.  Typically 50% if TEC
is below 1000km and the rest from the plasmasphere.  There are about 500 GPS
stations across the north American continent making these measurements for
the last 5-years continuously.  A density map is made every 5 minutes,
derived from over 150 measuring stations. This processing extends out to
L=4. Processing is expected to be automated in the next year.  IMAGE Extreme
Ultraviolet imager (EUV) observations can be used directly to obtain the (L &
MLT) equatorial projected locations of observed edge features, primarily the
plasmapause.  Densities are just beginning to be obtained using a variety of
techniques: forward modeling, genetic algorithm inversion, and arithmetic
reconstruction technique inversion.  These techniques remain to be broadly
applied to EUV observations, although this can now be expected soon.  The
observations are of 30.4nm solar ultraviolet light scattered by He+ in the
inner magnetosphere; presenting an optically thin medium.  Inversions depend
to varying degrees on a priori assumptions and are capable of varying spatial
resolutions in derived density.  IMAGE Radio Plasma Imager (RPI) measurements
directly sample remote densities.  Group index of refraction for X and Z mode
waves (primarily) are used to determine group velocity transit times for
transmitted wave pulses, which are directly measured.  Simultaneous fitting
of multiple-mode echo traces yields unique identification that echoes often
follow field-aligned paths.


Inner Magnetospheric Density Resutls

In situ measurements of plasmaspheric plasma have a long rich history, from
the ground and from space.  As part of a review, the suggestion was made
to use d(Dst)/dt in future statistical correlations.  It was also noted that
few observational missions have measured both ion populations and waves at
the same time and that more of these measurements are needed.  In situ and
remote techniques are yielding new information about field aligned density
distributions.  ULF waves together with in situ and remote IMAGE RPI
observations can be used to explore storm-time mass loading of field lines.
It was also noted that there might be some value in consolidating the
differing mathematical approaches currently being employed to describe
derived plasma density distributions.  Initial interpretation of IMAGE global
EUV images suggest that midnight eroded plasmaspheric plasma may be pushed
inward to lower L-shell, while evening plasmas can be seen to directly
convect toward the dayside magnetopause.  A clear determination of midnight
plasma transport remains to be accomplished.  Field aligned plasmaspheric
densities are often found not to follow diffusive equilibrium.  In one case
this was true even after 3 or more days of quite conditions at L=5.  The
question was posed as to when collisional and collisionless treatments are
appropriate for describing field-aligned distributions and how do
distributions transition between these treatments?  New models for polar cap
thermal densities and plasmaspheric densities are becoming available from the
IMAGE RPI.  Global observations from the IMAGE EUV are showing that
plasmaspheric structures include the convection tail, as previously
anticipated, but also strikingly include many types of azimuthal structures.
Some azimuthal structures have been successfully related to under/over
shielding of the convection electric field.  Some show evidence of
large-scale standing waves, perhaps externally driven by the solar wind and
not locally resonant.  Plasmapause behavior is still not well understood.
Observational techniques are now available which should be capable of
addressing this issue, which include GPS TEC measurements that were used to
find a hole in the plasmasphere/ionosphere over the magnetic equator with
TEC=20.


Inner-System Coupling: Observations

Sub-Auroral Polarization Streams (SAPS) are driven by a polarization electric
field that forms in the low conductivity region between the inner edge of
plasma sheet precipitation and the plasmasphere.  SAPS are found between dusk
and midnight, and have flow speeds of about 1 km/s.  They appear to be driven
by a ring current pressure gradient.  The inner edge of SAPS penetrate the
outer plasmasphere, participating in plasmaspheric convective tail formation.
Meso-scale inward and outward motions of the outer plasmasphere are
interpreted as the result of a westward electric field resulting from over or
under shielding of the externally driven convection electric field.
30-minute delays are found between changes in the IMF Bz and the
reconfiguration of electric fields in the inner magnetosphere.  Remote
plasmasphere, ring current, and auroral observations were presented that
suggest coupled interactions between these regions.  An overlap between the
plasmasphere and ring current appears to result in a significant reduction in
plasmaspheric densities in the overlap region.  This region also appears
connected to a localized equatorward extension of the auroral zone.  Evidence
suggests that these interactions occur during storm and recovery times.
Observations were presented of the build up of the ring current in response
to changes in IMF Bz, suggesting that the 30-minute time delay in the
transmission of convection electric field to the inner magnetosphere results
from the time required for ring current build up.  The implication is that
the inner magnetospheric electric field is purely pressure driven.  Ring
current asymmetry is found to persist well into storm-time recovery, based on
Dst.  It was suggested that IMF driven convection should be examined to
determine whether injection continues beyond early storm times, which would
explain persistent asymmetry.  The convection electric field as represented
by VB-south does not appear to control the rapid initial recovery of Dst.  If
flow-out is the primary loss process in the initial recovery, then the plasma
sheet density is a likely influence.  Examination of the AMPTE/CCE database
reveals that there is a lot of low-energy O+ at low L-shells during storms.


Inter-System Coupling: Modeling

Several new and near-new plasmasphere models were presented.  H+, He+, and O+
are modeled in these physics-based models.  Ionospheric inflow/outflow and
convection are uniformly included in these models.  Simulations of the
storm-time diffuse auroral using AMIE electric field were compared to PIXIE
and UVI.  Flux dependent diffusion is found to be the most realistic
regarding precipitating flux and local time variation.  It has been possible
to account for some, but not all, features.  A next step is to include a more
realistic time dependent magnetic field configuration.  RCM modeling results
were presented, which capture most of SAPS events: double peak structure of
SAPS electric field, MLT extent, and variation of SAPS location with MLT.
Modeling does not include changes in ionospheric conductance in the midnight
sector.  Modeling also does not have changes in charge exchange in strong
flow regions.  Future work requires coupling an ionospheric model with RCM.
Results from Michigan MHD and RCM modeling were presented.  The MHD code
directly drives RCM inputs, but RCM products are applied with a different
time-scale to MHD code operation.  This coupling results in a different
pressure distribution.  It also results in a change in the magnetic field
configuration and clearly gets the development of a region 2 current system.
Pressure pushes the reconnection site further down the tail.  Differences in
grid resolution that result in some computational diffusion have yet to be
dealt with.  In the future, comparison will be made between potentials and
Birkeland currents computed by RCM with those computed from MHD and RCM will
be further extended into the nightside, simulating a magnetic storm.


WG1 Plans for the future:

This last year has been one of transitions.  Important new observational
techniques have been developed, which now make possible much more global
representations of ionospheric, plasmaspheric, and ring current plasmas. At
the same time, new efforts to couple plasma population models have
successfully advanced physical modeling.  In this context, this year's
IMS/WG1 sessions made it possible to share many of these advances.

WG1 sessions for June 2003 will consequently refocus on GEM storms, both old
and new, for the purpose of encouraging the application of new observational
and modeling techniques.  GEM storm times will be revisited at the mini-GEM
meeting at Fall AGU 2002.  At that meeting, participants are invited to argue
for including specific storms for discussion in the following June 2003
Workshop.  The following storm times have been proposed so far as candidates:

Previously-chosen storms:
	January 10, 1997
	May 15, 1997
	September 25, 1998
	October 19, 1998
	October 4-7, 2000
	March 31-April 2, 2001
Newly-suggested storms:
	April 6, 2000
	September 17, 2000
	April 11, 2001
	October 21, 2001
	November 6, 2001
	April 17-24, 2002

You'll note that the list includes storms from the previous GEM list and new
storms.  Participants at the mini-GEM in December will down-select to a
shorter list of storm-times for community-wide examination.  It is
encouraged for those interested in promoting a particular storm to send
data, model results, and explanations to the GEM-Storms webmaster (go to
http://leadbelly.lanl.gov/GEM_Storms/GEMstorms.html).

Following the mini-GEM, providers of data products will be invited to
contribute data relevant to these down-selected storm-times for posting to
the GEM web pages.  These data products will then be available for model
comparison.  Modelers are solicited to quantify the important features of
their models, which will be used for modeling the selected events for GEM
2003, through submission of Model Vitas.  Web posted vitas need to be
submitted prior to the June 2003 meeting along with "key parameters" that
summarize modeling results.

The current plan is for the first WG1 session to consist of presentations by
data product providers who will summarize their data by storm.  Subsequent
sessions will focus, storm by storm, on modeling results.  Some of these
sessions are likely to be jointly held with other working groups.  Vita and
key parameter results will be used to facilitate the comparison of modeling
results and the identification of modeling technique strengths and
weaknesses.

The proposed Model Vita format contains:
 * A list of relevant inputs, processes, and calculational methods
     So others can reproduce results
     So others can understand strengths and weaknesses
     So others can contrast and compare various model results
 * A vita for each storm event simulation run

Model presenters are further encouraged to identify key parameters of the
physical system that result from their modeling efforts.  Discussions at the
June meeting will make it possible for the community to select among these
provided parameters to identify common community parameters for future work.


****************************************************************************

IMS WG2 (RADIATION BELTS) REPORT

WG2 Co-chairs: Geoff Reeves (reeves at lanl.gov)
               Richard Thorne (rmt at atmos.ucla.edu)

At the 2002 GEM summer workshop the Inner Magnetosphere-Storms 
Working Group 2 (IMS WG2) held several oral and poster sessions.

Inner Magnetospheric Magnetic and Electric Fields (joint with WG1)
	Anthony Chan & Mike Liemohn chairs
Recent Theory and Modeling of Radiation Belt Acceleration
	Richard Thorne and Anthony Chan, chairs
Radiation Belt Electron Loss Processes
	Richard Thorne and Kirsten Lorentzen, chairs
The Magnetosphere and Ionosphere under Extreme conditions: Points of 
Contact between Global MHD Simulations, IMS Modeling, and Data (joint with
GGCM and MIC)
       George Siscoe and Janet Kozyra, chairs
Observations of the Radiation Belts and the Radiation Belt-Storm connection
	Geoff Reeves and Paul O'Brien, chairs
Solar Wind Drivers of Radiation Belt events
	Xinlin Li and Geoff Reeves, chairs
End-to-end Models and "Cartoons" of Radiation Belt Events
	Geoff Reeves and Mary Hudson, chairs
And a Wrap-Up/Future Plans session which was also joint with Working Group 1.

A report on the sessions on "The Magnetosphere and Ionosphere under Extreme
conditions: Points of Contact between Global MHD Simulations, IMS Modeling,
and Data" has been published in THE GEM MESSENGER, Volume 12, Number 24.

The Working Group 2 tutorial was by Richard Horne of the British
Antarctic Survey who spoke about the contribution of wave particle
interactions to acceleration and loss of radiation belt electrons.

There were also approximately 20 posters presented on radiation belt 
dynamics, theory, modeling, and related processes. This year the 
poster presentations sessions were notably well-attended and rich in 
content.

Several important areas of consensus emerged from the discussion of 
the oral and poster presentations.

There is now wide-spread appreciation of the delicate balance between 
loss processes and acceleration processes, both of which appear to be 
enhanced during storms. One result is that any given storm can 
produce either an enhancement or a reduction in relativistic electron 
fluxes throughout the radiation belts. This has several important 
implications, one of which is that the amount of acceleration cannot 
be quantified without having a simultaneous quantitative 
understanding of electron loss. Models of electron acceleration 
without losses will underestimate the amount of acceleration. 
Estimates of electron precipitation rates as well as theoretical 
calculations of pitch angle diffusion suggest that loss processes could 
essentially empty the radiation belts in a matter of days under 
storm-time conditions (and if simultaneous acceleration/transport 
processes were not acting simultaneously.)

The investigation of acceleration and transport processes (both 
observational and theoretical) continued at this workshop but the 
emphasis has changed from evaluating the possibility of certain 
mechanisms to developing quantitative calculations of the amount of 
acceleration and transport.

Drift resonance with ULF waves has been shown to produce enhanced 
radial diffusion leading to radiation belt enhancements. Several 
papers therefore looked more closely at the ULF wave fields during 
geomagnetic storms. It was shown, for example, that ULF wave power is 
significantly higher during storms than at other times and such 
enhancements are seen both in ground magnetometer data and in MHD 
simulations. It was also shown that the location of the ULF wave 
fields as a function of L-shell is also highly variable with ULF wave 
power geometrically increasing with higher L-shell. Evidence was 
also shown for a minimum L with significant wave power with that 
cut-off related to the auroral boundary index. These ULF wave fields 
were used in particle simulations to demonstrate that the particle 
dynamics and diffusion rates obtained from the drift resonances can 
account for relativistic electron enhancements.

Wave particle interactions with VLF & EMIC waves were also 
investigated in more quantitative detail than in previous years. A 
prominent feature of this class of interactions is that it integrates 
acceleration and loss processes into a single physical model with 
both processes driven by the same wave fields. Another key difference 
between this class of interactions, compared to the ULF interactions, 
is the importance of continued substorm activity during the recovery 
phase of a storm. Substorm injections produce the substorm-associated 
VLF chorus emissions that are responsible for the electron 
gyro-resonance and energy+pitch angle diffusion process. Quantitative 
modeling of this scenario showed that, with continued enhanced wave 
fields, the electron spectrum could continually harden at L=4 over at 
least a week, even with significant losses occurring simultaneously. 

Analytic treatments of wave-particle interactions (both ULF and VLF) 
have developed considerably in the past year. Those studies show 
promise for developing better physical understanding and numerical 
simulation of particle acceleration and losses.

Solar wind-magnetospheric electron interactions were discussed in the 
context of radiation belt events. Those interactions were discussed 
in the context of semi-empirical models of electron fluxes that can 
be used for space weather applications as well as in the context of 
physical interactions between the solar wind and magnetosphere which 
could, for instance, produce the observed wave fields. The role of 
high solar wind velocity  and enhanced dayside reconnection were 
joined by the role of intrinsic fluctuations in the solar wind in 
capturing the attention of the working group. Efforts to disentangle 
these potential drivers  (which tend to occur together) by 
statistical analysis and by modeling of hypothetical conditions is 
leading to a tantalizing clues about which conditions are necessary 
and/or sufficient to produce relativistic electron events.

While these topics define some of the dominant themes of the meeting 
there were numerous and significant presentations that are not 
mentioned here. Many of those were in the nature of raising 
interesting observations for which clear explanations do not yet 
exist. In the course of our discussions we developed broad consensus 
on several questions that the campaign will focus on between now and 
the 2003 summer workshop. A partial list includes:

While the largest changes in relativistic electrons occur during 
storms can the same processes occur during non-storm times? These 
might include "small storms", high solar wind velocities with little 
or no Dst, intervals of strong solar wind ULF wave power with little 
or no geomagnetic activity.  These events may be rare but would 
provide a sensitive test of which conditions are necessary and/or 
sufficient.

What are the IMF and magnetospheric conditions during storm recovery 
phase? How do they relate to the ultimate flux levels of radiation 
belt electrons and what do they tell us about the acceleration and 
transport mechanisms? How do they influence the spectral, spatial, 
and temporal structure of the belts?

What is the role of the plasmasheet source population?  What are its 
characteristics? How are they different during efficient and 
inefficient electron acceleration events? How are they trapped?

What is the phase space density gradient as a function of L-shell and 
storm phase? How often and when are peaks observed? Are the peaks 
produced by localized acceleration or localized losses? What does the 
gradient look like at or outside the trapping boundary?

What are the statistical properties of the various relevant 
parameters for quantitative modeling of the radiation belts? Those 
properties include at least, wave fields (power and distribution) 
of ULF, VLF chorus, and EMIC waves, the rate and intensity (?) of 
substorms, the phase space density of the plasma sheet source 
population, the pitch angle distribution of radiation belt particles, 
the energy spectrum, etc. In particular we identified the need to 
compare MHD wave power calculations with ground-based observations.

Are there different processes for acceleration and loss that can have 
different magnitudes during storms? Or, are acceleration and loss 
intimately related by the same physical processes?

What determines the inner boundary of the radiation belts? Is the 
boundary primarily set by loss processes, by the location of an 
internal acceleration process, or the depth of penetration of the 
radial transport in a given event.

How do we quantify the relative magnitudes of acceleration and loss 
and the balance between the two?


In the summary session we reviewed the objectives of the campaign set 
in 1999. Those continue to be appropriate guides for our continued 
efforts. A small change recognizing the importance of loss as well as 
acceleration was incorporated so the objectives now read:
1) To evaluate the relative contribution of various proposed 
acceleration and loss processes through theory, modeling, and 
comparison with data.
2) To create time-dependent phase space density profiles of the 
radiation belts that will more accurately represent their structure 
and dynamics than fixed-energy profiles.

The evidence of published work on these topics shows that the 
campaign has been highly successful. The number and importance of 
outstanding questions shows that there is still considerable 
opportunity for improved scientific understanding.

****************************************************************************
						    
+-------------------------------------------------------------------------+
|To add name to the mailing list or for a message to the GEM community    |
| please contact: editor at igpp.ucla.edu                                    |
|                                                                         |
|URL of GEM Home Page: http://www-ssc.igpp.ucla.edu/gem/Welcome.html      |
|Please update your e-mail address.                                       |
+-------------------------------------------------------------------------+