Summary Working Group 2
SHINE 2001

The general task of working group 2 at the SHINE 2001 workshop was to
investigate the interplanetary connections of solar transient activity. In
particular, our aim was to address (in each of the three half-day
sessions) the following topics: (1) Complex versus simple CMEs; (2) Model
predictions; and (3) Are there two classes of (interplanetary) CMEs?

Session 1 (complex versus simple CMEs) was introduced by Len Burlaga, who
in an invited review during the plenary session, laid the groundwork for
interpreting interplanetary ejecta as either simple (cloud-like) or
complex (non-cloud-like). He discussed the difficulties in identifying
ejecta boundaries and suggested how best to locate them. He proposed two
possible evolutionary paths for simple CMEs into complex CMEs. In the
first, two or more CMEs interact with each other between the Sun and the
Earth, resulting in a complex structure. In the second, some instability
develops within the ejecta producing a complex ejecta.

During the first session presentations were made by Dave Webb and Ian
Richardson on complexity within CMEs. Richardson summarized common
interpretations of "complex ejecta,"  pointing out that the term was
sufficiently different to different groups that virtually all CMEs could
be considered complex! Webb illustrated this by noting that one of the
events in the working group's preliminary candidate list of simple CMEs
was also listed in the list of complex CMEs. Nat Gopalswarmy presented an
analysis of Radio burst measurements also suggesting that complex CMEs may
be produced by the interaction of two or more ejecta. Short presentations
summarizing datasets for a preliminary list of "compaign" events were made
by Mike Andrews, Garath Lawence, Len Burlaga, Dave Webb, and John
Steinberg.

These presentations stimulated several relevant discussions, one of which
concerned the importance and difficulties associated with identifying
ejecta boundaries. Using the event identified as both "simple" and
"complex," we realized that  its initial designation as complex relied on
a questionable interpretation of where counterstreaming suprathermal
electrons (CSEs) were present during the interval. On closer inspection we
were able to move the rear boundary forward in time, making the ejecta
"simple," and resolving the apparent discrepancy. The discussions then
moved to the question of what makes a complex CME complex? Are complex
CMEs just "composite" CMEs, produced from the interaction of two or more
ejecta? Are they "born" that way? Do they  evolve from a simpler
configuration via some instability? A complementary question was also
posed: What makes a simple CME simple? Are they also born that way? Do
they evolve from a more complex low-beta configuration through
relaxation? Is the only distinction between complex and simple ejecta an
observational selection effect: Does the trajectory of the spacecraft
through the ejecta dictate whether we observe a "simple" or
"complex" CME? Needless to say, we did not arrive at any definitive
answers to these questions! What we did learn though is that any CME will
be complex if you don't understand it!

In session 2 we turned our attention to the question of what the modellers
would/could predict in solar and/or in situ observations of CMEs,
particularly with regard to differentiating features between the
models. Presentations were made by Jim Klimchuk, Paul Bellan, Jon Linker,
Jack Gosling, Spiro Antiochos, Jonathan Krall, and Vic Pizzo. Although the
idealized nature of the models generally precluded such predictions,
Klimchuk ventured that while there is, in principle, no limit to the
degree of twist in a flux rope  produced in the "breakout" model, a
maximum of 2-3 turns is all that can be produced in the "flux
rope" model. We discussed whether the models would place prominence
material at different locations in relation to the flux rope and how this
fit with in situ composition measurements that suggest cold material is
sometimes present at the back of the flux rope, while in other cases, it
is present at the front. The importance of prominence material (perhaps
containing as much mass as the CME itself) was discussed and Antiochos
argued that a fundamental distinction between the "breakout" and "flux
rope" models was that in the latter, either all or none of the prominence
material must be ejected. We also asked whether the formation of a flux
rope is a necessary part of models, which it apparently is. In a
generalization of the topic of associating the "three part structure" seen
in coronagraph observations with in situ structures, we discussed the
topology of magnetic field lines, as inferred from the signature of
CSEs. Linker pointed out that the presence of an overlying arcade might
translate into  CSEs preceding the flux rope, whereas heat flux dropouts
(suggesting the complete disconnection of field lines) should trail the
ejecta. Gosling emphasized that this picture would likely be complicated
since the reconnection was expected to occur in a patchy fashion along the
neutral line. Finally, Pizzo discussed the significant interaction of the
ambient solar wind with the ejecta, and to what degree the flux rope would
maintain its integrity. 

In the third, and final session for working group 2 we returned to a
"challenge question" posed at the end of the SHINE 2000 workshop, namely,
are there two classes of CMEs?  To address this, we began with
presentations from Mike Andrews, David Alexander, and Seiji Yashiro. This
topic had formed the basis of a special session at the Spring AGU, and a
consensus view had already formed that while two classes may exist, the
now infamous "Sheeley Plot" may not be the best place to start. As
Alexander pointed out, to really try to resolve this from a kinematic
perspective, we need acceleration profiles at lower altitudes. Andrews
argued that mass and energy are better delineators than velocity profiles
and Thomas Zurbuchen suggested that we might compare in situ
"conserved" quantities with solar observations. To make further progress
on some of the topics discussed at the workshop, we agreed to screen the
original list of campaign events. In particular, a subgroup of the
participants will identify 4 events: a "simple" event that drives a shock
and one that does not, and a "complex" event that drives a shock and one
that does not. We completed the session by defining some "challenges" to
be addressed during the upcoming year. Our first challenges are directed
at the solar modellers: (1) to provide differentiating observational
signatures predicted from their models; and (2) to model a non-flux rope
CME. For ourselves, we posed the following question: What are the origins
of complexity within CMEs?