Some of the most important recent progress in polar outflow studies has been in the development of powerful new modelling approaches, chiefly under the categories of generalized transport (high-order moments, through parallel and perpendicular heat flux) models [e.g., Gombosi and Rasmussen, [1991]] and Generalized Semi-Kinetic(GSK) models [e.g., Wilson et al., [1992]]. (GSK transport modelling is taken here to mean all of those treatments which involve large-scale modeling in which the ions are treated as particles and electrons as a fluid.) In this connection, one of the most controversial issues at present is the degree to which such(computationally fast) generalized transport models produce results in agreement with predictions of semikinetic models. Demars and Schunk [[1992]] compared sixteen-moment generalized transport model results for the steady-state supersonic polar and solar wind, finding excellent agreement between all moments through parallel and perpendicular heat flux for these two models. On the other hand, Ho et al. [[1993a]] examined time-dependent problems of the expansion of the polar wind and effects of density perturbations for these two principal polar wind transport approaches. They found that steep gradients in bulk parameters, which were dissipated in the semikinetic results due to (stress-releasing) velocity dispersion effects, were much more persistent in the fluid model. However, these authors allowed for the possibility that directly solving the heat flow equations as part of the transport equation set could improve the agreement.
Various modeling methods were also applied to
phenomenon of convection-driven heating of the ionospheric
O
through friction with the neutral atmosphere, and the consequent
upward O
flow. Korosmezey et al. [[1992]]
applied generalized transport methods to this problem, and
found that their resulting
velocities tended to be smaller than thosed observed by radars and satellites.
At the time, such generalized transport methods were regarded
as essential to describe this region in which frequent collisions in the
lower
ionospheric region make kinetic methods cumbersome, yet important anisotropies
and heat flows may render low-order fluid methods inaccurate. Nevertheless,
Heelis et al. [[1993]] used a simpler isotropic
fluid model to study
the response of the F-region
plasma to Sub-Auroral Ion Drift(SAID) 2-4 km/s convective drift events.
The modeled peak upflow velocities varied more than a factor of ten
depending on whether the onset rise time to large convection
is fast(seconds) or slow(several
minutes), with the lowest upflow velocities being
associated with the slower rise time cases. In contrast, kinetic features of
the field-aligned flow of convection-heated
O
were examined by Loranc and
St.-Maurice [[1994]] and Wilson [[1994]], with
the results of Wilson [[1994]] apparently indicating upflows
between the smaller generalized transport results of
Korosmezey et al. [[1992]]
and the larger flow speeds of Heelis et al. [[1993]].
In addition to direct convection-driven frictional heating, Ganguli et al. [[1994]] have shown spacecraft measurements which indicate strong ion heating and upward flows to be located in convection flow reversal regions, where the frictional heating is small. Their theoretical analysis suggested that such convection shears can lead to waves in the ion cyclotron and lower hybrid frequency range which will heat the ions to drive the upflows in those regions.
During this reporting period, various methods were employed to model the
interaction of hot
magnetospheric plasmas and cold ionospheric plasmas on polar field lines, and
the possible consequences for the ionospheric plasma outflow.
The reader is again referred to Shelley [[this issue]] for
auroral studies.
An emphasis on the hot electron-affected polar
outflow characteristics was found in the semikinetic study of
Ho et al. [[1992a]],
who showed a topside O
escaping flux peaking at
about 10
ions/(cm
-s)
when the high-altitude electron
temperature was 2 x 10
K. This peak results from the competing
effects of
upward ambipolar electric fields which increase with local electron
temperature, and
downward electric fields, associated with electron temperature gradients.
In an interesting observational/modeling
companion set dealing with magnetosheath effects on ionospheric outflow,
Winglee et al. [[1993a],
[1993b]] first presented data showing that during
downward magnetosheath injections, the upwelling ionospheric flux decreased
while the energy decreased, and also they also
may have the first downward
ionospheric conics(conics are
particle flux distributions which peak at angles
between parallel and perpendicular to the magnetic field direction).
Using small-scale particle-in-cell (PIC)
simulations, Winglee et al. [[1993b]] indicated that
through wave-particle interactions,
the downflowing magnetosheath plasma can transfer momentum and transverse
energy to the ionospheric plasma, creating initially downward flowing conics
which later mirror into upflowing conics.
Long-overdue semikinetic modeling of the H
polar wind
collisional-to-collisionless flow
through O
was initiated during this period by Wilson [[1992]]
and Bargouthi et al. [[1993]].
Wilson [[1992]] studied
the transitional altitude range 500 km to 7000 km, considering
H
self-collisions and collisions with O
, and other effects.
Where H
-O
collisions
are more frequent than H
self-collisions(1900-2800 km altitude),
the bulk of the H
distribution is accelerated to
substantial flow velocity by the O
-electron ambipolar electric field,
but a portion of the lower-end tail of the distribution collides strongly
with the O
and is ``held'' near zero velocity. This is because the
collisional factors vary as
the inverse cube or fourth power of the relative velocity between
the colliding partners, an effect that is difficult to obtain with
the simplified collision terms often used in moment-based models.
Barghouthi et al. [[1993]]
used slightly different methods for
treating H
- O
collisions in a Monte Carlo simulation, but obtained
results similar to those of Wilson [[1992]].
Modeling efforts to understand the polar cap altitudinal density profile
and large O
outflow velocities, respectively, were
made by Ho and Horwitz [[1993]] and Horwitz et
al. [[1994]]. Ho and Horwitz [[1993]]
compared total
density profiles from an
H
/O
steady-state
polar wind model with the averaged electron density profiles from
earlier Dynamics Explorer-1(DE-1) plasma wave measurements,
finding remarkable agreement between model and empirical density profiles.
Horwitz et al. [[1994]]
subsequently incorporated the convection-driven centrifugal parallel force into
this model, showing that this force is a viable explanation for
the large 1-10 km/s O
outflow velocities observed by the
DE-1
and Akebono spacecraft, and also that centrifugal
acceleration
may greatly enhance the levels of outflowing oxygen ion fluxes, by
a factor of ten or more.
Surveys of ion outflows at low and moderate polar altitudes from the
Dynamics Explorer
mission contributed further to our detailed understanding of the polar
cap environment. The
extensive statistical studies of Loranc et al. [[1991]] and
Chandler et al. [[1991]] began to characterize the
vertical plasma flows at F-region to topside ionospheric altitudes.
Loranc et al. [[1991]] showed that the
morphology of regions of upflowing F-region O
matches with
the observed morphology of outflowing beams and conics at higher altitudes,
but with the typical F-region upfluxes being much larger. They also
suggested
that the principal driver for these F-region upflows is frictional heating
by convection. Chandler et al. [[1991]] performed
a statistical analysis
of the polar wind ion outflows from DE-1 measurements between 1000 and 4000 km
altitude, and provided the first statistically-averaged density profiles
of H
, He
and O
and velocity profiles of H
and He
in
this altitude range. An aspect of particular interest was
the unexpectedly large oxygen to hydrogen ion density ratio at high altitudes,
typically of the order ten or more at about 4000 km altitude.
For the mid-altitude polar cap magnetosphere,
Horwitz et al. [[1992a]] observed core O
outflowing streams
in the polar cap magnetosphere at 3-5 earth radii geocentric distance
and examined features in other aspects of
the polar cap environment using data from DE-1 and 2. They found that the
polar cap streams could be categorized into two fairly distinct types: (1)
high-speed streams(10-30 eV or more streaming energies), which tended to be
bursty and near field lines threading auroral arc features; and
(2) low-speed streams(streaming energies below about 10 eV), which were
relatively stable and on field lines threading the dark polar cap.
Giles et al. [[1994]] described the pitch angle distributions
of outflowing ionospheric ions from high latitudes.