Waves and Instabilities at Midlatitudes - The Earth's Ionosphere: Plasma Physics and Electrodynamics - page 326

and move parallel to B at the velocity u s cos I (as do the ions), the plasma just

moves with the neutral wind. Thus, to obtain the large electric fields, the patches
must have one dimension much larger than the other.
To summarize, the very existence of two-stream instabilities at midlatitudes
seems to require patchy but elongated sporadic E layers, and these indeed have
been observed. The driving mechanism is most likely to be large zonal neu-
tral winds and wind shears. But when the large scale F-region structures occur,
as discussed earlier in this chapter, large meridional electric fields exist, and
the F-region contribution to the load is greatly reduced where the plasma is at
high altitude (low airglow region). Both effects contribute to enhanced E-region
instabilities. Furthermore, Swartz et al. (2002) and Kelley et al. (2003a) show
examples in which F-region structures are highly correlated with strong E-region
irregularities. Since electric fields map both directions it is not obvious which
region is dominant when, as Swartz et al. (2002) found, the height-integrated
conductivities are comparable in the two zones.
Coupling between the E and F regions has been discussed for a long time,
going back to Farley (1959). At midlatitudes, the nighttime F-region conductiv-
ity exceeds that of the E region so it usually wins the electrodynamic battle. But
Tsunoda and Cosgrove (2001), Haldoupis et al. (2003), Cosgrove and Tsun-
oda (2004a, b), and Shalimov and Haldoupis (2005) have shown that E-region
sources can affect the F region.
At first, there was debate about the configuration of Fig. 6.39: Because the
polarization electric field would drive a strong secondary Hall current down
the length of the structure (southward), there appeared to be no way for this
current to close. Hence, it appeared that a secondary polarization effect would
quickly shut down the process, as indicated by (6.30c,d). However, Shalimov
et al. (1998) showed that if the structures were narrow in the east-west direc-
tion and extended in the north-south direction, then closure of the secondary
Hall current could occur through the F region. This closure mechanism relied
on the assumption that the zonal polarization electric field extended over a
short enough spatial scale that it would not map effectively to the F region.
Hysell and Burcham (2000) and Hysell et al. (2002) verified the Haldoupis et
al. (1996) theory with the Shalimov et al. (1998) current closure path using
simulations.
Tsunoda (1998) showed that the geometry of Fig. 6.39 can be generalized and
that the long direction of the plasma structure can have any alignment in the
horizontal plane, as long as there is a component of the Hall current (from the
background electric field or neutral wind) perpendicular to the long direction.
Also, Cosgrove and Tsunoda (2001) showed that if the mechanism was driven
by a sheared neutral wind, then closure of the secondary Hall current could
occur through other E s patches, eliminating the need for long narrow structures
and electrical contact with the F region. In fact, Cosgrove and Tsunoda (2002a)
showed that, simply given the configuration of an E s layer located in a vertical
wind shear (such as the one that supposedly formed the layer), generic alti-
tude or field-line-integrated (FLI) density distortions create polarization electric

−