The periapsis (the lowest altitude in the orbit) of PVO was maintained near an altitude of 150 km until 1980, when the onboard fuel supply was just about exhausted. Solar tides then pushed the periapsis to higher altitudes until it reached about 2300 km in 1986, after which the periapsis began to drop. During this period of high periapsis no in situ ionospheric measurements were possible, although radio occultation observations provided electron density profiles throughout the life of PVO. Beginning around early 1992 the periapsis of PVO was again inside the ionosphere allowing direct measurements in the nightside of Venus. The small amount of remaining fuel was used to keep the periapsis around 140 km, but it was finally exhausted before the orbit precessed to the dayside and PVO entered the atmosphere. Most of the results from the entry phase portion of the PVO mission were published together in the December 14, 1993 issue of Geophysical Research Letters. It should be noted that the bulk of the entry phase ionospheric measurements were restricted to nightside observations.
The retrograde rotation of Venus results in a very long nighttime period
(
58 earth days); therefore it was first thought that the absence of solar ionizing
(extreme ultraviolet, EUV) radiation for such a long period would virtually
eliminate the ionosphere. The detection of a robust nightside ionosphere by
Mariner 5 was thus a surprise. To explain this observed nightside ionosphere
became a major research objective. Two competing mechanisms responsible for
the maintenance of the nightside ionosphere were proposed (i) plasma transport
from the dayside and (ii) ionization by precipitating low-energy (
40-200 eV)
electrons. Model calculations and PVO data prior to that obtained from the entry
phase seemed to indicate that day-to-night plasma transport is the main source
during solar maximum conditions and impact ionization becomes the dominant
mechanism during solar minimum. The early phase of the PVO data base, 1978-1981,
corresponds to high solar activity (
), while the entry phase data
corresponds to low solar activity (
). The 10.7 cm radio emission,
, monitored at the surface of the Earth, has been used as a proxy
indicator of solar EUV emissions. The observed mean nightside electron
densities, at the higher altitudes, are significantly lower during moderate
than at high solar activity; however, the solar cycle differences are small
near the density peak [ Theis and Brace, 1993]. These observations support
the results of model calculations [ Cravens et al., 1983; Fox, 1992; Brannon
et al., 1993; Dobe et al., 1995], which indicated that day to night plasma
transport is responsible for most of the plasma during high solar activity,
but that during solar cycle minimum conditions the decrease in ionopause
altitude chokes off the flow and precipitation becomes the dominant ionization
source. The latter is sufficient to maintain a peak electron density close to
the solar maximum value (within maybe a factor of about two), but results in
significant decreases at the higher altitudes. This behavior is due to the fact
that the impact ionization rate peaks near 140 km and drops off rapidly with
the neutral density at higher altitudes. The observed solar cycle dependence of
the nighttime electron temperature is more puzzling. Theis and Brace [1993]
argue that the temperature increase at higher altitudes during low solar
activity might be due to a combination of decreased densities and a ``better''
thermal connection to the hot sheath plasma, while the decrease at the lower
altitude is caused by a reduction in the thermal coupling to the dayside.
The observed large decrease in electron densities and only limited increase
in plasma temperatures above about 160 km during low solar activity results in
the dominance of the magnetic pressure over the kinetic pressure in the upper
nighttime ionosphere [ Russell et al., 1993]; this is a change from solar cycle
maximum behavior when the kinetic pressure is the dominant one throughout the
ionosphere.
Ion composition measurements made during low solar cycle activity [ Cloutier
et al., 1993; Grebowsky et al., 1993] helps to elucidate further the sources
of nightside ionization. These observations show that the relative abundance
of atomic ions (e.g., O
and H
) decrease significantly compared to molecular
ones (e.g., O
). More than a decade ago Cravens et al. [1983], and more recently
Brannon et al. [1993] and Dobe et al. [1995], showed that electron impact
ionization can produce molecular ion densities near 140 km comparable to those
resulting from day to night transport, but produce much smaller O
densities
above the ionization peak. The recent model calculations of Brannon et al.
[1993], Kar et al. [1994] and Dobe et al. [1995] show quantitatively that using
precipitating low energy electron flux values, consistent with the limited
observational constraints, can reproduce most of the ion densities observed
during the PVO entry phase; however, the presence of a much reduced day-to-night
transport source cannot be excluded. In summary, the new, nighttime, in situ
ionospheric data, obtained during the PVO entry phase, helped to elucidate the
processes responsible for the maintenance of the nightside Venus ionosphere
during conditions of both moderate and high solar activity.
Recently model studies have also explored the question of which mechanism or mechanisms maintain the nightside ionosphere of Mars. Haider et al. [1992] and Fox et al. [1993] looked at the potential electron precipitation source; Fox et al. [1993] also studied plasma transport from the dayside. These two potential source mechanisms are not inconsistent with the limited observational constraints set by radio occultation measurements of electron densities [e.g., Zhang et al., 1990]; however, more information (e.g., ion composition measurements) is needed to establish what the major sources of the nightside ionosphere are.
The question whether Mars has an intrinsic magnetic field is still not resolved. Building on earlier studies, Shinigawa and Cravens [1992] used their ionospheric MHD (magneto-hydrodynamic) model to consider the effects of different orientations of a weak horizontal intrinsic field on the dayside ionosphere of Mars and concluded that such a field would have to be less than about 50 nT (nanoTesla) in order for its effects on the ionospheric electron density structure to be evident. Zhang and Luhmann [1992], based on the examination of peak ionospheric pressures at Venus and Mars, also concluded that any intrinsic magnetic field at Mars must be very small.
The relative importance of different potential energy sources and transport mechanisms of the dayside ionospheres of Venus and Mars has been debated for about two decades [cf. Nagy et al., 1983], ever since it became obvious that solar EUV heating and classical thermal heat conduction lead to electron and ion temperatures significantly smaller than the measured values. Model calculations lead to temperature values consistent with the observed ones if either an ad-hoc topside heat inflow into an unmagnetized ionosphere is assumed and/or the electron thermal conductivity is reduced from its classical value. There have been a number of reasonable suggestions for sources and mechanisms of a heat inflow [ Taylor et al., 1979; Gan et al., 1990; Szego et al., 1991] and the observed presence of fluctuating and small steady magnetic fields in the ionosphere is consistent with reduced thermal conductivity. However, the data base has been insufficient to establish the relative importance of these two mechanisms. It was hoped that further in situ data from the dayside ionosphere would be obtained from PVO during its entry phase. Unfortunately, atmospheric entry and burnup occurred before any meaningful dayside data could be obtained. This led Dobe et al. [1993] to reexamine the old PVO database. They concluded that elevated temperatures at Venus are most likely the result of reduced thermal conductivity, but no definitive conclusions can be made without further information. No new missions are planned to explore Venus further, but the great similarities between the ionospheres of Venus and Mars mean that data from the upcoming Mars 96 and Planet B missions may not only help us to advance our understanding of the ionosphere of Mars, but also help to answer some of the remaining questions concerning Venus.
Finally, it should be mentioned that ionospheric processes, such as dissociative
recombination of molecular ions (e.g., O
, N
), are among the most important
mechanisms responsible for atmospheric escape at Venus and Mars. Theoretical
calculations by Fox [1993] indicate that, assuming an initial CO
pressure of
2 bars at Mars, an exponential loss of the atmosphere, with a time constant
of 700 m.y., is consistent with the observed isotope ratio of atomic nitrogen.
Different isotopes of a given atom have different atomic masses and thus have
different escape rates. In another theoretical study Zhang et al. [1993]
examined the loss of oxygen from Mars, which is thought to be an indicator of
the loss of water, via ionospheric processes. They concluded that given their
assumptions (e.g., changing solar EUV intensities), these ionospheric processes
may have been responsible for the loss to space of about 30 meters of water
on an evolutionary time scale. This calculated escape rate is higher than the
escape rate due to O
and O
ions in the tail of Mars, as observed by one of
the ion mass spectrometers on Phobos-2 [ Lundin et al., 1990]. The measured
escape rate was of the order of 
ions/sec, which corresponds to a loss of
water of only about 1 meter over Mars' history.