The retarding potential analyzers on the Viking landers obtained the first in situ measurements of ions from another planetary ionosphere. Mars has an F1 ionospheric layer with a peak ion concentration of approximately 105 cm−3 just below 130‐km altitude, of which ∼90% are O2+ and 10% CO2+. At higher altitudes, O+ ions were detected with peak concentration near 225 km of less than 103 cm−3. Viking 1 measured ion temperatures of approximately 150°K near the F1 peak increasing to an apparent exospheric temperature of 210°K near 175 km. Above this altitude, departures from thermal equilibrium with the neutral gas occur, and Ti increases rapidly to >1000°K at 250 km. An equatorward horizontal ion velocity of the order of 100–200 m/s was observed near 200 km and near the F1 peak, with a minimum velocity at intermediate heights. Both landers entered the F1 layer at a solar zenith angle near 44°, though the local times of the Viking 1 and 2 entries were 16:13 and 9:49 LT, respectively. On Viking 2, considerably more structure was observed in the height profiles of ionospheric quantities, although they were similar in shape to the Viking 1 profiles.

A planar retarding‐potential analyzer will be included in each of the Atmosphere Explorer (AE) payloads. The primary functions of the instrument are to supply ion‐temperature and ion‐concentration data, which it will do at least once every 40 km of flight path in the region of interest with an expected accuracy of better than 2%. In addition, the instrument will determine ion‐drift velocities and energy spectra of both the thermal and suprathermal electrons. During eclipse, negative‐ion concentrations greater than 0.3 cm−3 can be detected, though the normal positive‐ion sensitivity is approximately 1.5 ions cm−3. Changes in the ion concentration along the flight path greater than 0.1% can be monitored with a spatial resolution of less than 40 m.

The nature of the results from the retarding potential analyzer on Ogo 6 are described. The device appears to be capable of measuring ion temperature to an accuracy of better than 10% in a quiet ionosphere. In the dawn‐dusk plane the ion temperature is observed to vary from 1000° to 4000°K; the higher temperatures are associated with higher altitudes in the winter hemisphere. Molecular ions are detected near perigee (400 km), but their concentrations seldom exceed 1% of the O+ ion concentration. At midlatitudes and higher altitudes (>700 km) the lighter ions H+ and He+ are both present at approximately the O+ concentration, but their relative importance decreases near and in the polar regions. Both polar regions display large spatial fluctuations in ion concentration, and this behavior extends somewhat into the plasmasphere. Fluxes of electrons with energy greater than 10 ev of the order of 108 cm−2 sec−1 are observed; the fluxes change rather smoothly within the plasmasphere but show rapid and large variations at higher latitudes. The device also operates in a mode that examines the horizontal changes in ion concentration (fractional changes as small as 10−3 can be observed) with a spatial resolution from 350 to as small as 40 meters, depending on the telemetry rate.

A simple, nearly passive technique is described that allows a spacecraft sensor to be driven to a potential close to that of a plasma where the ion concentration exceeds about 100 cm−3 even in the presence of a large vehicle potential. Such a technique may become increasingly useful in the event that compatibility constraints in the Space Shuttle demand a 28‐V negative ground spacecraft system.

Frank et al. [1986a] report that exceptionally dark spots in the ultraviolet images of the sunlit Earth provided by the Spin-scan Auroral Imager (SAI) on the Dynamics Explorer I satellite occur far more frequently than would be expected from Poisson counting statistics, which ordinarily govern the statistical distribution of counts from a photomultiplier tube.Frank et al. [1986b] postulate that the spots, which are usually confined to a single pixel, could be due to obscuration of the photometer field of view by a population of screening objects.They conclude that these objects may be water vapor clot{ds from a previously undetected, 20 per minute influx of 100-ton icy comets into the Earth's upper atmosphere.We present here a statistical analysis of "dark pixel" events in a data base of 182 images kindly provided by Frank et al. Based on various instrumental and geophysical considerations,Chubb [1986] argued that these events are probably instrumental artifacts. Theoretical analysis provides a quantitative basis for Chubb's considerations regarding the overlap of adjacent pixels along scan lines in the image, and shows that pixels adjacent to the very dark ones should themselves be darkened by at least 9% on the average, assuming that screening objects cause the events. In our data base, however, the adjacent pixels are not darkened beyond statistical error bars, which are less than 1% for a data set containing •500 dark spots. Frank et al. [1986a] Because the fields of view for successive pixels overlap, the GSO screening of the field of view required to explain an anomalously dark pixel turns out to imply a partial screening of the fields of view for the pixels immediately preceeding and following it in the scan line. Consider, for example, screening by a long absorbing band of width W and optical transmissivity T, aligned perpendicular to the scan direction, but exactly centered on one pixel's field of view. Detailed calculations that include the exposure-time weighting effect show that if T -0 and W is chosen so as to yield a 50% reduction of the expected count in this pixel, then W turns out to be 0.14 ø and the expected average count reduction in both adjacent pixels is 5.9%. If T -0.5 and W _• 0.52 ø, then the average center-pixel count would again be reduced by 50%, but the two adjacent pixel counts would both be reduced on average by more than 25%. Typical anomalously dark pixels have count reductions significantly greater than the 50% value assumed above for illustration.We have also considered screening due to a more geophysically reasonable class of GSO's, circular or spherical objects of various sizes and zero transmissivity with random positions 573