Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
ELECTRICAL STRUCTURE OF THE MIDDLE ATMOSPHERE 188 (left-hand panel). The stratospheric ion chemistry leading to the formation of nonproton hydrates has been omitted. In both cases the proton hydrates are dominant in the lower mesosphere with a fairly abrupt transition to NO+ at greater heights in the quiet case and to mostly in the SPE case. A thin layer containing mostly the intermediate clusters of or NO+ separates the two regions. The tendency toward horizontal layering of the principal proton-hydrate species is caused by the shift in equilibrium toward lighter species as the temperature increases. Figure 13.6 shows the result of similar calculations of negative-ion composition, again omitting both the upper meteoric layer above 80 km and the stratospheric region of heavy HSO4- derived ions. In the undisturbed case is dominant in most of the middle atmosphere, but the great enhancement in the rate of ion-ion recombination during the SPE inhibits the formation of and leads to an increase in the fraction of . The initial ions O-and are the main components at the top of the mesosphere. The ultimate loss of ions in the middle atmosphere takes place mainly through recombination with electrons or with ions of the opposite charge. In the case of negative ions, photodetachment by sunlight provides another loss mechanism about which little quantitative information is available for the main atmospheric species. Loss to aerosol particles is presumably also a significant sink, especially in the vicinity of the stratospheric aerosol layer and possibly of the polar aerosol layers near the summer mesopause. Neither photodetachment nor aerosol attachment were included in the calculations represented by Figure 13.6. MESOSPHERIC AEROSOLS Stratospheric aerosols were discussed in Chapter 12 (this volume). The aerosol content of the mesosphere and the effect of these aerosols on the electrical properties are much more speculative. Noctilucent clouds provide direct evidence that particulate material does exist at mesospheric heights, at least on some occasions. These silvery-blue translucent clouds appear sporadically at high latitudes during the summer months, when their great height allows them to be seen by scattered sunlight long after the Sun has set at the surface. Their phenomenology has been reviewed by Fogle and Haurwitz (1966). Optical measurements suggest that the particle concentrations are 1 to 50 cm-3 in these clouds, with an individual particle radius of the order of 0.1 Âµm. Satellite measurements of backscattered sunlight reveal the existence near the mesopause of a denser semipermanent particle layer over the summer polar cap (Donahue et al., 1972; Thomas et al., 1982), which is probably related to the noctilucent cloud layer. There is general agreement that the particles forming these layers are ice crystals formed in the extremely low temperatures of the summer mesopause region by condensation from the low background concentration of water vapor (Hesstvedt, 1961). Several model calculations have shown that it is reasonable to expect ice crystals to form under these conditions (e.g., Charlson, 1965; Reid, 1975; Turco et al., 1982) provided that suitable nucleation centers exist. Mass-spectrometer results suggest that nucleation does occur on positive ions (Goldberg and Witt, 1977; BjÃ¶rn and Arnold, 1981), and the particles themselves could then act as surfaces for ion capture. There is evidence for abnormally low electron concentrations in the high-latitude summer mesopause region, perhaps indicating enhanced electron loss through attachment to particles (e.g., Pedersen et al., 1970). Below the mesopause, the evidence is much less direct. Volz and Goody (1962) found evidence from twilight measurements of low concentrations of dust particles throughout the mesosphere. Hunten et al. (1980) calculated the flux of particles produced by condensation of meteor ablation products. Depending on the model conditions, this calculation predicted mesospheric concentrations of 102 to 103 cm-3 and individual particle radii of a few nanometers. The influence of such a particle distribution on the ion and electron concentrations would be small. Chesworth and Hale (1974) proposed the existence of mesospheric particle concentrations of 103 to 104 cm-3 to explain certain discrepancies in the electrical parameters. The evidence for such large concentrations was indirect, and further work is needed in this area. In particular, the relationship, if any, between the meteor-ablation particles of Hunten et al. (1980) and the particles suggested by Chesworth and Hale (1974) should be studied. CONDUCTIVITY IN THE MIDDLE ATMOSPHERE The current density, j, and the electric field, E, are related by the familiar Ohm's law expression where Ï is the conductivity. In the lower atmosphere and most of the middle atmosphere, Ï is a scalar, and the electric field and the current lie in the same direction. Above about 70 km, however, collisions between electrons and air molecules become infrequent enough that the bending of the electron path by the Earth's magnetic field becomes appreciable between collisions, and mo
ELECTRICAL STRUCTURE OF THE MIDDLE ATMOSPHERE 189 tion perpendicular to the field becomes less easy than motion along the field. In these circumstances, Ï becomes a tensor, and an applied electric field with a component perpendicular to the magnetic field drives a current in a different direction. This anisotropy of the conductivity becomes a dominant influence on the electrical properties of the thermosphere (see Chapter 14, this volume). Conductivity can be measured directly by rocketborne probes, and a substantial number of such conductivity values are reported in the literature, especially using the so-called blunt-probe technique (Hale et al., 1968). The problem of shock-wave effects associated with a supersonic rocket are usually avoided in these experiments by deploying the probe with a parachute at the top of the trajectory and making the measurements during the subsonic descent. In the presence of a mixture of ions, the conductivity can be expressed in terms of the mobilities of the individual species as where n and k are the concentrations and mobilities of the positive ions, negative ions, and electrons. [Equation (13.3) is identical to Eq. (12.1) in Chapter 12, this volume, except that in the middle atmosphere electrons come into consideration.] An experimentally derived relationship between reduced mobility and ion mass in nitrogen is shown in Figure 13.7 (Meyerott et al., 1980). Figure 13.7 Reduced mobility as a function of ion mass. The reduced mobility, k0, is the mobility in a standard atmosphere and is related to the actual mobility, k, by where T is absolute temperature and p is pressure in millibars. The ion mobility can be measured by the Gerdien condenser technique (e.g., Conley, 1974), and a number of measurements using rockets have been reported. As discussed by Meyerott et al. (1980), there is no general agreement among the various measurements, although the data of Conley (1974) and Widdel et al. (1976) suggest a single reduced mobility of about 2.7 Ã 10-4 m2 V-1 sec-1 for the entire middle atmosphere. This corresponds to an ion mass of about 30 amu according to Figure 13.7, which is clearly in disagreement with both the mass-spectrometer measurements and the model studies discussed above. Meyerott et al. (1980) suggested that the Gerdien condenser measurements were affected by the breakup of cluster ions by both shock-wave and instrumental electric-field effects. Much heavier ions have been seen in some flights. Rose and Widdel (1972), for example, reported a group of positive ions above 60 km whose altitude dependence gives a reduced mobility of about 3.7 Ã 10-5 m2 V-1 sec-1, corresponding to an ion mass of several thousand amu. The origin of these ions is unknown. Ion concentrations are also measured by the Gerdien condenser technique, and here again there are puzzling differences between observations and predictions. Above about 60 km, ionization of nitric oxide by solar Lyman-alpha radiation becomes an important source, and a considerable amount of variability in ion concentration is expected (even in quiet conditions) owing to the variability in NO concentration (Solomon et al., 1982b). At lower altitudes, however, the only significant source is cosmic-ray ionization, and the ion-production rate due to cosmic rays can be calculated with a fair degree of certainty. The loss rate through ion-ion recombination is also well known (Smith and Church, 1977; Smith and Adams, 1982), and the steady-state ion concentration can thus be calculated with corresponding accuracy. Figure 13.8 shows a typical set of results. The points are taken from experimental data reported by Widdel et al. (1976), and the solid line is the positive-ion concentration calculated from the same model used to produce the ion- composition profiles shown in Figure 13.6. The overall shape of the altitude profile shows reasonable agreement, but the measured values are generally lower than the calculated values by about a factor of 3. A similar discrepancy was found by Meyerott et al.
ELECTRICAL STRUCTURE OF THE MIDDLE ATMOSPHERE 190 (1980), who suggested that it might be due to the lack of the experimental technique to measure low-mobility ions. Further measurements are clearly needed. Figure 13.8 Results of measurements of ion concentration (Widdel et al., 1976) and of model calculations for quiet conditions. Figure 13.9 Middle-atmosphere daytime conductivity. The dotted curve shows direct measurements (Mitchell and Hale, 1973) of positive conductivity, and the other curves show the result of model calculations of the contributions of the electrons, positive ions, and negative ions. Another puzzling feature shown in Figure 13.8 is the abrupt increase in concentration occurring at about 65 km in the 1968 data and 45 km in the 1975 data. A similar abrupt increase in conductivity at about 65 km is seen in blunt- probe measurements and has been attributed to the transition from a cosmic-ray ionization source below to a Lyman- alpha source above (Mitchell and Hale, 1973). The model results show such a transition, but it is much less abrupt and smaller in magnitude than the one observed. The same explanation is certainly ruled out for the 45-km transition, since solar Lyman-alpha radiation is almost totally extinguished below 50 km. Figure 13.9 shows the theoretical profiles of positive-ion, negative-ion, and electron contributions to the daytime conductivity, using the same model as before. The dotted curve shows an average of several blunt-probe measurements of positive conductivity in quiet conditions (Mitchell and Hale, 1973) and is taken as representative of the current experimental situation. Clearly the overall shape of the theoretical positive-conductivity profile matches the observations reasonably wellâparticularly below 60 km where cosmic rays are the principal source of ionization. At higher levels, the expected variability of ion concentration should lead to a corresponding variability in positive conductivity. The role of electrons is noteworthy. The theoretical profiles show that electrons make a negligible contribution to the total conductivity below 50 km but completely dominate the conductivity above 60 km, where they give rise to an extremely steep upward gradient in conductivity [the ''equalizing" layer (Dolezalek, 1972)]. The electron mobility is so large that the electron contribution to the conductivity becomes equal to the ion contribution at a level where the electron concentration is less than 1 cm-3. In this region the model predictions of the electron-ion balance are not trustworthy. In particular, the model used here does not include negative-ion photodetachment as a source of electrons, as photodetachment cross sections of the principal atmospheric negative ions are not known. Even small amounts of photodetachment, however, will have an important effect on electron concentrations in the region of the stratopause, and hence on the model conductivity profiles. The conductivity is greatly reduced at night at all heights above 50 km. The decrease is partly due to the absence of solar ionizing radiation and partly to changes in neutral chemistry, notably the conversion of atomic oxygen to ozone in the mesosphere. Figure 13.10 shows model profiles of daytime and nighttime conductivity for quiet mid- latitude conditions, in which the nighttime sources of ionization are mostly solar Lyman-alpha radiation scattered from the geocorona and the zonally averaged flux of energetic electrons at 45Â° magnetic latitude given by Vampola and Gorney (1983). The night