Calculated photoelectron pitch angle and energy spectra

Calculations of the steady-state photoelectron energy and angular distribution in the altitude region between 120 and 1000 km are presented. The distribution is found to be isotropic at all altitudes below 250 km, while above this altitude anisotropies in both pitch angle and energy are found. The isotropy found in the angular distribution below 250 km implies that photoelectron transport below 250 km is insignificant, while the angular anisotropy found above this altitude implies a net photoelectron current in the upward direction. The energy anisotropy above 500 km arises from the selective backscattering of the low energy photoelectron population of the upward flux component by Coulomb collisions with the ambient ions. The total photoelectron flux attains its maximum value between about 40 and 70 km above the altitude at which the photoelectron production rate is maximum. The displacement of the maximum of the equilibrium flux is attributed to an increasing (with altitude) photoelectron lifetime. Photoelectrons at altitudes above that where the flux is maximum are on the average more energetic than those below that altitude. The flux of photoelectrons escaping to the protonosphere at dawn was found to be 2.6 × 10⁸ cm^?2 sec^?1, while the escaping flux at noon was found to be 1.5 × 10⁸ cm^?2 sec^?1. The corresponding escaping energy fluxes are: 4.4 × 10⁹ eV cm^?2 sec^?1 and 2.7 × 10⁹ eV cm^?2 sec^?1.     

Measurements of the ambient photoelectron spectrum from atmosphere explorer: I. AE-E measurements below 300 km during solar minimum conditions

The ambient photoelectron spectrum below 300 km has been studied for a sample of 500 AE-E orbits taken during the period 13 December 1975 to 24 February 1976. During this solar minimum period, the average and maximum Σ K_p were 19 and 35 respectively. The agreement between the measured spectral shape and several recent calculations is extremely good. The daytime photoelectron spectrum below 300 km from 1 to 100 eV is illustrated by a number of spectra. Detailed 0–32 eV spectra are presented at various altitudes and solar zenith angles. High resolution 10–32eV spectra show the widths of the photoelectron lines in the spectrum and the variation of the linewidth and intensity with altitude. Data from the entire 500 orbit sample are combined into plots of the average flux over a number of altitude ranges up to 300 km at various local times and solar zenith angles. The data show that the photoelectron flux below 300 km is remarkably constant (typical variation less than ±50%) over a period of several months. The photoelectron lines between 20 and 30 eV are extremely sharp when the total plasma density is low but broaden significantly at high altitudes as the plasma density builds up during the day. The N₂ vibration-rotation excitation dip at 2.3 eV is strongest at the lowest altitudes and decreases with increasing altitude and plasma density. The absolute accuracy of the experiment is discussed in detail and a correction factor for previously published AE-E fluxes is given.     

Observed heating effects of conjugate photoelectrons

A gridded spherical electrostatic analyzer aboard Injun 5 has been used to measure fluxes of thermal and hyperthermal electrons at subauroral latitudes in the midnight sector of the northern ionosphere between altitudes of 2500 and 850 km. Due to the offset between the geomagnetic and geographic poles hyperthermal fluxes, consisting of energetic photoelectrons that have escaped from the sunlit southern hemisphere are observed along orbits over the Atlantic Ocean and North America but not over Asia. The ambient electron temperatures (T_e) near 2500 km have their highest values at trough latitudes for all longitudes. At altitudes near 1000 km elevated electron temperatures in the trough were not a consistent feature of the data. Equatorward of the trough, in the longitude sector to which conjugate photoelectrons have access, T_e ～ 4000 K at 2500 km and ～ 3000 K at 1000 km. For regions with the conjugate point in darkness T_e ? 2300 K over the 1000–2500 km altitude range. The effective thermal characteristics of conjugate photoelectrons are studied as functions of altitude and latitude. The observations indicate that (1) at trough latitudes elevated electron temperatures in the topside ionosphere are mostly produced by sources other than conjugate photoelectrons, and (2) at subtrough latitudes, in the Alantic Ocean-North American longitude sector, conjugate photoelectrons contribute significantly to the heating of topside electrons. Much of the conjugate photoelectron energy is deposited at altitudes >2500 km then conducted along magnetic field lines into the ionosphere.     

Thermally driven escape from Pluto’s atmosphere: A combined fluid/kinetic model

A combined fluid/kinetic model is developed to calculate thermally driven escape of N₂ from Pluto’s atmosphere for two solar heating conditions: no heating above 1450 km and solar minimum heating conditions. In the combined model, one-dimensional fluid equations are applied for the dense part of the atmosphere, while the exobase region is described by a kinetic model and calculated by the direct simulation Monte Carlo method. Fluid and kinetic parts of the model are iteratively solved in order to maintain constant total mass and energy fluxes through the simulation region. Although the atmosphere was found to be highly extended, with an exobase altitude at ～6000 km at solar minimum, the outflow remained subsonic and the escape rate was within a factor of two of the Jeans rate for the exobase temperatures determined. This picture is drastically different from recent predictions obtained solely using a fluid model which, in itself, requires assumptions about atmospheric density, flow velocity and energy flux carried away by escaping molecules at infinity. Gas temperature, density, velocity and heat flux versus radial distance are consistent between the hydrodynamic and kinetic model up to the exobase, only when the energy flux across the lower boundary and escape rate used to solve the hydrodynamic equations is obtained from the kinetic model. This limits the applicability of fluid models to atmospheric escape problems. Finally, the recent discovery of CO at high altitudes, the effect of Charon and the conditions at the New Horizon encounter are briefly considered.     

Ionospheric photoelectrons observed in the magnetosphere at distances up to 7 earth radii

Photoelectrons of ionospheric origin have been observed for the first time at high altitudes (up to 7R_E geocentric distance) using the suprathermal plasma analysers (SPA) on the GEOS satellites. At such high altitudes the photoelectron flux is confined within a few degrees of the magnetic field direction. We show how this flux may be identified and extracted from the background which is a combination of locally produced photoelectrons and ambient plasma. GEOS-2 results are presented to illustrate the “turn-on” of the photoelectron flux at dawn in the ionosphere. Data from GEOS-1 are used to study the behaviour of the photoelectron flux with equatorial geocentric distance from 3 to 7R_E. The results compare favourably with theoretical models and with ionospheric observations at mid latitudes.     

Field-aligned suprathermal electron fluxes below 270 km in the auroral zone

Observations are reported of field aligned etectron fluxes in the energy range 50–500 eV at altitudes below 270 km from two rocket flights in the auroral zone. The regions of field aligned suprathermal electrons occurred in bursts of a few seconds duration, and in some instances the energy of the peak field aligned flux was in the range 100–500 eV. Theoretical calculations of the pitch angle distribution were made using the Monte Carlo technique for two model atmospheres having exospheric temperatures of 750 and 1500 K bracketing the expected auroral zone exospheric temperature. The calculations were made for the case of incident field aligned suprathermal fluxes with no local parallel electric field and also for the case of a local constant parallel electric field. Comparison of theoretical and experimental pitch angle distributions showed that in one case at 270 km a parallel electric field of 1–2 mV/m fitted the data whereas another burst at 210 km required a parallel electric field of about 10 mV/m to produce a field aligned distribution of 230 eV electrons as pronounced as was observed. Furthermore in this latter case the lack of strong field alignment at 500 eV pointed to localisation of the parallel electric field to an altitude range of 20–30 km about the rocket altitude.     

Photoelectron distribution determination from plasma line intensity measurements obtained at Nancay (France)

The plasma oscillations that can be observed by the French incoherent scatter system have small phase velocities and are excited by low energy photoelectrons, typically 2–5 eV. Consequently, the method used to determine the energy photoelectron distribution from plasma line measurements made at other observatories (e.g. Cicerone, 1974) cannot be applied here: it is necessary to chose a model energy distribution with a small number of parameters. The energy shape of the flux is assumed Maxwellian and the angular shape is assumed linear with the cosine of the pitch angle. Total flux values and mean energies are obtained as a function of altitude, in agreement with other determinations, and the difference between upshifted and downshifted plasma line intensities lead to the determination of the anisotropy of the photoelectron flux.     

Numerical interpretation of high-altitude photoelectron observations

The Electron Spectrometer (ELS) instrument of the ASPERA-3 package on the Mars Express satellite has recorded photoelectron energy spectra up to apoapsis (∼10,000 km altitude). The characteristic photoelectron shape of the spectrum is sometimes seen well above the ionosphere in the evening sector across a wide range of near-equatorial latitudes. Two numerical models are used to analyze the characteristics of these high-altitude photoelectrons. The first is a global, multi-species MHD code that produces a 3-D representation of the magnetic field and bulk plasma parameters around Mars. It is used here to examine the possibility of magnetic connectivity between the high-altitude flanks of the martian ionosheath and the subsolar ionosphere. It is shown that some field lines in this region are draped interplanetary magnetic lines while others are open field lines (connected to both the IMF and the crustal magnetic field sources). The second model is a kinetic electron transport model that calculates the electron velocity space distribution along a selected, non-uniform, magnetic field line. It is used here to simulate the high-altitude ELS measurements. It is shown that the photoelectrons are essentially confined to the source cone, as governed by magnetic field inhomogeneity along the field line. Reasonable agreement is shown between the data and the model results, and a method is demonstrated for inferring properties of the local and photoelectron source region magnetic field from the ELS measurements. Specifically, the number of sectors in which photoelectrons are measured is a function of the magnetic field intensity ratio and the field's angle with respect to the detector plane. In addition, the sector of the photoelectron flux peak is a function of the magnetic field azimuthal angle in the detector plane.     

Investigation of energetic proton penetration in Titan's atmosphere using the Cassini INCA instrument

Saturn's largest moon, Titan, provides an interesting opportunity to study how dense atmospheres interact with the surrounding plasma environment. Without an intrinsic magnetic field, this satellite's nitrogen-rich atmosphere is relatively unprotected from plasma interactions. Therefore, the energy-deposition rate is important for understanding chemistry and dynamics in Titan's atmosphere. Since the plasma environment can vary significantly we focus here on the T18 Titan encounter using in-situ data from instruments on board the Cassini spacecraft. These instruments cannot provide in-situ information below the spacecraft closest approach altitude (∼>960 km) so we use the Cassini magnetospheric imaging instrument (MIMI) ion-neutral camera (INCA) to remotely image energetic hydrogen particle fluxes (20-80 keV) at altitudes below Titan closest approach. We also use the MIMI low-energy magnetosphere measurements system (LEMMS) to measure the incident ion fluxes as the spacecraft approaches Titan and combine these data sets with an atmospheric model to first reproduce INCA images. We then use this model to calculate the energy-deposition profiles for the observed incident proton flux. Our model is able to reproduce the INCA observations and give the energy density deposited vs. altitude in Titan's atmosphere; however, we find that the incident fluxes and energy-deposition profiles vary significantly during the encounter.     

The diurnal heat budget of the thermosphere

Detailed numerical calculations of thermospheric heat sources and sinks are presented and their relative importance is discussed in reference to the energy balance phenomena of the neutral atmosphere. It is shown that the thermal energy available from the absorption in the Schumann-Runge continuum leading to photo-dissociation of O₂ is by far the largest energy source in the lower thermosphere. Other sources of varying importance in different altitude ranges are: (1) energy from photoelectrons; (2) energy exchange from thermal plasma; (3) chemical reaction (ion-electron dissociative recombination) energy gain; (4) kinetic and dissipative energy associated with the neutral wind. The energy sinks of importance are (1) thermal conduction at the lower boundary (120km); and (2) radiative cooling of atomic oxygen.It is shown that the combined energy from processes 2–4 constitutes only a small fraction of the total energy available from photoelectrons and is in phase with the latter. These secondary sources (processes 2–4), therefore, do not constitute a significant energy source and their contribution can be simply incorporated into photoelectron energy (process 1) by defining an effective photo-ionization heating efficiency. The heating efficiencies for photo-ionization (including processes 2–4) and photo-dissociation are estimated to be 0.5 and 0.3, respectively.As the important heat input (photo-dissociation) and loss (conduction and radiation) rates are basically governed by the O₂ and O densities, any diurnal or seasonal variation in these constituents at the lower boundary would have profound effects on the thermal structure of the overlying atmosphere. For this and other reasons, it is suggested that a choice of lower boundary much below 120km, e.g. near the mesopause level (90 km), should be more appropriate for general thermospheric studies.     

Observations of OI 7774 emission excited by conjugate photoelectrons

Observations and computer calculations of OI 7774 airglow emissions excited by conjugate photoelectrons have been carried out. The observations were made at McDonald Observatory, Texas using a 2m grille spectrometer from December 1972 to June 1973. The zenithal emission intensity during conjugate photoelectron precipitation was fairly constant at 2–4 R until conjugate sunset, after which it diminished steadily and ceased near a conjugate solar zenith angle (χ_c) of 105 ± 3°. A predawn enhancement in both OI 7774 and [OI] 6300 was observed to commence near χ_c ～ 102°.The computations utilize the two-stream technique of Nagy and Banks (1970) to obtain the escaping photoelectron flux and the local excitation rates of the oxygen emissions. Good agreement with the observations is obtained for the dependence of the emission rate on conjugate solar zenith angle. A lack of agreement in absolute intensity may not be due entirely to uncertainties in the excitation cross section. The discrepancy may indicate significant magnetospheric scattering of photoelectrons with energy greater than 15 eV.     

Solar EUV and Electron-Proton-Hydrogen Atom-Produced Ionosphere on Mars: Comparative Studies of Particle Fluxes and Ion Production Rates Due to Different Processes

The total photoelectron and secondary electron fluxes are calculated at different times and altitudes along the trajectory of Mars Global Surveyor passing through the nightside and dayside martian ionosphere. These results are compared with the electron reflectometer experiment on board Mars Global Surveyor. The calculated electron spectra are in good agreement with this measurement. However, the combined fluxes of proton and hydrogen atom as calculated by E. Kallio and P. Janhunen (2001, J. Geophys. Res.106, 5617-5634) were found to be 1-2 orders of magnitude smaller than the measured spectra. We have also calculated ionization rates and ion and electron densities due to solar EUV, X-ray, and electron-proton-hydrogen atom impacting with atmospheric gases of Mars at solar zenith angles of 75°, 105°, and 127°. In the vicinity of the dayside ionization peak, it is found that the ion production rate caused by the precipitation of proton-hydrogen atom is larger than the X-ray impact ionization rate while at all altitudes, the photoionization rate is always greater than either of the two. Moreover, X-rays contribute greatly to the photoelectron impact ionization rate as compared to the photoion production rate. The calculated electron densities are compared with radio occultation measurements made by Mars Global Surveyor, Viking 1, and Mars 5 spacecraft at these solar zenith angles. The dayside ionosphere produced by proton-hydrogen atom is smaller by an order of magnitude than that produced by solar EUV radiation. X-rays play a significant role in the dayside ionosphere of Mars at the altitude range 100-120 km. Solar wind electrons and protons provide a substantial source for the nightside ionosphere. These calculations are carried out for a solar minimum period using solar wind electron flux, photon flux, neutral densities, and temperatures under nearly the same areophysical conditions as the measurements.     

The calculated and observed diurnal variation of the ionosphere over Millstone Hill on 23–24 March 1970

A numerical model of current F-region theory is use to calculate the diurnal variation of the mid-latitude ionospheric F-region over Millstone Hill on 23–24 March 1970, during quiet geomagnetic conditions. From the solar EUV flux, the model calculates at each altitude and time step primary photoelectron spectra and ionization rates of various ion species. The photoelectron transport equation is solved for the secondary ionization rates, photoelectron spectra, and various airglow excitation rates. Five ion continuity equations that include the effects of transport by diffusion, magnetospheric-ionospheric plasma transport, electric fields, and neutral winds are solved for the ion composition and electron density. The electron and ion temperatures are also calculated using the heating rates determined from chemical reactions, photoelectron collisions, and magnetospheric-ionospheric energy transport. The calculations are performed for a diurnal cycle considering a stationary field tube co-rotating with the Earth; only the vertical plasma drift caused by electric fields perpendicular to the geomagnetic field line is allowed but not the horizontal drift. The boundary conditions used in the model are determined from the incoherent scatter radar measurements of T_e, T_i and O⁺ flux at 800km over Millstone Hill (Evans, 1971a). The component of the neutral thermospheric winds along the geomagnetic field has an important influence on the overall ionospheric structure. It is determined from a separate dynamic model of the neutral thermosphere, using incoherent scatter radar measurements.The calculated diurnal variation of the ionospheric structure agrees well with the values measured by the incoherent scatter radar when certain restrictions are placed on the solar EUV flux and model neutral atmospheric compositions. Namely, the solar EUV fluxes of Hinteregger (1970) are doubled and an atomic oxygen concentration of at least $\text{10}{}^{11}\text{cm}{}^3$ at 120 km is required for the neutral model atmosphere. Calculations also show that the topside thermal structure of the ionosphere is primarily maintained by a flow of heat from the magnetosphere and the night-time F2-region is maintained in part by neutral winds, diffusion, electric fields, and plasma flow from the magnetosphere. The problem of maintaining the calculated night-time ionosphere at the observed values is also discussed.     

Scale-dependent infrared radiative damping rates on Mars and their role in the deposition of gravity-wave momentum flux

Using a Curtis-matrix model of 15 μm CO₂ radiative cooling rates for the martian atmosphere, we have computed vertical scale-dependent IR radiative damping rates from 0 to 200 km altitude over a broad band of vertical wavenumbers ∣m∣ = 2π(1-500 km)⁻¹ for representative meteorological conditions at 40°N and average levels of solar activity and dust loading. In the middle atmosphere, infrared (IR) radiative damping rates increase with decreasing vertical scale and peak in excess of 30 days⁻¹ at ∼50-80 km altitude, before gradually transitioning to scale-independent rates above ∼100 km due to breakdown of local thermodynamic equilibrium. We incorporate these computed IR radiative damping rates into a linear anelastic gravity-wave model to assess the impact of IR radiative damping, relative to wave breaking and molecular viscosity, in the dissipation of gravity-wave momentum flux. The model results indicate that IR radiative damping is the dominant process in dissipating gravity-wave momentum fluxes at ∼0-50 km altitude, and is the dominant process at all altitudes for gravity waves with vertical wavelengths ?10-15 km. Wave breaking becomes dominant at higher altitudes only for “fast” waves of short horizontal and long vertical wavelengths. Molecular viscosity plays a negligible role in overall momentum flux deposition. Our results provide compelling evidence that IR radiative damping is a major, and often dominant physical process controlling the dissipation of gravity-wave momentum fluxes on Mars, and therefore should be incorporated into future parameterizations of gravity-wave drag within Mars GCMs. Lookup tables for doing so, based on the current computations, are provided.     

The penetration of soft electrons into the ionosphere

Calculations are presented of energy spectra and angular and spatial distributions of electron fluxes in the ionosphere resulting from precipitation ofmonoenergetic (E = 25, 50 and 100 eV) electrons. The incident electrons are assumed to be isotropic over the downward direction. It is found that the resulting steady-state electron fluxes above ca. 300 km are highly anisotropic, and that the pitch angle distribution is energy dependent. About 15 per cent of the incident electrons are backscattered elastically to the protonosphere. A much larger number of electrons escape after they have deposited a part of their energy in the atmosphere. The mean energy of the escaping electrons is about half that of the incident electrons. About 50% of the incident energy is absorbed in the atmosphere, the remainder being returned to the protonosphere. The rate of absorption of energy is a maximum at heights between 300 and 400 km. Most of the energy is absorbed in ionization and excitation of atomic oxygen. An appreciable amount of energy is, however, absorbed as heat by the ambient electron gas. Altitude profiles are presented of the rates of ionization, excitation, and electron heating caused by soft electron precipitation.     

Latitude variations of exospheric hydrogen and the polar wind

Several satellite experiments have measured the solar Lyman-α line, either in scattering from upper atmospheric atomic hydrogen (the Lyman-α airglow) or directly at line center (which determines the hydrogen column density along the line of sight). Recent analyses of data from the above experiments consistently reveal the presence of an atomic hydrogen depletion at high latitudes. In situ determinations of hydrogen at lower altitude show no evidence of such behaviour. This has led us to postulate two mechanisms which may be more effective in reducing the high-latitude density at the high altitudes of the exospheric measurements (500–2000 km). The first is the polar wind loss of protons, which depletes atomic hydrogen through a charge exchange reaction. The second is a high-latitude magnetospheric heating of protons, followed by charge exchange. Opposing the above loss mechanisms are the influences of ballistic lateral flow and mean meriodional winds. We have shown by means of a three-dimensional exospheric transport model that none of the above mechanisms can reconcile the disparate results in the two altitude regimes, nor can they provide the large outward hydrogen fluxes and the correct seasonal variations observed at high latitudes.     

Nighttime proton fluxes at Millstone Hill

Incoherent scatter measurements of electron density and vertical O⁺ fluxes over Millstone Hill (42.6°N, 71.5°W) previously have been used to study the exchange of plasma between the ionosphere and the magnetosphere. During the daytime there is usually an upward flux of O⁺ ions above about 450 km that can be measured readily and equated to the escaping proton flux. At night the O⁺ fluxes usually are downwards everywhere owing to the decay of the F-layer, and it becomes difficult to detect effects due an arriving proton flux. In a new study of the nighttime fluxes, appeal was made to the estimated abundance of the H⁺ ions in the upper F-region which can be extracted from the observations. From a study of the behavior on 25 days over the interval 1969–1973, we conclude that in the daytime the flux always is upwards and close to its limiting value. This situation persists throughout the night in summer at times of high sunspot activity (e.g., 1969). There is a period of downward flux prior to ionospheric sunrise on winter nights whose duration increases with decreasing sunspot number. As sunspot minimum is approached (e.g., in 1973) downward fluxes are encountered for a brief period prior to ionospheric sunrise in summer also. Thus, over most parts of sunspot cycle, it appears that the protonosphere supplies ionization to the winter night ionosphere, while being maintained from the summer hemisphere. This helps explain the smallness of the day-to-night variations reported for the electron content of magnetospheric flux tubes near L = 4 in the American sector.     

The formation of high energy electron fluxes in the inner radiation belt of earth

Measurements on board the low altitude polar orbiting satellite Intercosmos-17 /nearly circular orbit h = 500 km, i = 83.5°/ have shown relatively high fluxes of high energy electrons /E_e > 100 MeV and E_e > 300 MeV respeetively/ at minimum-B-equator. Computation of the electron production spectra assuming the interaction of high energy protons of the inner radiation belt with residual atmosphere is made. The considered mechanism can explain the enhanced flux of high energy electrons registered in the Brazil magnetic anomaly.     

Energy deposition and primary chemical products in Titan’s upper atmosphere

Cassini results indicate that solar photons dominate energy deposition in Titan’s upper atmosphere. These dissociate and ionize nitrogen and methane and drive the subsequent complex organic chemistry. The improved constraints on the atmospheric composition from Cassini measurements demand greater precision in the photochemical modeling. Therefore, in order to quantify the role of solar radiation in the primary chemical production, we have performed detailed calculations for the energy deposition of photons and photoelectrons in the atmosphere of Titan and we validate our results with the Cassini measurements for the electron fluxes and the EUV/FUV emissions. We use high-resolution cross sections for the neutral photodissociation of N₂, which we present here, and show that they provide a different picture of energy deposition compared to results based on low-resolution cross sections. Furthermore, we introduce a simple model for the energy degradation of photoelectrons based on the local deposition approximation and show that our results are in agreement with detailed calculations including transport, in the altitude region below 1200 km, where the effects of transport are negligible. Our calculated, daytime, electron fluxes are in good agreement with the measured fluxes by the Cassini Plasma Spectrometer (CAPS), and the same holds for the measured FUV emissions by the Ultraviolet Imaging Spectrometer (UVIS). Finally, we present the vertical production profiles of radicals and ions originating from the interaction of photons and electrons with the main components of Titan’s atmosphere, along with the column integrated production rates at different solar zenith angles. These can be used as basis for any further photochemical calculations.     

The topside ionosphere above Arecibo at equinox during sunspot maximum

The coupled time-dependent O⁺ and H⁺ continuity and momentum equations and O⁺, H⁺ and electron heat balance equations are solved simultaneously within the L = 1.4 (Arecibo) magnetic flux tube between an altitude of 120 km and the equatorial plane. The results of the calculations are used in a study of the topside ionosphere above Arecibo at equinox during sunspot maximum. Magnetically quiet conditions are assumed.The results of the calculations show that the L = 1.4 magnetic flux tube becomes saturated from an arbitrary state within 2–3 days. During the day the ion content of the magnetic flux tube consists mainly of O⁺ whereas O⁺ and H⁺ are both important during the night. There is an altitude region in the topside ionosphere during the day where ion-counterstreaming occurs with H⁺ flowing downward and O⁺ flowing upward. The conditions causing this ion-counterstreaming are discussed. There is a net chemical gain of H⁺ at the higher altitudes. This H⁺ diffuses both upwards and downwards whilst O⁺ diffuses upwards from its solar e.u.v. production source which is most important at the lower altitudes. During the night the calculated O⁺ and H⁺ temperatures are very nearly equal whereas during the day there are occasions when the H⁺ temperature exceeds the O⁺ temperature by about 300 K.