Strong Pitch Angle Diffusion Research Articles

Correction to “Particle drift and loss rates under strong pitch angle diffusion in Dungey's model magnetosphere” by Michael Schulz

Correction to “Particle drift and loss rates under strong pitch angle diffusion in Dungey's model magnetosphere” by Michael Schulz

Observations of black auroral patches and of their relationship to other types of aurora

A morphological study of black patches and newly reported black rings (black patches with islands of embedded diffuse aurora) was conducted using an archive of low‐light level all‐sky and narrow‐field television camera data. The overwhelming majority of black patches and rings drifted eastward, poleward of westward drifting pulsating aurora in the early recovery phase of the auroral substorm. Occasionally, black aurora would overlap with pulsating aurora. Sometimes the pulsations would add to the intensity in the black aurora more than in the surrounding diffuse aurora (called veiling). One example was documented where the pulsation was purely additive, adding to the intensity of the black aurora and the surrounding diffuse aurora equally. Cases of total veiling suggest an intensity limit, possibly associated with the strong pitch angle diffusion limit.

Particle drift and loss rates under strong pitch angle diffusion in Dungey's model magnetosphere

Strong pitch angle diffusion so thoroughly randomizes the first two adiabatic invariants of charged‐particle motion that it conserves (as a first approximation) the surrounded phase space volume Λ ≡ p3Ψ0, where Ψ0 = ∮(ds/B) is the flux tube volume per unit magnetic flux and p is the common scalar momentum of the particles involved. (This conservation law is widely used in plasma physics. Its nonrelativistic form is usually derived from the adiabatic gas law but is not really a “fluid” result. Since the conservation law does not couple particles with different energies in the same flux tube, such particles can have different values of Λ.) Here the flux tube volume Ψ is computed for selected field lines in Dungey's model magnetosphere (dipole moment μE ≈ −30.2 μT−RE3 plus uniform southward ΔBz ≈ −9.0 COS6Λ* μT, Λ* being the invariant latitude of the quiet time auroral oval, which maps outward to a circular neutral line at radial distance b = (μE/ΔBz)1/3 in the equatorial plane). The integral ∮(ds/B) is fitted within 0.2% at all L values by the expression Ψ0 ≈ (L4a4/μE){(32/35) − [2.045 + 1.045(r0/b)3 + 0.095(r0/b)6 + 0.075(r0/b)9] ln [1−(r0/b)3]}, where r0 (equatorial radius of the field line) is obtained by solving the cubic equation (r0/b) = (La/b)[1 + (1/2)(r0/b)3]. The particle lifetime τ against strong pitch angle diffusion is given in turn by τ ≈ [4ΨBnBs/(Bn + Bs)(1 – η)](m/p), where Bn and Bs (which can differ in an offset‐dipole field model) are the field intensities at the northern and southern foot points of the field line, m is the relativistic mass, and η is the backscatter coefficient. The same expression for Ψ thus enters both the quasi‐adiabatic Hamiltonian (to whose derivative with respect to 1/L the ensemble‐averaged gradient‐curvature drift rate is directly proportional) and the particle loss rate 1/τ (by which the phase space density ƒ = J4π/4πp2 is exponentially attenuated along the ensemble averaged drift trajectory). This construction provides a means of numerically simulating kinematical aspects of electron motion in the outer magnetosphere and precipitation in the diffuse aurora, as well as ion motion in part of the plasma sheet (regarded as a source of the ring current). Its use in magnetospheric plasma simulations should save computer time and reveal analytical structure relevant to magnetospheric dynamics.

Electron temperature profiles in the magnetosphere of Neptune

The electron temperature in the magnetosphere of Neptune decreases with decreasing radial distance over the same distance range over which a similar decrease of the proton temperature has been explained as resulting from interaction of protons with a cloud of neutral hydrogen. The effect on the electron temperature of the interaction with this cloud is studied, taking into account photoionization, electron impact ionization, radiation, and electron-ion Coulomb collisions. It is found that the electron temperature should decrease with decreasing distance, in agreement with observations, but that at the outer boundary the theoretically expected electron temperatures are considerably lower than what is observed, implying a heating process or a source of hot electrons which have not yet been unambiguously identified. In the inner region, where the adiabatic increase of the proton temperature has been taken as implying the absence of any significant density of neutral atoms, the electron temperature remains essentially constant, implying a powerful cooling process that acts only on electrons. This process is also at present an object of speculation; as a possibility, thermal coupling to the ionosphere via strong pitch angle diffusion of electrons is suggested.

Plasma wave characteristics of the Jovian magnetopause boundary layer: Relationship to the Jovian aurora?

The Jovian magnetopause boundary layer (BL) plasma wave spectra from 10−3 to 102 Hz have been measured for the first time. For one intense event the magnetic (B′) and electric (E′) spectra were 2 × 10−4 ƒ2.4 nT2/Hz and 4 × 10−9 ƒ2.4 V2/m2 Hz, respectively. Although no measurable wave amplitudes were detected above the electron gyrofrequency, ∼140 Hz, this finding may be due to the low signal strength characteristic of this region. The B′/E′ ratio is relatively frequency independent. It is possible that waves are obliquely propagating whistler mode waves. The B′ and E′ spectra are broadband with no obvious spectral peaks. The waves are sufficiently intense to cause cross‐field diffusion of magnetosheath plasma to create the BL itself. A Jovian BL thickness of 10,700 km is predicted, which is consistent with past in situ measurements. The Jovian boundary layer wave properties are quite similar to the BL waves at Earth (however, the Jovian waves are orders of magnitude less intense). It appears that the solar wind/magnetosphere dynamos at the two planets are similar enough to be consistent with a common wave generation mechanism. The predicted ionospheric latitudinal width of the BL of ∼100–200 km is quite similar to the Jovian auroral high‐latitude ring measured by Hubble. The location of the BL at and inside the foot point of the last closed field line may place the boundary layer and the aurora on approximately the same magnetic field lines. The Jovian BL waves are sufficiently intense to cause strong pitch angle diffusion for <5‐keV electrons and 1‐keV to 1‐MeV protons. The estimated energy precipitation rate from this interaction <1 erg cm−2 s−1, sufficient for a weak high‐latitude auroral ring. This intensity is 2 to 3 orders of magnitude too low to cause the main aurora ring, however. If it is found that this main aurora maps into the boundary layer, then other mechanisms such as (ionospheric) double layers must be responsible for the particle energization and precipitation.

One‐ and two‐dimensional simulations of whistler mode waves in an anisotropic plasma

We present results from self‐consistent, one‐ and two‐dimensional, electromagnetic simulations of the electron whistler mode instability relevant to the near‐Earth nightside plasma sheet region during geomagnetically disturbed times. Specifically, we study the evolution of energetic, anisotropic (T⊥ > T∥) electron distributions that are injected into the nightside ring current region at geomagnetically disturbed times, the resulting growth of electron whistler mode waves, and subsequent electron pitch angle diffusion via electron whistler wave‐particle interactions. Growth of whistler mode waves from an initial pitch angle anisotropy (T⊥ ≈ 4 T∥) is studied in the strong pitch angle diffusion regime (defined as having scattering times much shorter than a typical electron bounce time in the near Earth's dipolar field). The quasi‐linear and subsequent nonlinear evolution of waves and the corresponding migration of electrons in velocity space is followed over timescales such that ion motion may be neglected. Our simulations contain wave frequencies and growth rates that are a significant fraction of the electron gyrofrequency (ω ∼ 0.5ωce, γ ∼ 0.1ωce being typical) and the simultaneous evolution of waves propagating both parallel and nonparallel to the ambient magnetic field direction. Effects due to these are not usually accounted for in applications of quasi‐linear theory to the problem of electron whistler wave‐particle interactions, so that our self‐consistent simulations of the electron whistler instability provide an important insight into the applicability of quasi‐linear theory to the velocity space diffusion of electrons due to electron whistler wave‐particle interactions. We examine the dependence of whistler mode wave growth rates, nonlinear wave mode saturation, and pitch angle diffusion rates on the β value of the hot electron species which contains the resonant population, and we compare the differences between results of one‐ and two‐dimensional simulations. In general, we have found significantly larger average growth rates in a one‐dimensional than in a two‐dimensional geometry, by up to a factor of ∼2–3, with the difference between such growth rates becoming larger as β increases. As a result, we find pitch angle diffusion rates are significantly larger (by up to a factor of ∼10) in a one‐dimensional geometry, and pitch angle diffusion rates increase as β increases. During the early wave growth period of the instability, pitch angle diffusion rates Dα have been found to scale to magnetic wave energy Bw², approximately Dα ∝ Bw². We also show that in the self‐consistently evolved system (containing several wave modes), the pitch angle diffusion is still consistent with a usual quasi‐linear approach in which resonant particles are taken to diffuse along surfaces of constant wave frame energy.

The Uranian aurora and its relationship to the magnetosphere

About 32 h of Voyager Ultraviolet Spectrometer (UVS) observations of Uranus H2 band airglow emission (875 ≤ λ ≤ 1115 Å) have been analyzed using the singular value decomposition (SVD) approach to inversion, producing an intensity map showing aurora at both magnetic poles. An H Lyman α aurora may also be present but is difficult to separate from scattered solar and local interstellar medium components. SVD analysis of variance shows that the intensity estimate is significantly larger than the error estimate over both Uranographic poles and part of the equatorial region, fortuitously including both magnetic polar regions. The Goddard Space Flight Center Q3 magnetic field model correctly predicts that the aurora should be larger in area and emit more power at the weaker N magnetic pole than at the stronger S magnetic pole. However, the auroral emissions are quite localized in magnetic longitude and so do not form complete auroral ovals. The brightest auroral emission at each magnetic pole is confined to a range of ≈ 90° of magnetic longitude centered on the magnetotail direction, at moderate magnetic L parameter (5 ≤ L ≤ 10), but some emission at each pole is distributed over a range of more than 180° of longitude. The S polar auroral intensity maximum is coincident with the source of the broadband bursty and broadband smooth Uranian kilometric radio emission (UKR), while the N polar auroral intensity maximum may coincide with the dayside UKR source. The N and S auroral intensity maxima also lie at the conjugate magnetic footprints of the maximum intensities of whistler‐mode plasma wave emission and 22‐ to 35‐keV electron fluxes observed by Voyager. The magnetic longitudes of the aurora are completely inconsistent with the “windshield wiper” effect for either ions or electrons, indicating that some other effect, such as rapid depletion of the population of precipitating particles or highly localized strong pitch‐angle diffusion, may be acting to localize emission. The low apparent L of the precipitating particles indicates that their energies may be ≤ 10 keV. Hence magnetospheric convection is likely to be important, and thus particles exciting the aurora may not remain on constant L shells. The precipitating particles may be a relatively low‐energy population at high L that is heated to aurora‐exciting energy by adiabatic compression during convection to low L. We estimate that the total auroral power output at H Lyman α and shorter wavelengths is about 3 × 109 to 7 × 109 W, requiring about 10 times that much power for excitation.

Stimulated auroral precipitation by wave‐particle interactions: Implications for the CRRES satellite mission

The interaction of charged particles with plasma waves is one of the mechanisms which can cause the precipitation of energetic particles from the Earth's radiation belts. In the inner plasma sheet of the magnetotail, electromagnetic cyclotron waves can resonate with the bulk of the ambient energetic electron distribution only if an additional cold dense plasma population is present. In an attempt to artificially stimulate auroral electron precipitation, one experiment to be carried out during the Combined Release and Radiation Effects Satellite (CRRES) mission will be to release lithium close to the Earth's equatorial plane near geosynchronous orbit, along field lines that map to the diffuse auroral region. As the lithium is photoionized, a dense cold plasma cloud will be created within the background energetic population causing increased whistler wave growth. To support the CRRES Stimulated Electron Precipitation (STEP) experiment theoretically, a quantitative study is presented examining wave growth, propagation, and trapping of electromagnetic whistler waves within the released cold plasma cloud. Results show that for a background thermal electron temperature anisotropy of T⊥/T∥ > 1.2, whistler wave growth rates are greatly increased by the presence of the cold electrons, and the waves can be trapped in the cloud. Because of a maserlike effect, intense standing waves with amplitudes the order of a few nanotesla will be generated within a frequency range of about 50–500 Hertz. The waves will reach their largest amplitudes about 20 s after release, causing strong electron pitch angle diffusion into the loss cone, which can be detected as enhanced diffuse aurora at Earth.

Consequences of strong pitch-angle diffusion of particles in solar flares

The consequences of strong pitch-angle diffusion of energetic electrons and ions during the impulsive phase of a solar flare have been analyzed. Strong diffusion suggests that the regions with high levels of small-scale turbulence (turbulent mirrors) are formed in the magnetic trap. These additional mirrors scatter the energetic particles, and therefore, the propagation velocity of the particles toward the footpoints decreases. It has been shown that the strong diffusion of energetic particles by waves is realized very easily in a flare loop. The main consequences of strong pitch-angle diffusion are as follows: (1) the turbulent propagation time for greater than about 10 MeV ions is about one order of magnitude longer than for greater than about 30 keV electrons, (2) the energetic particle spectra in a flare loop reflect the source particle spectrum, and (3) the energetic particle density is constant along the flare loop axis. The strong diffusion approach makes it possible to explain, for example, the time delays in hard X-ray and gamma-ray emission and the properties of particle fluxes in the interplanetary space, and it offers new possibilities for flare plasma diagnostics. 17 refs.

Energetic Particles in a Flare Loop: Spectra and Radiation Signatures

It has been shown that high energy particle spectra, particle dynamics, and radiation in a flare loop are determined by wave-particle interactions. The electron-whistler interaction occurs under conditions of strong pitch angle diffusion that makes the particle distribution function isotropic. The flare loop electrons retain information about the particle source spectrum. The interaction of energetic ions with Alfven waves is characterized by strong, moderate, and weak diffusion. The time delays in hard X-ray and gamma-ray emission during one-step acceleration processes might be understood in terms of a trap-plus-turbulent propagation model. The density of precipitating particles is less than or equal to the trapping one. Radiation signatures of flare loop electrons are discussed.

Radial diffusion and losses of energetic protons in the 5 to 12 RS region of Saturn's magnetosphere

Phase space density profiles for protons with first invariants µ = 600 and 3000 MeV/G (integral invariants K = 0.30 and 0.003 G½ RS) previously derived from measurements by the University of Iowa detector G on Pioneer 11 during the 1979 Saturn encounter are analyzed using solutions of the time‐averaged radial diffusion equation in a dipolar planetary magnetic field. A series of loss models ranging from satellite absorption only to satellite plus ring E absorption and added distributed losses is assumed and the corresponding form of the time‐averaged radial diffusion coefficient D(L) (taken to be of the form D(L) = DoLn, where n is an integer) is determined in each case by a minimum‐variance fit to the data‐derived phase space density profiles. The inferred forms for D(L) are characterized by a low‐order L‐dependence (∼L³–L6) and a relatively high amplitude (Do ≃ 10−9–10−8 RS² s−1) in approximate agreement with earlier studies. Loss models that include only satellite absorption result in pronounced Dione macrosignatures in the model phase space density profiles that are not present in the data profiles, indicating a need for additional distributed losses. Absorption by ring E would be sufficient to explain the absence of a Dione macrosignature only if both the presently accepted maximum normal optical depth and mean ring particle size are increased by at least 1 order of magnitude. Charge exchange losses, estimated from observational upper limits on the concentrations of relevant neutral species, also appear to be only marginally important. Pitch angle scattering precipitation losses occurring within the orbit of Rhea at a rate of at least 0.01 times that of the strong pitch angle diffusion limit could in principle provide the necessary distributed losses.

Energy spectra of the major ion species in the ring current during geomagnetic storms

Using the University of Maryland/Max‐Planck‐Institut für Aeronomie charge‐energy‐mass (CHEM) spectrometer on the AMPTE Charge Composition Explorer (CCE) spacecraft, we have examined the nearly equatorial storm time energy spectra of four major magnetospheric ions, H+, O+, He+, and He++, over the energy range 1–300 keV/e in the L range 3–6. The data were obtained during the main and early recovery phases of all geomagnetic storms with minimum Dst less than −50 nT in the time period September 1984 to November 1985. When the spectra are organized by local time, certain features emerge. In particular, there is a dip in the spectra of all ions at 5–20 keV/e in the dawn‐to‐noon sector, while in the noon‐to‐dusk sector the proton phase space density drops off sharply below ∼5 keV. We have compared these spectra with those predicted by a model of ion drift and loss in the magnetosphere. The model calculates the drift paths in a Volland‐Stern electric field and dipole magnetic field and determines the losses due to charge exchange and strong pitch angle diffusion along the paths. We find that the spectra are most consistent with a Volland‐Stern electric field with γ = 2 and with a rotation of the nominal dawn‐to‐dusk electric field eastward by 2 hours local time. Charge exchange is found to be the dominant loss process during the main phase of the storm, producing qualitative agreement with the observed spectra for all species. There are some quantitative disagreements, particularly in the prenoon sector, which may be explained either by an additional loss process or by a modified drift model.

On the origin of time delays in hard X-ray and gamma-ray emission of solar flares

The possibility of a strong pitch-angle diffusion regime as well as of turbulent propagation of energetic ions and electrons in flaring loops has been shown. The strong diffusion regime suggests that two regions with a high level of small-scale turbulence are formed in the magnetic trap. Such additional ‘turbulent mirrors’ scatter energetic particles and, therefore, the flux of precipitating particles decreases and the mean lifetime of electrons and protons in a flaring loop grows. We cannot rule out that the turbulent propagation of the particles can be responsible for the energy dependence of hard X-ray delays as well as the time lag of the gamma-ray line peaks with respect to the hard X-ray peaks as the electrons and ions are accelerated simultaneously. The trap plus turbulent propagation model may also explain the lack of abundant population of 10–100 keV electrons in interplanetary space in proton-rich events and offers new possibilities for flare plasma diagnostics.

Energetic electron precipitation during a magnetospheric substorm and its relationship to wave particle interaction

Energetic electron precipitation was observed via X ray measurements from three simultaneously flown balloons in a region conjugate to the location of the geosynchronous satellite GEOS 2. Data from the particle, wave, and field experiments on this satellite were used to investigate the relationship between magnetospheric processes in the equatorial plane and the electron precipitation. A large part of the electron precipitation during a substorm on July 3, 1979, was accompanied by VLF waves at whistler frequencies and had a temporal structure grossly similar to the intensity variations of the equatorial electron fluxes. Some precipitation also occurred in the absence of VLF waves. During the wave associated precipitation the equatorial electrons were anisotropically distributed with maximum fluxes perpendicular to the geomagnetic field. The anisotropy was larger than the critical anisotropy and thus enabled wave growth. The growth rate itself was determined numerically from the observations and turned out to have a frequency dependence similar to that of the observed waves. The interaction of the electrons with the waves led to strong pitch angle diffusion. The temporal variations of the electron flux in the loss cone were similar to those of the electron precipitation observed with a balloon near the conjugate point to GEOS 2. It is concluded that most of the electron precipitation during this substorm was caused by wave particle interaction.

Experimental study of the relationship between energetic electrons and ELF waves observed on board GEOS: A support to quasi‐linear theory

The interaction between energetic electrons (E > 15 keV) and whistler mode ELF/VLF waves (0.2–2 kHz) is studied experimentally in the light of the quasi‐linear pitch angle diffusion theory. This work is carried out with data from the European magnetospheric satellites GEOS 1 and GEOS 2. A method is given to calculate from the data in a complete manner the particle anisotropy A and the temporal growth rate of the waves, γ, as defined by Kennel and Petschek (1966), A and γ being frequency dependent. The data are averaged over 5 min, which is slightly less than the minimum lifetime of the electrons. Therefore the temporal evolution of weak diffusion cases can be studied. The calculated anisotropy is compared to the critical anisotropy AC, which is the minimum anisotropy needed for wave growth, and to the measured wave spectrum. In the presence of waves, A is greater than AC and roughly follows AC in the frequency range of wave emission. The difference [A‐AC] is smallest during strong diffusion but remains positive, as expected from pitch angle diffusion in a steady state situation. The behavior of A does not change considerably between the geomagnetic equator (GEOS 2) and latitudes of about 20° (GEOS 1). The normalized wave growth rate γ/ωce varies between ∼10−4 and 10−2 at the equator, leading to the generation of more or less intense waves and thus to strong or weak electron pitch angle diffusion. Conversely, the constant level of γ/ωce (∼10−5) at a latitude of 20° indicates that the diffusion is most efficient at lower latitudes. A very detailed discussion in terms of the quasi‐linear theory of wave‐particle interactions shows that this theory is applicable for the explanation of the observations. The discussion includes a consideration of the conditions for the existence of a steady state emission in terms of a minimum level of the integrated spatial growth rate needed to compensate for the wave losses at the ionosphere.

Temporal evolution of whistler growth in a cold plasma injection experiment

Using a simple time dependent cold plasma density model and assuming a typical ambient radiation belt environment, we study the evolution of the whistler mode turbulence and particle precipitation in a cold plasma release experiment similar to one that may be conducted as part of the Active Magnetospheric Particle Tracer Explorers (AMPTE) program. It is known from earlier work that the release of cold lithium ions can significantly lower the critical energy Ec above which the resonant radiation belt electrons can pitch angle scatter. We study the time evolution of the one pass gain factor for a whistler wave packet and find that for parameters accessible to AMPTE type experiments the gain factor is large enough to ensure strong whistler turbulence and strong pitch angle diffusion of radiation belt particles with energies in the range between the ambient value of Ec and the reduced value of Ec. Estimates for the total power input to the ionospheric footprint of the release are of the order of an erg/cm²/s. This precipitated energy should produce a patch of visible aurora. This effect should also persist for many hours.

Are equatorial electron cyclotron waves responsible for diffuse auroral electron precipitation?

On the basis of theoretical calculations of electron diffusion coefficients and of OGO 5 data, Lyons [1974] suggested that electrostatic electron cyclotron harmonic waves had amplitudes large enough to cause the strong pitch angle diffusion of plasma sheet keV electrons and to be responsible for diffuse auroral precipitation. However, recent measurements of the wave location and amplitude performed aboard the GEOS spacecraft have brought new pieces of information challenging these conclusions. Our calculations are based on the theoretical tool developed by Lyons [1974] but take into account the recently observed wave confinement within a few degrees from the magnetic equator. Under these conditions, we avoid the numerical averaging of the pitch angle diffusion coefficient over the whole line of force, and we can derive an analytical expression of the minimum wave amplitude required to cause strong pitch angle diffusion for plasma sheet electrons. This expression has the advantage to be easily tractable in further calculations, and permits us to evaluate the dependence of the results on parameters that are not reported by observations, such as the wave number spectrum. Our results are found to differ from Lyons's [1974] predictions by a factor of about 2.5, and typically, a wave amplitude of more than 2 mV m−1 is required to put 1‐keV electrons on strong diffusion. On the other hand, on the basis of a statistical analysis of electron cyclotron wave amplitudes measured in the nightside plasmasheet by the GEOS 2 spacecraft, we show that most of the time (>91 %), this typical value of 2 mV m−1 is not reached. This result appears inconsistent with the hypothesis that diffuse auroras, which are a permanent feature of the auroral zones, are due solely to electrostatic electron cyclotron waves. This leads us to the conclusion that these waves are not the only cause of diffuse electron precipitation. It is suggested that other mechanisms involving for instance the dynamics of the ions (such as field‐aligned currents) could play an important role.

A theoretical approach to the morphology and the dynamics of diffuse auroral zones

Direct observations of the high‐latitude ionosphere have established the continuous presence of large‐scale emissions, referred to as diffuse auroras. Neglecting localized structures, such as discrete arcs, we focus upon a quantitative description of the effects of large‐scale convection on diffuse precipitation. The transport of 100 eV‐10 keV electrons from the geomagnetic tail earthward by the convection electric field, and their pitch angle diffusion into the loss cone by wave‐particle interactions are believed to be the main cause of diffuse auroral electron precipitation. A self‐consistent treatment balancing wave generation and particle diffusion appears to be still beyond the present state of the art. For the main purpose of diffuse precipitation modeling, we propose a simplified approach to this problem and test its validity against direct observations of the location and dynamics of diffuse auroral emissions. Using basically the strong pitch angle diffusion limit in the way proposed by Kennel [1969], we derive a set of fluid equations describing the three‐dimensional transport of plasma‐sheet electrons. Their integration provides the latitude and local time distribution of precipitation fluxes and characteristic energies at the top of the ionosphere as a function of the large‐scale dawn to dusk electrostatic potential drop. The calculated expansion of the auroral oval with magnetic activity, deduced from our model, approaches the experimental results. So this simplified theoretical study permits us to reproduce and explain the main characteristics of the diffuse auroral zone. However, for high values of magnetic activity, the theoretical results are found to be shifted poleward relative to the observations. This suggests that the assumption of strong pitch angle diffusion overestimates the efficiency of wave‐particle interactions.

Satellite and ground observations of a pre‐substorm phase on May 4, 1977

On May 4, 1977, a polar orbiting low‐altitude DMSP satellite passed close to the Finnish riometer stations at the same time as an absorption bay preceding an onset of substorm was observed. The combined ground and satellite measurements show that the absorption bay was characterized by intense precipitation of energetic electrons at the inner border of the compressed plasma sheet. The precipitation occurred in a narrow band of latitude, perhaps less than 50 km wide. We suggest that the pre‐bay precipitation was due to strong pitch angle diffusion and/or scatter of energetic electrons at the inner border of the plasma sheet, probably from the outer radiation zone.

Loss cone fluxes and pitch angle diffusion at the equatorial plane during auroral radio absorption events

The flux and pitch angle distribution of energetic electrons near the loss cone have been investigated over the energy range 15–300 keV, using measurements on the geosynchronous satellite GEOS-2 at the times of auroral radio absorption events detected by riometers in Scandinavia. It is shown that conditions of strong pitch angle diffusion apply only during the most intense absorption events ( 6 dB at 30 MHz) which are relatively infrequent. During most events the loss cone is partially depleted, with the degree of depletion increasing as the absorption becomes weaker. The variation of the pitch angle diffusion coefficient with the observed radio absorption is estimated. A consequence of loss cone depletion is a tendency to overestimate the smaller events when computing the radio absorption from flux measurements in the 0°–5° range of detector pointing angles. An empirical law is derived which enables the computation of radio absorption consistent with measurements. D-region recombination laws are discussed and limits are set on the height profile of the effective recombination coefficient.