Abstract. On board the four Cluster spacecraft, the Cluster Ion Spectrometry (CIS) experiment measures the full, threedimensional ion distribution of the major magnetospheric ions (H + , He + , He ++ , and O + ) from the thermal energies to about 40 keV/e. The experiment consists of two different instruments: a COmposition and DIstribution Function analyser (CIS1/CODIF), giving the mass per charge composition with medium (22.5 • ) angular resolution, and a Hot Ion AnalCorrespondence to: H. Rème (Henri.Reme@cesr.fr) yser (CIS2/HIA), which does not offer mass resolution but has a better angular resolution (5.6 • ) that is adequate for ion beam and solar wind measurements. Each analyser has two different sensitivities in order to increase the dynamic range.

Using four months of tail data obtained by the three‐dimensional plasma instrument on board the AMPTE/IRM satellite in 1986, we have done a statistical survey on the behavior of ion and electron moments in the central plasma sheet. Almost 80,000 spin averages of plasma density, ion bulk velocity, ion and electron temperature, and plasma β were analyzed with respect to differences between their values in the inner and outer central plasma sheet as well as their dependence on magnetic activity. The ion temperature increases with increasing magnetic activity while the ion density decreases during disturbed intervals, except in the neutral sheet neighborhood at smaller radial distances. The ion and electron temperatures in the central plasma sheet are highly correlated, with Ti/Te being constant over a wide range of temperatures and about twice as large as in the distant tail. The average ion flow speeds in the central plasma sheet are below 100 km/s and nearly identical to those found in the plasma sheet boundary layer, although the distribution functions usually are quite different. High‐speed flows do occur, but in bursts of most often less than 1 min duration with intermittent intervals of nearly stagnant plasma. The distribution of flow directions strongly favors sunward flow for velocities above 300 km/s, indicating that a near‐earth neutral line is rarely, if ever, located inside of XGSM = −19 RE.

Eleven passes of the ISEE satellites through the frontside terrestrial magnetopause (local time 9 -'17 h; GSM latitude 2 0 -43 0 N) have been identified, where the plasma velocity in the magnetopause and boundary laver was substantially larger than in the magnetosheath. This paper examines the nature of°the plasma flow, magnetic field, and energeticparticle fluxes in these regions, with a view to determining whether the velocity enhancements can be explained by magnetic-field reconnection.

Further Heos 2 plasma and magnetic field data obtained in the frontside boundary layers of the magnetosphere are presented. They reveal that the low‐latitude extension of the entry layer is of a somewhat different nature. The most pronounced difference with respect to the entry layer in the cusp region is the substantial density jump at the magnetopause. Furthermore, the low‐latitude boundary layer tends to be thinner and less turbulent, and the flow velocity inside the layer is always lower than that of the adjacent magnetosheath. This observation excludes large‐scale reconnection at the front of the magnetosphere as the origin of the layer. It is suggested that diffusive entry of magnetosheath plasma and/or heating of detached plasma from the plasmasphere leads to the formation of the layer. It appears likely that reconnection is dominantly occurring as a transient process in the cusp region and accompanies the eddy convection inside the entry layer. As a consequence, magnetic flux is being eroded from the front of the magnetosphere. This is in agreement with the signature of short‐term large‐amplitude magnetic perturbations observed in the low‐latitude boundary layer.

No abstract

A total of 21 passes of AMPTE/IRM through the dayside (0800–1600 hours local time) low‐latitude magnetospheric boundary region have been examined, all of which were characterized by large magnetic shear across the magnetopause. The purpose of the study was to use the improved accuracy and time resolution of the IRM plasma measurements to reexamine the occurrence of high‐speed flows as signatures of the reconnection process at the magnetopause. A total of 12 of the 21 passes showed magnetopause crossings with high‐speed flows. The duration of these flows was sometimes as short as 10 s and rarely more than 30 s. The occurrence of high‐speed flows is inversely correlated with β, the ratio of plasma to magnetic pressure. For β < 2 the ratio, ΔV*, of observed to theoretically predicted flow velocity changes is high (ΔV* = 0.74 on average), while for β > 2 the average ratio is low (ΔV* = 0.18). We tentatively interpret this to indicate that reconnection may occur preferentially for low β values. A total of nine high‐speed flows were subjected to quantitative momentum and energy balance test. The agreement between observations and theoretical predictions was generally quite good. The signs of the normal magnetic field, Bn, deduced from the momentum balance agreed with the sense of magnetic connection inferred independently from the measured proton heat flux direction. In one case the observations suggest that the satellite may have traversed the diffusion region of the reconnection process. The momentum balance results indicate that at times the effective mass of the ions in the magnetopause and boundary layer was of the order of twice the proton mass.

A survey of two-dimensional electron velocity distributions, f(¾), measured near the earth's bow shock using Los Alamos/Garching plasma instrumentation aboard ISEE 2 is presented. This survey provides clues to the mechanisms of electron thermalization within the shock and the relaxation of both the upstream and downstream velocity distributions. First, near the foreshock boundary, fluxes of electrons having a power law shape at high energies backstream from the shock. Although most often they appear as a monotonically decreasing extension of solar wind distributions in the backward hemisphere along the magnetic field direction, /•, they occasionally appear as a resolved peak in energy. Within the interior of the foreshock, in addition to the hot, isotropic electrons at higher energies, field-aligned depressions in oe (V) are observed at the lowest energies (E • 15 eV) and twin angular peaks centered on/• are observed at intermediate energies (15 eV • E • 45 eV). Such distributions are associated closely with 1-Hz whistler waves. Second, within the shock, cuts through oe(V) along/•, oe(V•), often show single maxima offset toward the magnetosheath by speeds comparable to, but larger than, the upstream thermal speed. When sequences of such distributions ar• observed in a single shock transition, offset speeds increase and peak heights of f(V•) decrease with increasing penetration toward the downstream (magnetosheath) side. Third, magnetosheath distributions generally have flat tops out to an energy, Eo, with maxima substantially lower than that in the solar wind. Occasionally, cuts through f(¾) along/• show one and sometimes two small peaks at the edge of the flat tops making them appear concave upward. The magnetosheath distributions often have strong angular anisotropies which depend on energy. For energies less than Eo, f(V•) > f(V•_) at constant E, whereas for E > Eo, f(V•) < f(V•_). The electron distributions characteristic of these three regions are interpreted as arising from the effects of macroscopic (scale size comparable to or larger than the shock width) electric and magnetic fields and the subsequent effects of microscopic (scale size small in comparison with the shock width) fields. In particular, our results suggest that field-aligned instabilities are likely to be present in the earth's bow shock. 1. INTRODUCTION Much work has been done over the past two decades to understand the physics of collisionless shocks (see e.g., Sagdeer and Galeev [1969], Tidman and Krall [1971], Biskamp [1973], and Galeev [1976]). For any shock above a critical Mach number, Mc, some type of dissipation is necessary to provide a transition between upstream and downstream parameters and satisfy the Rankine-Hugoniot relations [Tidman and Krall , 1971]. Since a collisionless shock has, by definition, a width much less than a mean free path, the dissipation cannot be due to classical collisions. The usual mechanism which is invoked for heating electrons is that of an anomalous, or more precisely, wave-particle dissipation due t...

A comprehensive set of experimental observations of a high β (2.4), supercritical (Mf = 3.8), quasi‐perpendicular (ΘBn1 ∼ 76°) bow shock layer is presented, and its local geometry, spatial scales, and stationarity are assessed in a self‐consistent, Rankine‐Hugoniot‐constrained frame of reference. Included are spatial profiles of the ac and dc magnetic and electric fields, electron and proton fluid velocities, current densities, electron and proton number densities, temperatures, pressures, and partial densities of the reflected protons. The transformation of the apparent time scales to the actual spatial scales is performed with unprecedented accuracy. The observed layer profile is shown to be nearly phase standing and one dimensional in a Rankine‐Hugoniot frame, empirically determined by the magnetofluid parameters outside the layer proper. One or both of these properties appear to collapse at the time resolution of 1.5 s in the specific geometry considered in this study. Several pieces of evidence are used to show this stationarity: (1) the similarity of the average magnetic structures seen on the two ISEE spacecraft; (2) the close agreement between the electric currents directly determined from the plasma data and those inferred from the magnetic data assuming the layer is one dimensional and time stationary; (3) the close agreement of the empirically determined scale lengths of the most prominent substructures with those determined by numerical simulations and previous laboratory studies; and (4) the close agreement between the theoretical Rankine‐Hugoniot‐determined normal plasma pressure jump and that of the observed electron and proton fluids. The resolved cross‐field electrical currents (with empirical error estimates) are observed to peak within the main magnetic ramp at a level well below the first stabilization threshold for ion acoustic turbulence suggested for low β shocks by Galeev (1976); clear evidence is also provided for smaller parallel currents throughout the main ramp and overshoot, with a predominant sense as if the shock electric field has caused the lighter electrons to lead the ions along the local magnetic field direction. The width of the shock depends on what structures are used to define it. The upstream pedestal or “foot” is nearly two upstream ion skin depths wide, but the main magnetic ramp is only 1/5 the upstream ion skin depth and thus considerably smaller than “conventional wisdom” and most simulations. The ramp scale length is directly corroborated by the current densities determined from the plasma instruments.