Lion roars and nonoscillatory drift mirror waves in the magnetosheath

A complete set of ISEE plasma wave, plasma, and field data are used to identify the plasma instability responsible for the generation of extremely low frequency (ELF) electromagnetic lion roars. Lion roars detected close to the magnetopause are generated by the cyclotron instability of anisotropic ( T ⊥ − / T ∥ − ≃ 1.2) thermal electrons when the local plasma critical energy, E M = B ²/8π N , falls to values ( E M ∼ 10–30 eV) close to or below the electron thermal energy, 25 eV, as a result of decreases in B . A companion theoretical paper, Thorne and Tsurutani (1981), demonstrates that the convective growth rates of lion roars under these conditions is greater than 100 dB R E −1 . The lion roars are terminated by increases in the ambient magnetic field magnitude and consequential increases in E M to values greater than 100 eV. Because there are few resonant particles at these high energies, the growth rate decreases by 3 orders of magnitude and measurable growth ceases. The value of the absolute upper limit of the frequency of unstable waves predicted by theory, ω max = A − Ω − /( A − + 1), is compared with observations. The predictions and observations are found to be in general, but not exact, agreement. Several possible explanations are explored. The quasi‐periodic, ∼20‐s magnetic and plasma oscillations which cause the variations in E M and hence alternately drive the cyclotron waves unstable and then stable are also investigated. The plasma and field pressures are shown to be out of phase, while the total pressure (electron + ion + field) remains relatively constant. Most of the pressure is associated with the particle thermal motion. The large 2∶1 variations in field strength cause large oscillations in β (8π P / B ²), from β = 1–2 at field maximum to β = 10–25 at field minimum. Analysis of the high‐resolution magnetic fields at the two closely separated spacecraft, ISEE 1 and 2, rule out the possibility that these field and plasma oscillations could be due to magnetopause motion. Cross‐correlation analyses of the magnetic fields at the two spacecraft and the time delays for maximum correlation are shown to be consistent with the magnetic structures being quasi‐static in nature. The temporal variations of the plasma and fields are due to spatial structures convecting past the spacecraft at the magnetosheath flow speed. The quasi‐periodic structures are ∼20 proton gyroradii in scale in the plasma rest frame. Magnetic structures with similar scale lengths are also shown to exist in the magnetosheaths of Jupiter and Saturn (Pioneer 11 data). The results are consistent with the interpretation that these magnetohydrodynamic structures are nonoscillatory ‘waves’ generated by the drift mirror instability. The condition for instability, β ⊥ /β ∥ > 1 + (1/β ⊥ ), is met for the cases studied in this paper. The electron and ion instabilities are intimately coupled. The generation of high β (>10), low critical energy ( E M = 10–30 eV) regions by the drift mirror instability leads to the electrons becoming cyclotron unstable. The consequential whistler mode lion roars can then be ducted by the enhanced‐density, low‐field regions. Thus lion roar durations may not represent the propagation time for an electromagnetic wave packet travelling at the group velocity, but may correspond to the convection of a magnetosheath duct (drift mirror wave) past the spacecraft. The cyclotron and drift mirror instabilities occurring in the magnetosheath are natural relaxation processes that reduce the plasma pressure anisotropies created by preferential heating of the solar wind plasma as it passes through the bow shock and the further compression that takes place as the plasma and fields approach the near‐subsolar magnetopause. One consequence of the onset of the instabilities and isotropization of the plasma is the enhanced expulsion of the plasma along field lines toward the flanks of the magnetosheath. It remains to be determined if this mechanism is a general process of ‘plasma removal’ from planetary magnetosheaths. Furthermore, the presence of nonoscillatory drift mirror waves and the convection of these structures to the magnetopause may have important consequences for magnetic merging. The alternating high and low β regions and the ( T ⊥ > T ∥ ) plasma temperature anisotropies may lead to patchy, sporadic reconnection.