Observational test of local proton cyclotron instability in the Earth's magnetosphere

We present a study of the proton cyclotron instability in the Earth's outer magnetosphere, L > 7, using Active Magnetosphere Particle Tracer Explorers/Charge Composition Explorer (AMPTE/CCE) magnetic field, ion, and plasma wave data. The analysis addresses the energy of protons that generate the waves, the ability of linear theory to predict both instability and stability, comparison of the predicted wave properties with the observed wave polarization and frequency, and the temperature anisotropy/parallel beta relation. The data were obtained during 24 intervals of electromagnetic ion cyclotron (EMIC) wave activity (active) and 24 intervals from orbits without EMIC waves (quiet). This is the same set of events used by Anderson and Fuselier [1994]. The active events are drawn from noon and dawn local times for which the wave properties are significantly different. For instability analysis, magnetospheric hot proton distributions required the use of multiple populations to analytically represent the data. Cyclotron waves are expected to limit the proton temperature anisotropy, A(p) = T-perpendicular to p/T-parallel to p - 1, according to A(p) < alpha beta(parallel to p)(c) with a similar to 1 and c similar to 0.5, where T-perpendicular to p, T-parallel to p, and beta(parallel to p) are the perpendicular and parallel proton temperatures and the proton parallel beta, respectively. During cyclotron wave events, A(p) should be close to alpha beta(parallel to p)(c) whereas in the absence of waves A(p) should be below alpha beta(parallel to p)(c). The active dawn cases yielded instability in 9 of 12 cases using the measured plasma data with an average growth rate gamma/Omega(p) = 0.025 and followed the relation A(p) = 0.85 beta(parallel to p)(-0.52). The active noon events gave instability in 10 of 12 cases, but only when an additional similar to 2 cm(-3) cold plasma was assumed. The noon wave events fell well below the dawn events in A(p)-beta(parallel to p) space, slightly above the A(p) = 0.2 beta(parallel to p)(-0.5) curve. The lower A(p) limit for the noon cases is attributed to the presence of unmeasured cold plasma. The quiet events were all stable even for additional assumed cold ion densities of up to 10 cm(-3), the upper limit implied by the plasma wave data. The quiet events gave A(p) < 0.2 beta(parallel to p)(-0.5). At noon, the unstable component has T-perpendicular to p similar to 20 keV and A(p) similar to 0.8. At dawn the unstable component has T-perpendicular to p similar to 4 keV and A(p) similar to 2.3. Observed wave frequencies agree with the frequencies of positive growth, and the difference in frequency between noon and dawn is attributable to the combined effects of the different hot proton T-perpendicular to p and A(p) and the inferred higher cold plasma density at noon. The dawn events had significant growth for highly oblique waves, suggesting that the linear polarization of the dawn waves may be due to domination of the wave spectrum by waves generated with oblique wave vectors.