Solar wind-magnetosphere interactions cause significant energization of plasma in the magnetotail, which is later convected earthward with large anisotropies. Various plasma instabilities, modeled using both fluid theory and kinetic theory, draw on the free energy of these particle distributions to generate ULF waves [ Xue et al., 1993]. These waves can either themselves carry energy into the ionosphere, or can in turn affect (and often destabilize) other portions of the magnetospheric plasma population. This report reviews only Pc 1-2 waves, with frequencies between 0.1 and 5 Hz.
Systematic studies of AMPTE CCE satellite observations of Pc 1-2 waves by Anderson et al. [1992a,b] provided a significant new perspective on their morphology and on conditions needed for their generation. They found the highest occurrence probability beyond L = 7 in the afternoon sector, with a secondary maximum near the plasmapause. Their analysis of linear growth rates confirmed that, in addition to the commonly studied source at the plasmapause, where both hot ring current ions and cooler plasmaspheric ions are present, temperature anisotropies that develop in convecting plasma sheet ions provide an efficient free energy source for electromagnetic ion cyclotron instabilities, especially at large L values. Theoretical studies by Denton et al. [1992] and Horne and Thorne [1994] appear to explain these and other, more detailed features of the observed wave distributions.
Anderson and Hamilton [1993] showed that even for these waves
external influences are important: they used correlated magnetic
field and energetic particle data from the AMPTE CCE satellite to
show that magnetic field compressions in the outer dayside
magnetosphere enhance the instability of energetic proton
distributions to electromagnetic ion cyclotron waves, as predicted
by Mandt and Lee [1991]. Compressions generate waves most
readily close to the magnetopause, but are effective in to at least
L = 6 on the dayside. In addition to rapid field increases often
causing onset of wave activity, Anderson and Hamilton [1993]
showed that field decreases were almost equally effective in
terminating existing wave events. Evidence of the importance of
more sustained compressions was given by Erlandson et al.
[1994], who described a spatially and temporally extended 1--3 Hz
wave event that occurred during a long magnetospheric compression (a
6-hour increase in the equatorial D
magnetic storm index).
The event, which lasted over 4 hours, was observed with many similar
properties at ground stations separated by over 9 hours in local
time and from L 
5 to >15.
Documentation of the influence of these waves on plasma populations
was extended by Anderson and Fuselier [1994], who detailed a
modest, nonresonant heating of protons and strong, off-equatorial
gyroresonant heating of He
ions during periods of Pc 1-2 wave
activity.
At higher latitudes, Menk et al. [1992] used ground data along with DMSP (Defense Meteorological Satellite Program) particle data to develop a schematic picture of the source regions for various types of Pc 1 and 2 waves observed on the dayside at high and very high latitudes, with particular focus on those which might be used as cusp and cleft identifiers. Efforts to verify this picture are underway.