Three-component dynamic spectrograms (0--80 mHz) of AMPTE/CCE magnetic field data from August 24, 1984, through December 7, 1985, have been used to survey ULF pulsation activity occurring from L=5 to 9 in the equatorial magnetosphere (¿16¿ magnetic latitude) at all local times. The data were scanned visually, and each half-hour interval was categorized by spectral type, approximate polarization, spectral intensity, and spacecraft location to produce a data base representing 7231 hours of observations. Coherent pulsations are divided into nine categories which fall into four basic classes: (1) harmonic toroidal resonances (17.9% of the data), (2) fundamental mode toroidal resonances (11.0%), (3) radially polarized pulsations (5.2%), and (4) compressional low-frequency pulsations (10.6%). Classes (1) and (2) are not mutually exclusive. Intervals devoid of coherent pulsation are divided into five levels according to noise intensity and account for 56.8% of the observations. Uncategorized events represented 4.2% of the data. The spatial occurrence distributions of each pulsation category and the latitude distribution of selected categories have been examined. The basic conclusions of the study are as follows: (1) Harmonic torodial resonances are found to be the dominant coherent activity on the dayside, particularly in the prenoon hours, where they occur 60% of the time. Their region of excitation is uniformly distributed in radial distance and extends in local time from 0600 to 1600 but cuts off sharply at the local time boundaries. The pattern is consistent with excitation by a dayside energy source, possibly upstream waves, but the spatial distribution place constraints on models of energy transmission. (2) Fundamental toroidal resonances are observed with an occurrence rate of 40% to 50% for L>8 at dawn but were observed less than 10% of the time at dusk for L>8.

Dusk is also the site of several types of pulsations of sufficient average intensity to obscure fundamental mode resonances, so this trend may be exaggerated by masking effects. Nonetheless, the dawn/dusk asymmetry is sufficiently remarkable to warrant further study. The occurrence distribution exhibits a pronounced node at the magnetic equator, and the occurrence rate for MLAT>13¿ at dawn is ≈80%, suggesting that the fundamental mode resonances are present continually at the dawn flank. The preference for occurrence at the flank suggest that the Kelvin-Helmholtz instability is ultimately responsible for driving the resonances. (3) Radially polarized pulsations are observed≈10% of the time except in the 0400--1100 MLT, sector where their occurrence rate is less than 5%. The occurrence distribution of radially polarized waves suggests that they are driven by wave-particle interactions but not byfreshly injected particles. (4) Storms time Pc 5 type waves were observed most often at dusk for L>8 where they ocurred≈30% of the time. A secondary maximum in storm time Pc 5 type occurrence near dawn was also found. The spatial distribution of storm time Pc 5 type waves suggests a relationship to substorm injections, although the secondary maximum at dawn may not be explained by this mechanism alone. (5) Quiet periods represent 30% to 50% of nightime but less than 10% of daytime observations, while noisy periods represent ≈20% of the data at night and 40% to 50% during the day. ¿American Geophysical Union 1990