The AMPTE/CCE spacecraft, with an apogee of approximately 8.8R(E) and
an inclination of approximately 4.3-degrees, sampled the outer dayside
equatorial magnetosphere for extended time periods and often crossed
into the magnetosheath whenever the solar wind pressure was sufficient
ly high to compress the magnetopause to <8.8R(E). We have analyzed ion
distributions on both sides of the magnetopause in order to investiga
ted any local time and energy dependence, giving information about phy
sical processes at the magnetopause and the bow shock. Particle measur
ements are from the CHEM (1.5 to 300 keV/e) and MEPA (10 keV to 2 MeV)
instruments. The wide total energy range (1.5 keV to 2 MeV) covered d
escribes the magnetospheric distributions quite well, and for the purp
ose of this study, adequately describes the high-energy part of the sh
ocked solar wind. Thus both solar wind and magnetospheric components c
an be recognized in a mixed particle distribution. Case studies of rep
resentative magnetopause crossings at dawn, noon, and dusk, as well as
a survey of several other crossings, indicated: (1) a local time and
energy dependence of magnetosheath spectra at energies greater-than-or
-equal-to 50 keV; spectra were harder at the duskside than at the dawn
side and also correlated with magnetospheric activity, (2) constantly
much higher intensities in the magnetosphere than in the magnetosheath
at energies >10 keV and an earthward gradient in the subsolar magneto
sheath. In addition to the steady state magnetosheath population there
exists a burst-type component indicative of a magnetospheric source,
and most of the time this is recognized as a flux transfer event. Over
all, the results about the origin of the greater-than-or-equal-to 50 k
eV magnetosheath ions are consistent with the continuous leakage of ma
gnetospheric particles across a tangential discontinuity magnetopause,
locally distributed according to magnetospheric drift paths. Magnetic
reconnection, although present, should not be a dominant source on av
erage, because it is not continuous in time. Fermi acceleration should
not be dominant because it predicts the opposite local time asymmetry
, and shock drift acceleration should be a minor contributor at E grea
ter-than-or-equal-to 50 keV because of upper-energy cutoff limitations
. Our observations also indicate a significant magnetospheric contribu
tion to energies as low as approximately 10 keV, where the magnetosphe
re-magnetosheath intensity gradient reverses. However, in order to exa
mine the relative strength and local time distribution of all possible
sources at these energies, a detailed analysis is required.