Following Hollweg and Johnson [1988], Isenberg [1990], and Li et al. [1999a
], we postulate that the Sun launches a flux of low-frequency Alfven waves,
which dissipate via a turbulent cascade to high frequencies where the ener
gy is absorbed by ion cyclotron resonant interactions. The plasma consists
of two proton beams, which are proxies for the resonant and nonresonant hal
ves of their distribution function, two He++ beams, which are proxies for t
he strongly and weakly resonant halves of their distribution, and a single
beam of O+5 with vanishing density. The level of the power spectrum at the
high resonant frequencies is determined by the condition that the protons a
nd He++ resonantly absorb energy at the same rate at which the low-frequenc
y waves are dissipating. Once the level of the high-frequency power spectru
m is determined, the resonant heating and acceleration of the O+5 can be ca
lculated. For both Kolmogorov and Kraichnan scalings of the turbulent dissi
pation the model yields results for the protons that are in reasonably good
agreement with the UVCS/SOHO results. The He++ becomes; more than mass pro
portionally heated and flows faster than the protons, close to the Sun. How
ever, our model is unable to reproduce the UVCS/SOHO observation that the O
+5 temperature is still increasing with heliocentric distance r out to 3.5
r(s). Instead, the O+5 becomes very hot initially, experiences a strong mir
ror force, and accelerates to high speed, which in turn leads to rapid adia
batic cooling. Put another way, the O+5 observations imply that (dT(perpend
icular to)/dt)(res) must bean increasing function of r, while it is the nat
ure of the resonant interactions to make (dT(perpendicular to)/dt)(res) dec
rease with increasing r.