J. Giacalone et al., INJECTION AND ACCELERATION OF THERMAL PROTONS AT QUASI-PARALLEL SHOCKS - A HYBRID SIMULATION PARAMETER SURVEY, J GEO R-S P, 102(A9), 1997, pp. 19789-19804
Collisionless shocks that propagate along the mean magnetic field are
known to accelerate some fi action of the incident charged particles d
irectly from the thermal pool to energies that are considerably higher
than the energy at which the plasma rams into the shock. Using hybrid
simulations, we address two issues: (1) the dependence of the injecti
on/acceleration of thermal protons to energies much higher than the pl
asma ram energy on various shock parameters such as Mach number, plasm
a beta, etc., and (2) the effect of the high-energy particles, acceler
ated directly from the thermal population by the shock, on the macrosc
opic Properties of the shock, most notably, on the density compression
. We find that for supercritical Mach numbers the acceleration of the
thermal plasma is efficient enough that the back pressure due to the e
nergetic particles can significantly increase the density compression
across the shock, above the value expected from the simple Rankine-Hug
oniot prediction. Additionally, at low Alfven Mach number, where the a
cceleration of the thermal plasma is inefficient, the density compress
ion is smaller than the simple Rankine-Hugoniot prediction owing to th
e nonresonant fire hose instability. The acceleration efficiency incre
ases with Mach number except at very high Alfven Mach numbers, where i
t begins to decrease for Mach numbers greater than similar to 10. This
is due to the presence of a fixed, free-escape boundary that limits t
he size of the foreshock region measured in units of the mean-free pat
hs of the accelerated particles. Additionally we find that regardless
of the upstream plasma parameters,the acceleration efficiency increase
s with both the density compression ratio across the shock and the dis
tance to the free-escape boundary measured in units of the mean-free p
ath of the energetic particles. Both of these are consistent with anal
ytic theory and numerical models that use a phenomenological scatterin
g law.