We have shown that the onset of deceleration is marked by a large increase in rest-frame emissivity and a major change in the jet collimation. It is not merely that the jet becomes gradually brighter as it decelerates and Doppler suppression is reduced: there is also a discontinuity at the inner boundary.
One possibility is that the jet is supersonic, over-pressured and
expanding freely in the inner region. In that case, the internal pressure
would fall until it drops below that of the external medium, at which
point a reconfinement shock forms (Sanders 1983). The reconfinement shock
is followed by a second shock at which the jet becomes overpressured
with respect to the external medium and it this feature which is most
plausibly identified with the flaring point. For a relativistic jet, Komissarov 1994 shows that the shock forms at a distance
We see no evidence for any simple shock structure at the inner boundary,
although the emission there is not completely resolved and there are
obvious (non-axisymmetric) knots at the beginning of the flaring region.
If the inner region is in free expansion, we can estimate the initial Mach
number of the flow from the opening angle:
where
is
the generalized Mach number defined by Königl 1980,
,
is the internal sound speed, and
. The observed value of
corresponds to
and hence to
if
the inner jet has the sound speed
of an
ultra-relativistic gas. This initial velocity is considerably faster than
we have inferred for the inner region but, as mentioned in
Section 5.1, we cannot exclude the presence of such
higher-velocity material there.
A second possibility which has frequently been discussed in the literature is that the flaring point marks the onset of turbulence, or the position at which Kelvin-Helmholtz instabilities become non-linear (e.g. Begelman 1982 ; Bicknell 1986 ; Baan 1980 ; Rosen et al. 1999 ; De Young 1996 ; Bicknell 1984 ; Rosen & Hardee 2000).
We will show elsewhere (Laing & Bridle 2002) that conservation-law analysis favours the hypothesis that the flaring point is associated with a stationary shock, primarily because it suggests that the jet is over-pressured at the beginning of the flaring region. This does not, of course, exclude the subsequent development of entrainment (and presumably turbulence), as we now discuss.