Bulk Comptonization of the Cosmic Microwave Background by Extragalactic Jets as a Probe of their Matter Content
Extracted source text ยท arXiv astro-ph/0502201
%\documentclass{aastex}
%% preprint produces a one-column, single-spaced document:
%\documentclass{article}
%\usepackage{emulateapj}
\documentclass[12pt,preprint]{aastex}
%% preprint2 produces a double-column, single-spaced document:
%\documentclass[preprint2, 12pt]{aastex}
\def\gsim{\mathrel{\raise.5ex\hbox{$>$}\mkern-14mu
\lower0.6ex\hbox{$\sim$}}}
\def\lsim{\mathrel{\raise.3ex\hbox{$<$}\mkern-14mu
\lower0.6ex\hbox{$\sim$}}}
%\received{}
%\accepted{}
%\journalid{}{}
%\articleid{}{}
\slugcomment{Submitted to the Astrophysical Journal.}
\shorttitle{The matter content of quasar jets}
\shortauthors{Georganopoulos, Kazanas, Perlman, \& Stecker}
\begin{document}
\title{ Bulk Comptonization of the Cosmic Microwave Background by Extragalactic Jets as a Probe of their Matter Content}
\author{Markos Georganopoulos\altaffilmark{1,2}
, Demosthenes Kazanas\altaffilmark{2},
Eric Perlman \altaffilmark{1},
Floyd W. Stecker\altaffilmark{2}}
\altaffiltext{1}{Department of Physics, Joint Center for Astrophysics, University of Maryland-Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, USA}
\altaffiltext{2}{Laboratory for High Energy Astrophysics,
NASA Goddard Space Flight Center,
Code 661, Greenbelt, MD 20771, USA.}
\begin{abstract}
We propose a method for estimating the composition, i.e. the relative amounts
of leptons and protons, of extragalactic jets which exhibit
X-ray bright knots in their kpc scale jets. The method relies on measuring,
or setting upper limits on, the component of the Cosmic Microwave
Background (CMB) radiation that is bulk-Comptonized by cold electrons
in the relativistically flowing jet.
These measurements, along with modeling of the broadband knot emission
that constrain the bulk Lorentz factor $\Gamma$ of the jets, can yield
estimates of the jet power carried by protons and leptons. We provide
an explicit calculation of the spectrum of the bulk-Comptonized
(BC) CMB component and apply these results to
PKS 0637--752 and 3C 273, two superluminal quasars with
{\sl Chandra} -- detected large scale jets.
What makes these sources particularly suited for such a procedure is the
absence of significant non-thermal jet emission in the {\sl `bridge'},
the region between the core and the first bright jet knot,
which guarantees that most of the electrons are cold there,
leaving the BC scattered CMB radiation as the only significant source
of photons in this region.
At $\lambda=3.6-8.0 \; \mu m$,
the most likely band to observe the BC scattered CMB emission,
the {\sl Spitzer}
angular resolution ($\sim 1''-3''$) is considerably smaller than
the `bridges' of these jets ($\sim 10''$),
making it possible to both measure and resolve this emission.
\end{abstract}
\keywords{galaxies: active --- quasars: general --- quasars: individual (PKS 0637-752, 3C 273) --- radiation mechanisms: nonthermal --- X-rays: galaxies}
\section{Introduction \label{intro}}
One of the primary issues that impedes our understanding of extragalactic
jets is that of their composition.
The power and momentum transferred by the jets from their point of origin,
near the central black hole, to the radio
lobes at Mpc scales is transported by an electrically neutral,
yet unspecified
combination of leptons (both electrons and positrons or $e^\pm$), protons,
and Poynting flux carried by the magnetic field of the flow. For a given
radiative output, the jet power depends on the composition of the outflowing
matter, with leptonic ($e^\pm$) jets demanding less overall power and mass
and being easier to accelerate to relativistic bulk flow velocities
than hadronic ($e-p$) ones. Uncertainty in the matter
content results in uncertainty in the jet power, which bears on our
understanding of the jet dynamics and the radio lobe energetics, their
interaction and influence on the host galaxy and possibly the embedding cluster
core (e.g. Omma \& Binney 2004; McNamara et al. 2005). The composition is also related to
the fundamental problem of jet formation: the plasma in jets powered by an
accretion disk (Blandford \& Payne 1982; K\"onigl 1989) would be expected to
be baryon loaded, while jets powered by the rotational energy
of a black hole are more likely to result in $e^\pm$ jets (Blandford \&
Znajek 1977).
A number of attempts have been made over the years toward measuring, or at
the least constraining, the matter content of jets, in particular the
fraction of kinetic energy stored in protons and low energy or cold leptons,
whose low radiative efficiencies fail to provide direct evidence of their
presence. To this end, a variety of arguments
have been presented, based on the synchrotron - self Compton (SSC)
formalism (e.g. Celotti \& Fabian 1993; Reynolds et al. 1996; Hirotani 2005),
on the circular polarization of the radio emission measured in 3C 279 and
other sources (e.g. Wardle et al. 1998),
and on the pressure balance of the radio
lobes and the X-ray emitting confining plasma (e.g., Gizani \& Leahy 2004). These largely indirect arguments have produced rather contradictory results,
with different works supporting leptonic and hadronic
contributions of varying proportions to the jet content.
A direct estimate of the cold lepton content of blazar jets was proposed
by Begelman \& Sikora (1987), Sikora \& Madejski (2000), and Moderski
et al. (2004). The gist of their argument is the following: The observed
non-thermal blazar emission is thought to be produced at distances $\sim 10^{17}
-10^{18}$ cm from the central engine (e.g. Sikora 1997); the jet leptons
providing the blazar emission at these distances need to be transported
practically cold by a relativistic flow of bulk Lorentz factor $\Gamma \sim 10$
from the black hole vicinity to the blazar emission site; as these cold jet
leptons propagate through the quasar broad line region (BLR) they would
Compton -- scatter the BLR optical-UV photons (of energy $E_{\rm O-UV}
\sim 10$ eV) to energies $E_{\rm BC} \simeq \Gamma^2 E_{\rm O-UV}
\simeq 1$ keV, to produce a black--body type hump in their X-ray spectra.
The fact that such a feature has not been observed in the inverse-Compton
dominated X-ray spectrum of blazars was used by the above authors to conclude
that the power in these jets is carried mainly by protons,
although cold leptons dominate the number of particles in the jet.
While this idea is well founded and appealing, concrete answers
are hindered by unknowns such as the distance at which the jet is formed,
its sub-pc scale opening angle and the actual photon energy density of
the BLR, as well as by the presence of a strong X-ray non-thermal continuum
that apparently could ``hide" the proposed bulk-Comptonized component.
Arguments based on the bulk Compton (BC) emission used by the above authors
can be applied to any
astrophysical site involving relativistic flows. One can then obtain more
concrete conclusions provided that the flow geometry and the target photon
density are better determined. Such a site is presented by the large scale
(100's of kpc) jets of superluminal quasars such as PKS 0637--752 (Schwartz
et al. 2000; Chartas et al. 2001) and 3C 273 (Marshall et al. 2001;
Sambruna et al. 2001; Jester et al. 2002) detected by the {\sl Chandra
X-ray Observatory}. Such large scale jets are tightly collimated and propagate through
the very well understood CMB.
The effect of the CMB scattering
by the jet cold electrons is akin to the well known Sunyaev-Zeldovich (SZ)
effect associated
with clusters of galaxies, with the difference being that in this case
the CMB photons are scattered by electrons moving with a bulk
relativistic velocity (see below) rather than by thermal electrons.
These objects and their jets have morphologies which are conducive to
applying the arguments referred to above. They exhibit radio, optical,
and X-ray emission from the
quasar core and then from well separated knots along the jet at angular
distances $\sim 8''$ for PKS 0637--752 and $\sim 13''$ for 3C 273.
The region between the core and the first knot, which we will refer to
as the {\sl `bridge'} in the rest of this work, radiates only very weakly
in radio, optical, and X-ray energies. This is very important because:
({\it i}) It shows
that most of the leptons propagating through the `bridge' are `cold'.
({\it ii})
It provides a region free from unwanted contamination by unrelated broad
band non-thermal radiation.
% In this respect, this effect is far cleaner than the cluster SZ effect, in that one is not searching for small deviations in the CMB spectrum produced by an external agent, but for a whole new component in a part of the EM spectrum where no other emission is expected.
As in the SZ effect, where the X-ray cluster properties provide values of
the parameters involved in the scattering of
the CMB photons, the jet properties (power and kinematics) are provided
by the properties (spectrum and luminosity) of the {\sl Chandra} --
detected knots of these jets.
%In its very first
%observation, {\sl Chandra} detected X-ray emission in the superluminal
%quasar PKS 0637--752 (Schwartz et al.2000; Chartas et al. 2000), from knots
%at a projected distance of $\sim 70-100$ kpc from the nucleus.
Schwartz et al. (2000) noted that {\sl Chandra}-detected X-ray emission from
the jet knots
at a projected distance of $\sim 70-100$ kpc from the nucleus of PKS
0637--752 is part of a spectral component separate from the knot radio-optical
synchrotron emission and that it is too bright to be explained through
SSC emission from electrons in energy equipartition with the jet magnetic
field. Observational evidence now indicates that this property is common to
the jets of many other quasars (Sambruna et al. 2002, 2004;
Siemiginowska et al. 2002, 2003; Marshall et al. 2005; Jorstad \& Marscher
2004; Yuan et al. 2003; Cheung 2004).
Tavecchio et al. (2000) and Celotti et al. (2001) proposed that the X-ray
emission is due to external Compton (EC) scattering of CMB photons off
relativistic electrons in the jet, provided that the jet flow is sufficiently
relativistic ($\Gamma \sim 10$) to boost the
CMB energy density in the flow frame (by $\Gamma^2$) to the level
needed to reproduce the observed X-ray flux.
This was the first suggestion,
based on extended X-ray emission, that powerful jets retain significantly
relativistic velocities at large distances from the core, a very important
feature because it boosts the level of the anticipated
BC emission by $\sim \Gamma^2$.
Here we argue that our recently gained understanding that the
{\sl Chandra}-detected quasar jets remain relativistic on scales of
hundreds of kpc, together with the IR capabilities of {\sl Spitzer},
can be used to measure or substantially constrain the
matter content of these jets. In \S \ref{section:bc} we
calculate the BC spectrum of the CMB as a
function of the cold lepton kinetic power and Lorentz factor $\Gamma$,
as well as
its polarization.
In \S \ref{section:xmech} we present the mechanisms that have been proposed
to explain the X-ray knot emission, and we argue that EC off the CMB is the
most probable mechanism.
In \S \ref{section:power}
we use the simple analytical arguments of Dermer \& Atoyan
(2004a; hereafter DA04) to derive
the flow velocity and estimate the number of electrons carried by the jet.
We then apply our calculations to the quasars PKS 0637--752 and 3C 273 in
\S \ref{appl}.
Finally, in \S \ref{section:discussion} we discuss our findings along
with some caveats and present our conclusions.
\section{Bulk Compton Spectrum and Polarization \label{section:bc}}
The expected level of BC emission depends on the jet power in cold leptons,
its length, its Lorentz factor $\Gamma$ and its angle to the observer's line
of sight. We now proceed to calculate the spectrum and luminosity of the
radiation produced by the propagation of a collimated cold electron beam
(mean electron Lorentz factor $\langle \gamma \rangle \approx 1$) of
Lorentz factor $\Gamma$ and power $L_{e}$ through an isotropic photon field.
We start with a simple estimate of the peak energy and peak luminosity
as a function of observing angle assuming a monoenergetic photon field
with dimensionless energy $\epsilon_0$ (in units of $m_ec^2$) and
energy density $U$. Assuming that the jet has a length $l$, the fractional
energy loss of a single electron after propagating this distance $l$ is
%
\begin{equation}
{\Delta\Gamma\over\Gamma}={4\over 3}\, {\sigma_{\rm T} \over m_e c^2}
\,\beta \Gamma U l,
\end{equation}
%
where $\beta$ is the beam velocity in units of $c$ and $\sigma_{\rm T}$
is the Thomson cross section. Given that the CMB
photon energy density $U$ scales with redshift $z$ as $U=U_0 (1+z)^4$,
where $U_0=4.18 \;10^{-13}$ erg cm$^{-3}$, for a
large scale quasar jet propagating through the CMB, the fractional
energy losses are
%
\begin{equation}
{\Delta\Gamma\over\Gamma}\approx 1.4 \times 10^{-6} \Gamma_{10}\;
l_{100 Kpc}\; (1+z)^4.
\end{equation}
%
Therefore, the bulk deceleration due to Compton drag off the CMB can be
safely ignored for $z \lesssim 10$, even for the extreme case of a
purely leptonic large scale jet.
The inferred isotropic BC luminosity $L_{BC}$ at a given observing
angle $\theta$ is given by the fractional electron losses multiplied
by $L_{e}$ and by the beaming pattern $\delta^3/\Gamma$ of a continuous
flow (e.g. Sikora et al. 1997),
%
\begin{equation}
L_{BC}=L_e \; {\Delta\Gamma\over\Gamma} \; {\delta^3\over \Gamma}=
{4\over 3}\, {\sigma_T \over m_e c^2} \; L_e \beta U l \, \delta^3,
\label{eq:LBC}
\end{equation}
%
where $\delta=1/[\Gamma(1-\beta\cos\theta)]$ is the well known Doppler
factor. The observed luminosity peaks at
%
\begin{equation}
\epsilon_{BC}\approx 2 \epsilon_0 \Gamma \delta.
\label{eq:fBC}
\end{equation}
%
If we assume $\theta=1/\Gamma$, which
corresponds to $\delta=\Gamma$, the BC emission has a power
%
\begin{equation}
L_{BC}\approx 1.4 \times 10^{-4} \;\; l_{100 Kpc} \; \Gamma_{10}^3 (1+z)^4 L_e,
\end{equation}
%
and peaks in the IR regime, $\nu_{BC}\approx 4 \times 10^{13} \; \Gamma^2_{10}$
Hz. For constant Lorentz factor jets, this emission due to the BC scattering
of the CMB will be evenly distributed between the core and the first knot,
which for $\theta=1/\Gamma$ will have a projected length $l/\Gamma$, and
a luminosity per unit projected length scaling as $\Gamma^4$. Note that:
({\it i}) the BC surface brightness is independent of $z$, because its
cosmological decrease by
$(1+z)^{-4}$ is exactly compensated by the increase of the CMB energy density
by $(1+z)^{4}$ (Schwartz 2002). ({\it ii}) The observed peak frequency
is also independent of $z$ because, while the CMB photon energy and
therefore the peak emission energy scales as $(1+z)$, the received
photon energy scales as $1/(1+z)$. ({\it iii}) The independence of the
surface brightness of the BC component on the redshift $z$ and the fact
that the angular size of an object is roughly constant for $z \gsim 1$,
imply that the flux of this feature will also remain constant independent of
the source redshift (Schwartz 2002), in distinction to the core and remaining
synchrotron
jet emission that decrease significantly with increasing $z$.
{\it The Spectrum.}
To derive the spectrum of the BC emission off the CMB, we extend
the calculation of section \S 7.3 of Rybicki \& Lightman (1979)
from a monoenergetic to a blackbody photon field. The intensity of the
blackbody photon field in terms of number of photons is:
%
\begin{equation}
I(\epsilon)=F_0\; \frac { \epsilon^2 } {\exp(\epsilon /\epsilon_0)-1},
\; F_0 \equiv \frac{2 m_e^3 c^4}{h^3},\; \epsilon_0 \equiv kT/m_ec^2
\end{equation}
%
where $T$ is the blackbody temperature. Using the invariance of $I/
\epsilon^2$, we obtain the incident intensity field in the (primed)
frame of the electron beam:
%
\begin{equation}
I'(\epsilon',\mu'_0)=F_0 \frac{\epsilon'^2}{\exp[\epsilon'\Gamma (1+\beta\mu'_0)/\epsilon_0]-1},
\end{equation}
%
where $\mu'_0=\cos\theta_0'$ and $\theta_0$ is the angle the incoming
photons form with the direction of the electron motion.
The emission function in the beam frame, assuming isotropic Thomson
scattering with
$d\sigma'/d\Omega'_0=\sigma_{\rm T}/4\pi$, is
%
\begin{equation}
j'(\epsilon_1')=\frac{1}{2}n'\sigma_{\rm T}
\int_{-1}^{+1} I'(\epsilon',\mu'_0) d\mu'_0,
\end{equation}
%
where $n'$ is the electron number density in the beam frame.
Performing the integration, we obtain:
%
\begin{equation}
j'(\epsilon_1')=\frac{n'\sigma_{\rm T}F_0 \epsilon_0}{2\beta\Gamma}
\epsilon'_1\;
\ln \frac{1-\exp[-\epsilon'_1\Gamma(1+\beta) /\epsilon_0]}
{1-\exp[-\epsilon'_1 /(\epsilon_0\Gamma(1+\beta))]}.
\end{equation}
%
Using the invariance of $j/\epsilon$ we obtain:
%
\begin{equation}
j(\epsilon_1,\mu)=\frac{\epsilon_1}{\epsilon'_1}j'(\epsilon_1')=
\frac{n'\sigma_{\rm T}F_0 \epsilon_0 }{2\beta\Gamma}\epsilon_1\;
\ln \frac{1-\exp[-\epsilon_1\Gamma^2(1+\beta)(1-\beta\mu) /\epsilon_0]}
{1-\exp[-\epsilon_1 (1-\beta\mu) /(\epsilon_0(1+\beta))]}.
\end{equation}
%
The power of the cold electron beam is $L_e\approx S \beta c \Gamma^2 n'$,
where $S$ is the jet cross section. Using this in the above equation
and multiplying by the beam volume $V=Sl$, by the final photon energy $\nu=m_ec^2\epsilon_1 /h$, and by $4\pi$, we obtain the BC specific
luminosity of the source
%
\begin{equation}
L_{\nu}=\frac{L_{e}\sigma_{\rm T}l kT}{m_ec^5\beta^2\Gamma^3}\; \nu^2 \;
\ln \frac{1-\exp[-h\nu\Gamma(1+\beta)/(\delta\, kT)]}
{1-\exp[-h\nu /(\Gamma \delta \, kT(1+\beta))]}, \label{app.lum}
\end{equation}
%
where we have also used the definition of $\delta$ and the relation
$\epsilon_1/\epsilon_0=h\nu/kT$.
As can be seen in Figure \ref{BC} the simple estimates of the
luminosity and peak frequency agree relatively
well with the results of the spectral calculation, and can be safely used
for simple estimates.
{\it The Polarization.} The emission we consider, being the result of
Compton scattering, is expected to be polarized, perhaps very highly so, since
in the jet flow rest frame the seed photons are essentially unidirectional.
The polarization can be easily calculated in the approximation of a
perfectly collimated cold electron beam with $\Gamma \gg 1$ (for a general
treatment of BC emission polarization
see Begelman \& Sikora 1987). The degree of polarization $\Pi$ is a Lorentz
invariant and at the beam frame can be written as $\Pi=(1-\mu'^2)/(1+\mu'^2)$.
Using the light aberration relation $\mu'=(\mu-\beta)\Gamma\delta$, one can
obtain immediately the degree of polarization $\Pi$ in the observer's frame:
%
\begin{equation}
\Pi=\frac{1-(\mu-\beta)^2\Gamma^2\delta^2}{1+(\mu-\beta)^2\Gamma^2\delta^2}.
\end{equation}
%
As can be seen in Figure \ref{polosi}, the degree of polarization is a
very sensitive function of orientation, reaching $100 \% $ at $\theta=1/
\Gamma$ and dropping rapidly to $0 \%$ at $\theta=0$. Using the above
expression along with the relation $\epsilon_{BC}=2\Gamma\delta\epsilon_0$
between the initial $\epsilon_0$ and final $\epsilon_{BC}$ photon energies, one
can obtain the values of both $\beta$ (or equivalently $\Gamma$)
and $\theta$ in terms of the observables $\Pi$ and $\epsilon_{BC}$.
Setting $A^2=(1-\Pi)/(1+\Pi)$ and $B=2\epsilon_0/\epsilon_{BC}$, we obtain
%
\begin{equation}
\beta=\frac{-AB\pm\sqrt{A^2B^2+4(1-B)}}{2}, \;\; \mu=\cos\theta=AB+\beta,
\end{equation}
%
where the positive sign corresponds to an approaching beam. Such polarization
measurements of the BC emission can in principle be performed in the near IR
to provide an independent measurement of both the Lorentz factor $\Gamma$
and the angle $\theta$ of the jet to the observer's line of sight that can
break the degeneracy normally present in, e.g., VLBI measurements.
\section{The X-ray emission mechanism\label{section:xmech}}
The BC emission of the CMB in the `bridge' region
depends on the power carried by cold leptons, the flow Lorentz
factor $\Gamma$ and the angle of the jet to the observer's line of sight
$\theta$, as shown by Equation (\ref{eq:LBC}). Estimates or constraints
on these quantities are provided by the non-thermal emission at the
knot marking the end of the `bridge'. The
{\sl Chandra} detections are critical in providing such estimates.
This brings on the issue of the X-ray emission process in the knots
of the {\sl Chandra-}detected quasar jets, as different mechanisms produce
different constraints for the quantities on which the level of BC
emission depends. While EC of the CMB (Tavecchio et al. 2000; Celotti et
al. 2001) appears the most promising process, we feel that a critical
examination of the alternatives is at this point necessary.
Dermer \& Atoyan (2002) suggested that the observed X-rays are due to
synchrotron emission from electrons cooling by EC off the CMB in the
Klein--Nishina (KN) regime. Due to the reduced KN-losses of the highest
energy electrons on the thermal CMB spectrum, their distribution function
develops a ``hump" at these energies, which manifests itself in
the synchrotron emission of these electrons as an increase in the spectral
luminosity of this component between UV and X-ray energies. The continuity
of the electron distribution function, then, implies also the continuity
of the spectrum between optical, UV, and X-rays.
As a result, the extrapolation of the observed X-ray spectrum to lower
frequencies, namely UV and optical, must {\sl always} lie below the
observed UV and optical fluxes.
This is actually contrary
to observation, as indicated by the optical detections or upper limits
at several such knots (e.g. PKS 0637--752, Chartas et al. 2000; Knot A of
1354+195, Sambruna et al. 2004; Knot B of 1150-089, Sambruna et al. 2002;
Knot C4 of 0827+243, Jorstad \& Marscher 2004). This interpretation
could be valid, however, for the X-ray
emission of knot A of 3C 273, as the near-IR - optical -UV spectrum
indeed turns upward, pointing toward the X-ray point (Jester et al. 2002).
Alternatively, it has been suggested (e.g. Schwartz 2000) that the
X-rays are due to synchrotron radiation by a second, very energetic
electron population with a low energy cut-off at sufficiently high
energy that its synchrotron emission results in a low energy cut-off
at UV energies to comply with observation. However, even if an unknown
mechanism can produce the injection of an election population with
the above properties,
they would cool in less than a knot-crossing time to energies below
those corresponding to optical synchrotron emission. This fact has two
unfavorable implications: ($i$) Given the observed X-ray spectral indices
($\alpha_x = (\gamma - 1)/2 \sim 0.5 - 0.8 $; e.g. Sambruna et al. 2004),
the {\sl injected} electron distribution must have an index $p$
flatter by one unit than that observed , i.e., $p = \gamma +1 =2\alpha_x
\sim 1-1.6 $ (see also Aharonian 2002), significantly flatter than
the asymptotic values predicted by particle
acceleration theories ($ p \simeq 2 - 2.3$, e.g. Kirk et al. 2000).
($ii$) These high energy electrons will cool below their low energy
cut-off to produce, in this energy range, an electron distribution $N_e(
\gamma)\propto \gamma^{-2}$; the synchrotron emission of these electrons
would then lead to a $\nu^{-1/2}$ spectrum that in
many cases, such as PKS 0637--752, overproduces the observed optical fluxes.
%ADDITION STARTS
These problems can be overcome if instead of a simple injection
one invokes the continuous acceleration of the radiating electrons in
multiple shocks or spatially distributed stochastic
acceleration (Stawarz et al 2004). In both cases such models can produce
a pile-up of high energy electrons at the upper end of the electron
distribution which could lead to X-ray synchrotron consistent with
observations. These models, like that of Dermer \& Atoyan (2002), can
only model successfully emission by sources in which the extrapolation
of the X-ray spectrum to optical frequencies lies below the observed
optical flux, such as the knot A of 3C 273.
%ADDITION ENDS
The EC interpretation of the knot X-ray emission is not entirely without
problems either. Multiwavelength observations have shown that in
many cases the emission profiles at the knot regions are largely achromatic.
This behavior is unexpected (Tavecchio, Ghisellini, \& Celotti 2003;
Stawarz et al. 2004), because the cooling length of the EC X-ray
emitting electrons ($\gamma \sim$ few hundreds) is longer than that
of the radio emitting ones ($\gamma \sim$ few thousands) and comparable
to or longer than the size of the entire jet; this would lead one to
expect longer jet emission in X-rays than that in the radio.
%
%However, the simplest EC interpretation is not without problems, as became
%apparent in multiwavelength morphological studies of several jets.
However, in most {\sl Chandra-}detected quasar jets, the radio-to-X-ray
logarithmic slope $\alpha_{rx}$ increases downstream along the jet.
Indeed, some (e.g. 3C 273 in Sambruna et al. 2001
and Marshall et al. 2001; PKS 1136--135 and 1354+195
in Sambruna et al. 2002; PKS 1127--145 in Siemiginowska et al. 2002;
0827+243 in Jorstad \& Marscher 2004) show anti-correlated X-ray and
radio maps, with the X-ray emission peaking closer to the core,
gradually decreasing outward, while the radio emission increases outward
to peak practically at the jet terminus.
This problem is alleviated, however, if the large-scale jet gradually
decelerates (Georganopoulos \& Kazanas 2004) downstream from the first
knot. Then, the X-ray brightness decreases along the jet because the
CMB photon energy density in the flow frame decreases. At the same time,
the deceleration leads to an increase of the magnetic field in the
flow frame, which enhances the radio emission with distance. As a result
the radio emission is shifted downstream of the X-rays and the radio
to X-ray spectral logarithmic slope $\alpha_{rx}$ increases along the jet,
in agreement with observations. The notion of relativistic and decelerating
flows in the large scale quasar jets is in agreement with the recent
suggestion (Georganopoulos \& Kazanas 2003) that the flow in the terminal
hot spots of powerful jets must also be mildly relativistic
($\Gamma\sim 2-3$) and decelerating to sub-relativistic velocities.
In conclusion, we consider the EC model (Tavecchio et al. 2000;
Celotti et al. 2001), with the modification of the bulk
flow deceleration proposed by Georganopoulos \& Kazanas (2004)
to be the most favorable process accounting for the observed knot
emission in this class of sources.
%addition begins
However, the synchrotron
interpretation for the X-ray emission is still viable, at least for
the sources consistent with a continuous underlying electron
distribution, manifest by the continuity between the optical to
X-ray spectra, as discussed above. This possibility will be considered
in the determination of the jet properties of 3C 273.
%addition ends
\section{The Jet Power \label{section:power}}
Having discussed the mechanism responsible for the X-ray
emission of the jet knots, we now turn to the determination of their
dynamical parameters, subject to the constraint that they account
for the observed X-ray emission of the knots. These estimates are
based principally on the energetics of the emission rather than the
details of the spectra, as the latter can be reproduced by appropriate
choice of additional parameters pertaining to the particle distributions.
However, even at this level,
the number of parameters exceeds that of the observables. One
therefore resorts to minimum energy arguments in order to further
constrain the available values of the magnetic field $B$, the Doppler
factor $\delta$ and the energy flux in or the power of the jet flows.
A comprehensive set of constraints for the jet power and beaming based
on multiwavelength observations of knots in the extended jet has been recently
presented by DA04 (see also Ghisellini \& Celotti 2001).
These authors produced analytic relations for the
jet Doppler factor $\delta_{min}$ that minimizes the jet power and are
consistent with all constraints imposed by the knot multiwavelength
emission. In this respect one should note that, because the quantity that is
minimized is the total knot power and {\sl not} the power in relativistic
electrons and magnetic field only, the value of $\delta_{min}$ depends
on the matter content of the jet.
\subsection{Minimum Jet Power for Knot X-ray Emission Due to EC}
The work of DA04 models the knots as homogeneous sources moving with a
Lorentz factor $\Gamma$ at an angle $\theta$ to the line of sight.
The knot matter content is described through the ratio $k_{pe}$
of the power carried by protons to the power carried by leptons in
the knot. For a pure $e^{\pm}$ composition, $k_{pe}=0$, while for an
$e-p$ jet this parameter reaches its maximum value of
%
\begin{equation}
k_{pe}={m_p(p-2)\over m_e \gamma_{min} (p-1)},
\label{kpe}
\end{equation}
%
where $\gamma_{min}$ and $p~ (>2)$ are the minimum Lorentz factor
and energy index of the power law electron energy distribution (EED),
and the protons are considered to be cold in the knot comoving frame
(note that DA04 use $k_{pe}=m_p/(m_e\gamma_{min})$, valid only for $p\gg1$).
Assuming that the X-rays are due to EC scattering off the CMB and that
$\delta=\Gamma$, DA04 calculate
(their Eq. (12)) the Doppler factor $\delta_{min}$
that minimizes the power that has to be supplied to the knot by the jet.
Their expression can be written as
%
\begin{equation}
\delta_{min}=f_1 \left({1+k_{pe} \over \gamma_{min}^{p-2}}\right)^{1/(5+p)},
\label{delta}
\end{equation}
%
where $f_1$ depends on the source redshift, the radio and
X-ray fluxes and spectral index (the radio and X-ray index are assumed to
be the same, as would be the case if they are due to a
single power-law electron distribution)
and the linear size of the knot ($f_1$, along with the subsequently used $f_2,\; f_3,\; f_4$
are reproduced in Appendix \ref{fs}).
In particular, as can be seen from Eq. (\ref{kpe}) and (\ref{delta}),
the ratio of hadronic to leptonic Doppler
factors depends only on $\gamma_{min}$ and $p$:
%
\begin{equation}
{\delta_{min, e-p}\over \delta_{min,e^{\pm}}}=(1+k_{pe})^{1/(5+p)}\approx
\left[{m_p(p-2)\over m_e \gamma_{min} (p-1)}\right]^{1/(5+p)}, \;\mbox{for} \;
\; k_{pe} \gg 1
\end{equation}
%
The lowest possible value for $\gamma_{min}$ is constrained by the
requirement that the EC emission does not extend to frequencies as low
as optical.
%$\epsilon_{0}\gamma_{min}^2 \delta_{min}^2 > \epsilon_{opt}$, where
%$\epsilon_{0}$ and $\epsilon_{opt}$
%are the CMB photon energy at redshift $z=0$ and the energy of the {\sl HST}
%optical observations.
Similarly the maximum possible value for $\gamma_{min}$ is constrained by the
requirement to be sufficiently small that its Comptonization of the CMB
leads to the observed X-ray emission.
%($\epsilon_{0}\gamma_{min}^2 \delta_{min}^2 < \epsilon_{X}$, where
% $\epsilon_{X}$ is the energy of the {\sl Chandra} X-ray observations).
In Figure \ref{deltafig} we plot $\delta_{min}$ as a function of $\gamma_{min}$
for the extreme cases of an $e^{\pm}$ (solid curve) and an $e-p$
composition (dashed curve) for the knot WK7.8 of PKS 0637--752.
The observational parameters needed for the calculation are
taken by Chartas et al. (2000) and Schwartz et al. (2000) and can be found in
Appendix \ref{fs}.
For those values of $\gamma_{min}$ for which both solutions are permitted,
the hadronic knot is characterized by substantially higher $\delta_{min}$.
We now turn to the minimum knot power $L_{min}$ that corresponds to
$\delta_{min}$, which, following DA04, is written as:
%
\begin{eqnarray}
&L_{min}=L_{part}+L_{B}, \nonumber \\
&L_{part}=f_2(1+k_{pe}) \; \gamma_{min}^{2-p}\;\,\delta_{min}^{-(1+p)} \nonumber\\
&L_B=f_3\;\delta_{min}^4,
\label{eq:power}
\end{eqnarray}
%
where $L_{part}$ is the power in particles and $L_B$ is magnetic field power,
and $f_2,\; f_3$ are functions of observables used in DA04 (see Appendix
\ref{fs}). In Figure \ref{power}, we plot the total minimum knot power as
a function of $\gamma_{min}$ for an $e^{\pm}$ (thin solid curve) and an
$e-p$ (thick solid curve), for the case of knot WK7.8 of PKS 0637--752.
As can be seen, the minimum power for a hadronic knot flow (equal numbers
of $p$'s and $e$'s) is always larger
than that for a leptonic one. Interestingly, the leptonic power (dashed line)
needed to produce the observed emission in minimum total power
conditions is larger for a leptonic jet. This is mostly due to the lower
Doppler factor $\delta$ of leptonic relative to hadronic knot flows and
the strong dependence of the knot emission on $\delta$.
In conclusion, {\sl for a given choice of $\gamma_{min}$, the Doppler factor
and the jet power are only a function of the matter content, and can be
used to calculate the BC emission using the formalism of
\S \ref{section:bc}}.
\subsection{ Minimum Jet Power for Knot X-ray Emission
Due to Synchrotron \label{synch}}
If we assume a synchrotron interpretation for
the broadband knot spectrum, then, as DA04 point out,
the power needed to produce the second high energy component is
only a small fraction of the power
needed to produce the low energy synchrotron emission. This is mostly because
the electron radiative efficiency is much smaller at radio energies.
Therefore, we focus on the power needed to produce the radio emission.
Here, for a given observed synchrotron radio spectrum,
the Doppler factor at minimum power conditions cannot be uniquely defined.
Instead, it is the product $B\delta$ that can be derived.
%
Following DA04, their Eq. (8) can be written as
%
\begin{equation}
\delta\epsilon_B=f_4\left( { \gamma_{min}^{2-p}\over \tilde{k}_{eq}}\right)^
{2/(5+p)}, \label{f4}
\end{equation}
%
where $\epsilon_B$ is the magnetic field in units of the critical magnetic
field $B_{cr}=m_e^2c^3/e\hbar $, $f_4$ is a function of observables
(Appendix \ref{fs}), and $\tilde{k}_{eq}$ is the ratio of leptonic
to magnetic field energy density in the knot. For $\tilde{k}_{eq}=1$,
$y_{eq}=(\delta\epsilon_B)_{eq}= f_4 \gamma_{min}^{2(2-p)/(5+p)}$ and
$\delta\epsilon_B=y_{eq}/\tilde{k}_{eq}^{2/(5+p)}$.
The jet power flowing through the knot is
$L=\pi r_b^2 c \delta^2(u_{part}+u_B)$, where $u_B$ is the magnetic field
energy density, $u_{part}=(1+k_{pe})\tilde{k}_{eq}u_B$ is the total particle
energy density, and we have made the usual assumption $\delta=\Gamma$.
%
The jet power can then be written as
\begin{equation}
L=\pi r_b^2 c u_{B_{cr}} [(1+k_{pe})y_{eq}^{(5+p)/2}y^{-(1+p)/2}+y^2],
\label{eq:lum}
\end{equation}
where $y=\delta\epsilon_B$ and it is minimized for
\begin{equation}
y_{min}=(\delta\epsilon_B)_{min}=f_4 \gamma_{min}^{-2(p-2)/(5+p)}\left[(1+k_{pe})(p+1)\over 4 \right]^{2/(5+p)}.\label{eq:ymin}
\end{equation}
Setting $y=y_{min}$ in Equation (\ref{eq:lum}) we can calculate the jet power
as a function of $\gamma_{min}$ and the jet matter content expressed through $k_{pe}$.
\section{Applications\label{appl}}
We focus our attention on the {\sl Chandra}-detected superluminal quasars
PKS 0637--752 and 3C 273, whose superluminal nature guarantees that their
jets must be at relatively small viewing angles $\theta\sim 1/\Gamma$, where
$\Gamma\approx \beta_{app}$ is the minimum
bulk Lorentz factor that corresponds to the detected superluminal velocity
$ \beta_{app}$ (see e.g. Urry \& Padovani 1995). This suggests actual
`bridge' lengths $\sim \Gamma$ times longer than their projections on
the plane of the sky, resulting in separations of hundreds of kpc between
the core and the first knot.
Using the analytic expressions of \S \ref{section:power} we first estimate
the power $L_{lept}$ in leptons and Doppler factor $\delta$ required
to reproduce the emission from the knot at which the radiatively inefficient
`bridge' terminates, under minimum power conditions.
We then proceed to calculate the BC flux using the formalism
of \S \ref{section:bc} under two different, general assumptions.
{\sl Case A:}
The lepton
power $L_{lept}$ required in the knot is provided by the cold leptons in the
flow ($L_{e}=L_{lept}$). This then requires that only a minority of these leptons
get accelerated at the knot to create the X-ray producing population by tapping
a small fraction of the kinetic energy of the remaining `cold' leptons.
This represents one of the most optimistic cases for detecting the BC CMB emission
in the `bridge' region as it requires a large number of cold leptons in the
jet.
{\sl Case B:} The most conservative case for detecting the anticipated
BC emission is that in which the jet provides only the number of
leptons needed to produce the observed X-ray emission at knot; the leptons
are accelerated there using exclusively the energy of other agents such
as the magnetic field and/or the jet hadrons. In this case $L_{e}=
L_{lept}(p-2)/(p-1)\gamma_{min}$, and the cold lepton luminosity,
and subsequently the BC emission is lower by a factor of $\gamma_{min}
(p-1)/(p-2)$ compared to case A.
In both cases the jet composition can range from purely leptonic to
equal number of electrons and protons. In the following we discuss
only these two extreme cases as they bracket all other combinations
of protons and leptons.
In both cases we assume that the Doppler factor of the flow in the
`bridge' is not significantly different from that derived for the knot.
We defer a discussion of this assumption for \S
\ref{section:discussion}.
\subsection{PKS 0637--752\label{section:0637}}
In the case of PKS 0637--752, the optical flux from knot WK7.8, the knot
at which the radiatively inefficient `bridge' terminates, is clearly below
the extrapolation of the X-ray spectrum at lower frequencies, and,
as we argued in \S \ref{section:xmech}, this favors the EC
interpretation for the X-rays.
For knot WK7.8 we adopt $\gamma_{min}=20$, which, using Equation
(\ref{delta}) and the observational data in Appendix \ref{fs},
corresponds to minimum
power Doppler factors $\delta_{min}=17.4$ for a leptonic composition
and $\delta_{min}=27.8$ for a hadronic composition,
as can be seen in Figure \ref{deltafig}.
It also corresponds to a jet
minimum power $L_{min}=9.7 \times 10^{45} $ erg s$^{-1}$ for the leptonic
jet and $L_{min}=6.3 \times 10^{46} $ erg s$^{-1}$ for the hadronic jet.
The corresponding lepton power is $L_{lept}=3.7 \times 10^{45} $
erg s$^{-1}$ for the leptonic
jet and $L_{lept}=6.8 \times 10^{44} $ erg s$^{-1}$ for the hadronic one.
As can be seen in Figures \ref{deltafig} and \ref{power},
these numbers are only weakly affected by our choice of $\gamma_{min}$,
as long as $\gamma_{min}\gtrsim 10$.
To estimate the actual length of the `bridge',
we make the usual for superluminal sources assumption that the source is observed at an angle
$\theta=1/\Gamma$.
At the redshift of the source ($z=0.651$),
$1''$ corresponds to $6.9$ Kpc assuming standard cosmology
(Spergel et al. 2003) and the deprojected `bridge' length
is $l \approx $ 930 Kpc for a leptonic jet and $l \approx 1.5$
Mpc for a hadronic jet. We are discussing the implications of these length
estimates in \S \ref{section:discussion}.
To calculate the BC flux we proceed using the formalism
of \S \ref{section:bc} for an $e-p$ and an $e^{\pm}$ composition
for both cases A and B for the cold lepton power described above.
As can be seen in Figure \ref{0637}, in case A
the emission for a leptonic jet peaks at mid IR energies,
while that for a hadronic jet peaks at near IR - optical energies.
For both compositions the anticipated mid IR flux is above the {\sl Spitzer}
sensitivity limits; the hadronic case however violates the {\it HST} 3$\sigma$
detection limits for both a $0.5''$ and $0.1''$
thin jet. These limits are derived from 2001 STIS observations at 7219 \AA,
P.I. Meg Urry.
In the second, most conservative case, the BC emission is still above
the {\sl Spitzer} sensitivity limit for the two shorter wavelength bands.
However, the existing {\it HST} optical limits cannot be used to argue
against a hadronic jet in this case.
%{\bf As with the conventional SZ effect, it is important that the effects of the Comptonization on the CMB radiation be angularly resolved.}
The angular length of the `bridge' of PKS 0637--752 is $\sim 8''$. This is
easily resolved by {\sl HST}. Most importantly, at $\lambda=3.6-8.0 \; \mu m$,
the most likely band for the BC scattered emission to appear,
the {\sl Spitzer} angular resolution ($\sim 1''-3''$) is considerably
smaller than the `bridge' size, and we anticipate that {\sl Spitzer}
will resolve the BC emission along the `bridge'. Additional constraints
on the BC SED can be imposed by NICMOS {\ sl HST} observations at
$\lambda=1.6 \mu m$ (see Figure \ref{0637}).
\subsection{3C 273 \label{section:273}}
The X-ray emission from knot A of 3C 273 has been interpreted
as both synchrotron (Marshall et al. 2001) and EC off the CMB
(Sambruna et al. 2001). The discrepancy can be, at least partially,
resolved by studying the spectrum of the near IR - optical - UV emission.
{\sl HST} and {\it VLA} observations by Jester et al. (2002) showed that
the near IR - optical - UV spectrum of knot A is flatter than the radio -
near IR spectrum, indicating the presence of a high energy component, which,
they suggest, can be interpreted either as EC off the CMB or as a second,
synchrotron component from an independent high energy electron population.
The constraints on the jet power
and matter content depend on the interpretation we adopt.
{\sl EC off the CMB.}
We examine first the possibility that the X-rays in knot A
are due to EC off the CMB. We adopt again $\gamma_{min}=20$, which,
using Equation (\ref{delta}) and data from Marshall et al. (2001)
listed in Appendix \ref{fs}, corresponds to a minimum
power Doppler factor $\delta_{min}=16.6$ for a leptonic and
$\delta_{min}=26.5$ for a hadronic composition (these values
of $\delta$ are significantly higher those inferred from observations
of superluminal motions ( $\delta\sim 10$; Pearson et al. 1981) assuming
$\delta=\Gamma$,
a problem we discuss in \S \ref{section:discussion}).
This corresponds to a jet
minimum power $L_{min}=3.3 \times 10^{45} $ erg s$^{-1}$ for a leptonic
jet and $L_{min}=2.1 \times 10^{46} $ erg s$^{-1}$ for a hadronic jet.
The corresponding lepton power is $L_{lept}=1.6 \times 10^{45} $ erg s$^{-1}$
for a leptonic
jet and $L_{lept}=2.9 \times 10^{44} $ erg s$^{-1}$ for a hadronic jet.
As in the case of PKS 0637--752, these numbers are only weakly affected by
our choice of $\gamma_{min}$, as long as $\gamma_{min}\gtrsim 10$.
The deprojected `bridge' length, derived under the
assumption $\theta=1/\Gamma$ (at the redshift of the source, $z=0.158$,
$1''$ corresponds to $2.7$ Kpc) is $l \approx $ 580 Kpc for a leptonic jet
and $l \approx 930 $ Kpc for a hadronic jet.
Using these values for the Doppler factor and `bridge' length
we calculate the BC flux
for an $e-p$ and an $e^{\pm}$ composition
for both cases A and B for the cold lepton power.
As can be seen in Figure \ref{273}, the situation is similar with that
in PKS 0637--752.
The BC emission of a leptonic jet peaks at mid IR energies,
while that of a hadronic jet at near IR - optical energies.
In case A, the anticipated mid IR flux is well above the {\sl Spitzer}
sensitivity limits for both compositions;
the hadronic composition for case A, however, violates the {\it HST} 3$\sigma$
detection limits for both a $0.5''$ and $0.1''$
thin jet.
In the most conservative case B, the BC emission is still above
the {\sl Spitzer} sensitivity limit for the two shorter wavelength bands.
The expected optical emission is lower than the {\sl HST} 3$\sigma$
detection limits for both compositions, offering no additional constraints.
As in PKS 0637--752 , the large angular size of the `bridge' ($\sim 13''$)
guarantees that the BC emission of the `bridge' can be resolved
by {\sl Spitzer}.
{\sl Synchrotron.} We turn now to the synchrotron interpretation.
Using Equations (\ref{eq:lum}, \ref{eq:ymin}), we
calculate the jet minimum power as a function of $\gamma_{min}$.
An upper limit on $\gamma_{min}$ is set by the requirement that the
lowest energy electrons are energetic enough to produce the lowest observed
synchrotron emission
at $\nu_{s,min}=408 $ MHz (Foley \& Davis 1985) associated with knot A:
$ h \nu_{s,min}/m_ec^2 < y_{min}\gamma_{min}^2(1+z)$.
As can be seen in Figure \ref{synch273}, in the case of a leptonic
composition the minimum power required in the leptonic component
(thin dashed line)
drops as $\gamma_{min}$ increases, down to
$\approx 4\times 10^{44}$ erg s$^{-1}$, a factor of $\approx 4$ below the
minimum power in the leptonic component under the EC interpretation.
In the case of a hadronic composition, the minimum power required in the leptonic component
(thick dashed line) also decreases with $\gamma_{min}$; even at its lowest
value, however, it remains more powerful than the the minimum power
in the leptonic component under the EC interpretation.
Note that the difference in leptonic power
between the leptonic and the hadronic compositions decreases with
increasing $\gamma_{min}$ as the energy per electron gradually becomes
comparable to the proton rest mass energy.
Calculating the BC component in the synchrotron case requires a choice
of $\delta$.
Small values of $\delta$ will render the BC emission undetectable, since
$L_{BC}\propto \delta^3$.
Adopting the same values for $\delta$ as those derived in the
EC off the CMB case, results in detectable by {\sl Spitzer} BC emission for both leptonic
and hadronic jet compositions in case A.
In case B, the
BC flux can drop below the {\sl Spitzer} detectability limits, because the
large permitted values of $\gamma_{min}$ reduce the power of the cold lepton
beam ( $L_{e}=L_{lept}(p-2)/(p-1)\gamma_{min}$).
\section{Discussion and Conclusions \label{section:discussion}}
In the preceding sections we have formulated and examined the process
of bulk Comptonization of the CMB photons by the `cold' electrons of the
relativistic flows of {\sl Chandra-}detected extragalactic jets.
%We refer to this effect as the relativistic SZ effect, as it involves, like the conventional SZ effect, the same process, i.e. scattering of the CMB by energetic electrons.
The physical process of BC scattering of the CMB is certainly taking place,
as long as there are cold electrons propagating in a jet. One, however, has
to focus on systems where the BC signature is expected to be
($i$) strong and ($ii$) minimally contaminated by other emissions.
We argued that these conditions are favorably met in
the {\sl Chandra} - detected superluminal quasars PKS 0637--752 and 3C 273,
sources in which the jet radiates very weakly in radio, optical, and X-ray
energies for $\sim 10''$ between the core and the first knot. Based on the
fact that the `bridge' connecting the core to the first knot is only weakly
radiating, we argued that most of the leptons in this radiatively inefficient
section of the jet are transported practically cold. We then calculated the
power and Doppler factor of the flow required in the first knot to produce
the observed broadband spectrum under minimum energy conditions (DA04),
adopting EC scattering of the CMB
by relativistically moving plasma as the X-ray emission mechanism.
For 3C 273 we also examined the possibility that the X-ray emission
is due to synchrotron, a viable alternative for this source.
Using these power and Doppler factor estimates,
we calculated the BC emission for an $e-p$ and $e^{\pm}$ jet composition,
in each case considering two ways for energizing the electrons in the knot:
in case A we assumed that the lepton
power needed in the knot is provided by the cold leptons in the
beam alone, while in the most conservative case B that
the jet provides simply the number of leptons needed to produce the
knot emission (while the required power is provided by another agent,
i.e. protons, magnetic fields).
The resulting BC mid IR emission is above the {\sl Spitzer}
detectability limits in both cases and for both compositions,
and actual {\sl Spitzer} measurements of the `bridge' mid IR emission,
together with optical - near IR observations,
possibly including near IR polarimetry will
measure or, at worst, substantially constrain the
matter content of these jets. As we showed in \S \ref{section:bc},
a measurement of the polarization of the BC component,
together with an estimate of its peak frequency, can break the
degeneracy between the orientation of the jet and its bulk Lorentz factor.
Existing {\sl HST} limits for both 3C 273 and PKS 0637--752 already
disfavor case A $e-p$ models, in agreement with similar conclusions from
blazar studies Sikora \& Madejski (2000).
Additional constraints for pure $e-p$ jets come from the
large Lorentz factors required.
Although values of $\delta_{min}\sim 30$ are still compatible with the
apparent superluminal motions observed in some blazars
(e.g. Jorstad et al. 2002), the number of such
highly relativistic sources should not overproduce the
parent (misaligned) source population (e.g. Lister 2003).
Additionally, as we mentioned in \S \ref{section:0637},
\ref{section:273},
the large Doppler factors required for pure $e-p$ jets, suggest jet lengths
over $1$ Mpc long, a value barely compatible with
the largest jets of known radio galaxies
(e.g. Subrahmanyan, Saripalli, \& Hunstead 1996).
Measuring the BC emission of the 'bridge' can be used to measure
what fraction of the cold electrons propagating in the 'bridge' are picked up
by the particle acceleration mechanism in the knot and are accelerated to
high energies. This ``injection efficiency'' of particle acceleration
is, so far, a theoretically not well understood and observationally not
strongly constrained quantity (e.g. Gallant 2002).
A failure to detect the BC emission has definite implications for the matter
content of jets; however, given the limited sensitivity of {\it
Spitzer}, such a nondetection will leave several possibilities open. For 3C
273, where a synchrotron interpretation of the knot X-ray emission is a
plausible alternative, a non-detection would still be compatible with
synchrotron X-ray emission in Case B.
A possible source of the separate high energy electron population required
could be the decay of a neutron beam (Dermer \& Atoyan 2004b).
If the knot X-ray emission is EC in nature (this
mechanism is plausible for 3C 273 and favored for PKS 0637--752, as we discuss
in \S \ref{appl}),
deviations from equipartition (e.g. Kataoka \& Stawarz 2004) and/or
from the adopted orientation $\theta=1/\Gamma$ can reduce the BC flux below
the {\sl Spitzer} detection limits in case B, in which case we will be able
set an upper limit on the number of cold leptons in the jet as a function
of $\delta$.
%new stuff
We note here that the minimum power needed in the EC
model is already $ \gtrsim 10^{46}$ erg s$^{-1}$ and one cannot deviate
significantly from that without requiring jet powers
greater than the Eddington luminosity of a $10^9 \; M_{\odot} $ black hole.
If significant deviations from equipartition are needed to explain future
non-detection of the BC emission by {\sl Spitzer}, they will impose severe
constraints on the EC knot emission model,
particularly in the case of the higher power hadronic jets.
%end of new styff
Even if {\sl Spitzer} does not detect the BC emission, our method can still
be applied in the near future using the {\sl JWST}; this instrument is
expected to be $\sim 3$ orders of magnitude more sensitive than {\sl Spitzer}
and therefore probe much fainter levels of BC emission.
Additional constraints on the BC bridge emission can also be provided by
$\lambda=1.6 \mu m$ NICMOS {\sl HST} observations (see Figure \ref{0637}).
An assumption made in our calculations is that the Doppler factor of the
jet flow in the `bridge' between the core and the knot is the same as the
Doppler factor of the knot. This can happen if the flow
does not decelerate substantially at the knot, which can happen
if the knot is the site of an oblique shock.
%New stuff
Note, however, that if the knot is a separate entity propagating
in the jet, as in one of the cases examined by Tavecchio et al. (2003)
and Stawarz et al. (2004), the physical properties of the knot
will be unrelated to those of the ``bridge" and our method will not
applicable for the determination of any of the jet parameters.
%end of new stuff
In the case of PKS 0637--752,
VLBI observations of superluminal velocities with $v_{app}=17.8\pm 1 \,c $
in the core of the source (Lovell et al. 2000)
set limits for $\Gamma>17.8$, $\theta<6^{\circ}.4$, in agreement with the
Doppler factor $\delta=17.4$ derived from minimizing the jet power in an
$e^{\pm}$ jet. Similar values for the Doppler factors in the core and
the first knot were also derived by Tavecchio et al. (2004) using spectral
modeling of both the core and the first knot for blazars PKS 1510-089 and
1641+399 and by Jorstad \& Marscher 2004 for the superluminal
source 0827+243; these authors concluded that there is no bulk flow
deceleration between the core and the first knot in these sources.
The situation, however, seems different for 3C 273, where the superluminal
velocities observed in the VLBI core suggest $\Gamma\sim 10$ (e.g. Pearson et
al. 1981), significantly lower than the Doppler factor $\delta=16.6$
needed for an $e^{\pm}$ jet in minimum power. This leaves open the
possibility for a synchrotron interpretation of the X-ray emission,
which, as we argued in \S \ref{synch}, \ref{section:273}, does not allow
for a unique determination of $\delta$ and, subsequently, for a firm
estimate of the BC emission.
We have also assumed in our calculations the leptons in the radiatively
inefficient `bridge' between the core and the first knot are cold ($\langle
\gamma \rangle \approx 1$). Considering that electrons in the large
scale jet cannot cool down to $\gamma \sim 1$ in the flow frame once
accelerated to relativistic energies, the cold electrons in the ``bridge''
region, in the situation we envisage, are cold not because they have been
radiatively or adiabatically cooled down, but because they have never been
accelerated. In fact, this is in agreement with our understanding of
stochastic particle acceleration, according to which, generally, only a
small fraction of the available particles is accelerated to high energies.
Clearly there are non-thermal electrons, since the broadband
emission in the `bridge' of both PKS 0637--752 and 3C 273 is weak but not
absent. However the majority of the electrons must be at Lorentz factors
smaller than those required to obtain EC off the CMB at optical energies.
For $\Gamma\approx 10 $ this implies $\gamma \lesssim 4 $.
If the electrons are not really cold,
but they have a distribution around some small $\gamma$, because
the electron losses scale as $\gamma^2$, the power and the peak frequency
of the bulk Compton emission would be higher by the same factor.
Our assumption, therefore, represents a lower limit on the expected
BC emission.
As was discussed by Schwartz (2002) the knot X-ray emission due to EC
off the CMB will remain visible at the same flux level independent of
redshift. This is also the case for the BC emission from cold leptons
in relativistic
jets. This suggests an exciting possibility for jets that have
a very low radiative efficiency past the core (practically sources like
PKS 0637--752, but without the acceleration events that produce the broadband
non-thermal knot emission): their IR-optical BC emission will be detectable
independent of redshift, and it will be the only observable signature of
these otherwise invisible jets.
\appendix
\section{Functions and observables \label{fs}}
We reproduce here the functions $f_1, \, f_2, f_3, f_4$ that result from
the formalism of DA04. The function $f_1$ appears in Equation (\ref{delta})
- eq. (12) of DA04 - that relates the minimum power Doppler factor
$\delta_{min}$ to the knot
matter content as expressed through $k_{pe}$ and the minimum Lorentz factor
$\gamma_{min}$ of the electron distribution:
\begin{eqnarray}
& f_1=\left({9 (1+p) d_L^2 m_e c^2(1+z)^{(p-3)/2} \over
8 (p-2)c\sigma_{\tau}r_b^3}\right)^{1/(5+p)}
\left[ (f_{\epsilon_{EC}}^{EC})^{5+p}u_{B_{cr}}^{3-p} \over
(f_{\epsilon_s}^s)^4 u_{\star}^{p+5}\right]^{1/(1+p)(5+p)} \times \nonumber \\
& \epsilon_s^{2(3-p)/(p+1)(p+5)} \times
\left(2\epsilon_{\star}\over \epsilon_{EC}\right)^{(3-p)/2/(p+1)}.
\end{eqnarray}
$f_2$ arises in the calculation of minimum jet power, by
combining eq. (11) of DA04 for the particle energy content with the expression
$L_{part}=\pi r_b^2\Gamma^2c W'_{part}$ and setting $\Gamma=\delta=\delta_{min}$:
\begin{equation}
f_2={9\pi m_ec^2 d_L^2f_{\epsilon_{EC}}^{EC} \over 2 \sigma_{\tau}u_{\star}r_b(p-2)}
\left( (1+z)\epsilon_{EC}\over 2\epsilon_{\star}\right)^{(p-3)/2}.
\end{equation}
$f_3$ arises in the calculation of minimum jet power, by
combining eq. (10) of DA04 for the magnetic field with the expression
$L_{B}=\pi r_b^2\Gamma^2c W'_{B}$ and setting $\Gamma=\delta=\delta_{min}$:
\begin{equation}
f_3=\pi r_b^2 c u_{B_{cr}} \left( f_{\epsilon_s}^s u_{\star} \over
f_{\epsilon_{EC}}^{EC}u_{B_{cr}}\right)^{4/(p+1)}
\left(\epsilon_{EC} \over 2 \epsilon_s \epsilon_{\star}\right)^{2(3-p)/(p+1)}.
\end{equation}
$f_4$ appears in Equation \ref{f4} (eq. (8) of DA04):
\begin{equation}
f_4=\left[ 9 m_e c^2 d_L^2 f_{es}^s \epsilon_s^{(p-3)/2}\over
2 c \sigma_{\tau} u_{B_{cr}}^2 (p-2) r_b^3\right]^{2/(5+p)}.
\end{equation}
These functions depend on the following quantities:\\
$z$, the source redshift\\
$d_L$, the source luminosity distance\\
$p> 2$, the electron index, related to the radio and X-ray spectral index $\alpha=(p-1)/2$\\
$r_b$, the knot radius\\
$u_{B_{cr}}=B_{cr}^2/8\pi=7.75\; 10^{25}$ erg cm$^{-3}$, where $B_{cr}=m_e^2c^3/e\hbar$ is the
critical magnetic field\\
$u_{\star}=4\,10^{-13} (1+z)^4$ erg s$^{-1}$, the CMB photon energy density at
$z$\\
$\epsilon_{\star}=2.70kT_{CMB}(1+z)/m_ec^2=1.24\,10^{-9}(1+z)$, the dimensionless CMB photon energy at $z$ \\
$\epsilon_s=h\nu_s/m_ec^2$, where $\nu_s$ is the radio (synchrotron) observation frequency\\
$\epsilon_{EC}=h\nu_{EC}/m_ec^2$, where $\nu_{EC}$ is the X-ray (EC) observation frequency\\
$f_{\epsilon_s}^s=\nu_s f_{\nu_s}$, where $ f_{\nu_s}$ is the observed radio flux\\
$f_{\epsilon_{EC}}^{EC}=\nu_{EC} f_{\nu_{EC}}$, where $ f_{\nu_{EC}}$ is the observed X-ray flux.\\
In the case of PKS 0637--752 ($z=0.651$ corresponding to $d_L=1.9 \, 10^{28}$
cm or to $6.9$ kpc per arcsecond) the following values for the knot W7.8,
taken by Schwartz et al. (2000) and by Chartas et al. (2000),
were used in \S \ref{section:power} (used also by DA04):
$\nu_s=4.8$ GHz, $ f_{\nu_s}=54 mJy$, $\alpha_r=0.8$,
$\nu_{EC}=3.8 \, 10^{17}$ Hz, $ f_{\nu_{EC}}=6.6\,10^{-9}$ Jy. The knot W7.8
is not resolved in the optical and it is assumed that $r_b=1 $ Kpc,
corresponding to a knot diameter of $\sim 0.3 ``$.
For 3C 273, ($z=0.158$ corresponding to $d_L=2.3 \, 10^{27}$
cm or to $2.7$ kpc per arcsecond) the following values for knot A,
taken by Marshall et al. (2000)
were used in \S \ref{section:power}:
$\nu_s=1.65$ GHz, $ f_{\nu_s}=0.42 Jy$, $\alpha_r=0.76$,
$\nu_{EC}=2.4 \, 10^{17}$ Hz, $ f_{\nu_{EC}}=3.8 \,10^{-8}$ Jy. For knot A,
$r_b=1 $ Kpc is assumed , corresponding to a knot diameter of $\sim 0.75 ``$.
\begin{thebibliography}{}
\bibitem[Aharonian (2002)]{aharonian02} Aharonian, F. A. 2002, \mnras, 332, 215
\bibitem[Begelman \& Sikora (1987)]{begelman1987} Begelman, M. C. \& Sikora, M. 1987, \apj, 322, 650
\bibitem[Blandford \& Znajek (1977)]{blandford77} Blandford, R. D. \& Znajek, R. L. 1977, \mnras, 179, 433
\bibitem[Blandford \& Payne (1982)]{blandford82} Blandford, R. D. \& Payne, D. G. \mnras, 1982, 199, 883
\bibitem[Celotti \& Fabian (1993)]{celotti93}Celotti, A. \& Fabian, A. C. 1993, \mnras, 264, 228
\bibitem[Celotti, Ghisellini \& Chiaberge (2001)]{celotti01}Celotti, A., Ghisellini, G., \& Chiaberge M. 2001, \mnras, 321, L1
\bibitem[Chartas et al. (2001)]{chartas01} Chartas, G. et al. 2001, \apj, 542, 655
\bibitem[Cheung (2004)]{cheung04} Cheung, C. C. 2004, \apj, 600, L23
\bibitem[Dermer \& Atoyan (2002)]{dermer02} Dermer, C. D. \& Atoyan, A. M. 2002, \apj, 568, L81
\bibitem[Dermer & Atoyan (2004a)]{dermer04a} Dermer, C D. \& Atoyan, A. 2004a, \apj,
611, L9 (DA04)
\bibitem[Dermer & Atoyan (2004b)]{dermer04b} Dermer, C D. \& Atoyan, A. 2004b, \apj,
613, 151
%\bibitem[Ferrari (1998)]{ferrari98} Ferrari, A. 1998, \araa, 36, 539
\bibitem[Foley \& Davis (1985)]{foley85}
Foley, A. R. \& Davis, R. J. 1985, \mnras, 216, 679
\bibitem[Gallant (2002)]{gallant02} Gallant, Y. A. 2002, in
Relativistic Flows in Astrophysics, eds. A.W. Guthmann, M. Georganopoulos, A. Marcowith, \& K. Manolakou, Lecture Notes in Physics, vol. 589.
\bibitem[Georganopoulos \& Kazanas (2003)]{georganopoulos03} Georganopoulos, M. \& Kazanas, D. 2003, \apj, 589, L5
\bibitem[Georganopoulos \& Kazanas (2004)]{georganopoulos04} Georganopoulos, M. \& Kazanas, D. 2004, \apj, 604, L81
\bibitem[Ghisellini \& Celotti (2001)]{ghisellini01} Ghisellini, G., \& Celotti, A. 2001, \mnras, 327, 739
\bibitem[Gizani \& Leahy (2004)]{gizani04} Gizani, A. B. N. \& Leahy, J. P. 2004, \mnras, 350, 865
\bibitem[Hirotani (2005)]{hirotani04} Hirotani, K. 2005, \apj, 619, 73
\bibitem[Jester et al. (2002)]{jester02}Jester, S., R\"oser, H.-J., Meisenheimer, K., \& Perley, R. 2002, \aap, 385, L27
\bibitem[Jorstad et al. (2002)]{jorstad02} Jorstad, S. G., Marscher, A. P., Mattox, J. R., Aller, M. F., Aller, H. D., Wehrle, A. E., \& Bloom, S. D. 2002, \apj, 556, 738
\bibitem[Jorstad \& Marscher (2004)]{jorstad04}Jorstad, S. G. \& Marscher, A. P. 2004, \apj, 614, 615
\bibitem[Kataoka \& Stawarz (2004)]{kataoka04} Kataoka, J, \& Stawarz, \L. 2004, \apj, in press, also in astro-ph/0411042
\bibitem[kirk et al. (2000)]{kirk00} Kirk, J. G., Guthmann, A. W., Gallant, Y. A., \& Achterberg, A. \apj, 542, 235
\bibitem[K\"onigl (1989)] {konigl89} K\"onigl A. 1989, \apj, 342, 208
\bibitem[Lister (2003)]{lister03} Lister, M. L. 2003, ApJ, 599, L105
\bibitem[Lovell et al. (2000)]{lovell00} Lovell, J. E. J. et al. 2000, in Astrophysical Phenomena Revealed by Space VLBI, ed. H. Hirabayashi, P. G. Edwards, \& D. W. Murphy (Sagamihara: ISAS), 215
\bibitem[Marshall et al. 2001]{marshall01} Marshall, H. L. et al. 2001, \apj, 549, L167
\bibitem[Marshall et al. 2005]{marshall05} Marshall, H. L. et al. 2005, \apjs, 156, 13
\bibitem[Martel et al. 2003]{martel03} Martel, A. R. et al. 2003, \aj, 125, 2964
\bibitem[McNamara et al. 2005]{mcnamara03} McNamara, B. R., Nulsen, P. E. J.,
Wise, M. W., Rafferty, D. A., Carilli, C., Sarazin, C. L., \& Blanton, E. L. 2005, \nat, 433, 45
\bibitem[Moderski et al. (2004)]{moderski04} Moderski, R., Sikora, M.,
Madejski, G. M., \& Kamae, T. 2004, \apj, 611, 770
\bibitem[Omma \& Binney (2004)]{omma04} Omma, H. \& Binney, J. 2004, \mnras, 350, L13
\bibitem[Pearson et al. (1981)]{pearson81} Pearson, T. J. et al. 1981, \nat, 290, 365
\bibitem[Reynolds et al. (1996)]{reynolds91} Reynolds, C. S., Fabian, A. C., Celotti, A. \& Rees, M. J. 1996, \mnras, 283, 873
\bibitem[Rybicki \& Lightman (1979)]{rl79} Rybicki, G. R. \& Lightman, A. P. 1979, Radiative Processes in Astrophysics (Willey: New York)
\bibitem[Sambruna et al. 2001]{sambruna01} Sambruna, R. M., Urry, M. C., Tavecchio, F., Maraschi, L., Scarpa, R., Chartas, G., \& Muxlow, T. 2001, \apj, 549, L161
\bibitem[Sambruna et al. 2002]{sambruna02} Sambruna, R. M., Maraschi, L., Tavecchio, F., Urry, M. C., Cheung, C. C., Chartas, G., Scarpa, R., \& Gambill, J. K. 2002, \apj, 571, 206
\bibitem[Sambruna et al. 2004]{sambruna04} Sambruna, R. M., Gambill, J. K., Maraschi, L., Tavecchio, F., Cerutti, R., Cheung, C. C., Urry, M. C., \& Chartas, G. 2004, \apj, 608, 698
\bibitem[Schwartz et al. 2000]{schwartz2000} Schwartz, D. E. et al. 2000, ApJ, 540, L69
\bibitem[Schwartz (2002)]{schwartz02} Schwartz, D. E. 2002, \apj, 569, L23
\bibitem[Siemiginowska et al. (2002)]{siemiginowska02} Siemiginowska, A., Bechtold, J., Aldcroft, T. L., Elvis, M., Harris, D. E., \& Dobrzycki, A. 2002, \apj, 570, 543
\bibitem[Siemiginowska et al. (2003)] {siemiginowska03} Siemiginowska, A., Smith R. K., Aldcroft T. L., Schwartz, D. A., Paerels, F., \& Petric, A. O. 2003, \apj, 598, L15
\bibitem[Sikora et al. (1997)]{sikora97} Sikora, M., Madejski, G., Moderski, R. \& Poutanen, J. 1997, \apj, 484, 108
\bibitem[Sikora \& Madejski (2000)]{sikora00} Sikora, M. \& Madejski, G. 2000, \apj, 534, 109
\bibitem[Spregel (2003)]{spregel03} Spergel, D. N. et al. 2003, \apjs, 148, 175
% \bibitem[Stawarz \& Ostrowski (2002)]{stawarz02} Stawarz, \L. \& Ostrowski, M. 2002, \apj, 578, 763
\bibitem[Stawarz et al. (2004)]{stawarz04} Stawarz, \L., Sikora, M., Ostrowski, M., \& Begelman, M. C. 2004, \apj, 608, 95
\bibitem[Subrahmanyan, Saripalli, \& Hunstead (1996)]{sub96} Subrahmanyan, R., Saripalli, L., \& Hunstead, R. W. 1996, \mnras, 279, 257
\bibitem[Tavecchio et al. (2000)]{tavecchio00} Tavecchio, F., Maraschi, L., Sambruna, R. \& Urry, C. M., 2000, \apj, 544, L23
\bibitem[Tavecchio, Ghisellini, \& Celotti (2003)]{tavecchio03} Tavecchio, F., Ghisellini, G., \& Celotti, A. 2003, \aap, 403, 83
\bibitem[Tavecchio et al. (2004)]{taveccio04}Tavecchio, F., Maraschi, L.,
Sambruna, R. M., Urry, C. M., Cheung, C. C., Gambill, J. K., \& Scarpa, R.
2004, \apj, 614, 64
\bibitem[Wardle et al. (1998)] {wardle98} Wardle, J. F. C., Homan, D. C., Ojha, R., \& Roberts, D. H. 1998, \nat, 395, 457
\bibitem[Urry \& Padovani (1995)]{urry95}
Urry, C. M. \& Padovani, P. 1995, \pasp, 107, 803
\bibitem[Yuan et al. (2003)]{yuan03} Yuan, W., Fabian, A. C., Celotti, A., \& Jonker, P. G. 2003, \mnras, 346, L7
%---------------------------------------------------------------
\end{thebibliography}
\clearpage
\begin{figure}
\epsscale{0.8}
\plotone{f1.eps}
\caption{The bulk Compton luminosity of a cold electron beam of power
$L_e=10^{46}$ erg s$^{-1}$ propagating with a bulk Lorentz factor $\Gamma=10$
for a distance of 100 Kpc, through the CMB at $z=1$. The three curves
correspond to the luminosity given by Equation (\ref{app.lum}) for observing
angles $\theta=0$ (top curve), $\theta=1/\Gamma$ (middle curve), and $\theta=2/\Gamma$ (bottom curve). The three diamonds correspond to the simple
power and peak frequency estimates of Equations (\ref{eq:LBC}) and
(\ref{eq:fBC}).}
\label{BC}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f2.eps}
\caption{The bulk Compton emission polarization from a cold electron beam
as a function of observing angle for $\Gamma=20$ (solid line), $\Gamma=10$
(long dashed line), and $\Gamma=5$ (short dashed line). }
\label{polosi}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f3.eps}
\caption{The Doppler factor $\delta_{min}$ that minimizes the power required by
the knot Wk7.8 of PKS 0637--752 as a function of the permitted values of
the Lorentz factor $\gamma_{min}$ of the EED low energy cutoff.
The solid curve corresponds to an $e^{\pm}$ composition, while the dashed
curve to an $e-p$ composition. The observational data used in calculating
$f_1$ in Equation (\ref{delta}) are those used by DA04 and are taken by
Schwartz et al. (2000) and Chartas et al. (2001). Note that the range
of permitted $\gamma_{min}$ is a function of the knot matter content.}
\label{deltafig}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f4.eps}
\caption{The minimum power of a leptonic (hadronic) jet required by
the knot Wk7.8 of PKS 0637--752 is plotted with a solid thin (thick)
line as a function of the permitted values of
the Lorentz factor $\gamma_{min}$ of the EED low energy cutoff.
The power carried by radiating leptons in a minimum power configuration
is shown with a broken thin (thick) line for a leptonic (hadronic)
composition. The thin (thick) dot-dash lines show the Poynting flux in
the case of a leptonic (hadronic) jet, also in minimum power conditions.
The observational data needed in calculating
$f_1$ in Equation (\ref{delta}) are those used by DA04 and are
taken by Schwartz et al. (2000) and Chartas et al. (2001).
The two dotted lines show the Eddington luminosity that corresponds to
a $10^8\; M_{\odot}$ (lower dotted line) and $10^9\; M_{\odot}$. }
\label{power}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f5.eps}
\caption{The BC emission for a leptonic (hadronic) jet composition of
PKS 0637--752 is plotted with a solid thin (thick) line for case A,
in which the lepton power $L_{lept}$
required in the knot is provided by the cold leptons in the beam.
The dashed lines correspond to case B, in which the jet provides
simply the number of leptons needed in the knot, with the thin (thick) line
representing a leptonic (hadronic) jet composition. The {\sl Spitzer}
sensitivity limits, existing $3\sigma$ STIS {\sl HST} limits (see text),
and expected $3\sigma$ NICMOS {\sl HST} $\lambda=1.6 \mu m$ limits for
a 3-orbit exposure, assuming a $0.1''$ or a $0.5''$ jet radius are also
shown.}
\label{0637}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f6.eps}
\caption{Same as in Figure \ref{0637} for 3C 273. ACS {\sl HST} limits
derived from Martel et al. (2003).}
\label{273}
\end{figure}
\begin{figure}
\epsscale{0.8}
\plotone{f7.eps}
\caption{The total power required to reproduce the synchrotron emission
of knot A is 3C 273 is shown as a thin (thick) solid line for a leptonic (hadronic) jet composition. The thin (thick) broken line corresponds to the leptonic power for a leptonic (hadronic) composition.}
\label{synch273}
\end{figure}
\end{document}
A way to test our assumption that the jet flow does not decelerate
significantly in the knot is to detect the BC emission not only in the
`bridge', but also at the knot and its downstream region. The BC
component there is expected to be weaker, because at least a fraction
of the cold leptons will have to be accelerated to higher energies.
It is also expected to have a lower peak frequency, with the decrease
of the peak frequency depending on the actual flow deceleration.
This may be a unique way to probe the interplay of the
flow dynamics and particle acceleration: measuring the BC emission upstream
and downstream of the knot provides directly the particle acceleration
injection coefficient, the fraction of leptons that are picked up by the
acceleration process, a theoretically not understood and observationally not
constrained quantity in particle acceleration (e.g. Gallant 2002).