Powerful relativistic jets in spiral galaxies

Abstract

The discovery of high-energy (E>100 MeV) gamma rays from Narrow-Line Seyfert
1 Galaxies (gamma-NLS1s) has confirmed the presence of powerful relativistic
jets in this class of active galactic nuclei (AGN). Although the jet emission
is similar to that of blazars and radio galaxies, gamma-NLS1s have some
striking differences: relatively small masses (10^6-10^8 M_sun), high accretion
rates (0.1-1 times the Eddington limit) and are generally hosted by spiral
galaxies. It is now possible to study a rather unexplored range of mass and
accretion rates of AGN with relativistic jets. Specifically, in this work I
present some results obtained by comparing a sample of blazars and gamma-NLS1s
with another sample of Galactic binaries with relativistic jets (stellar mass
black holes and neutron stars).

Full text

Page 1

arXiv:1107.2785v1 [astro-ph.HE] 14 Jul 2011
July 15, 20110:16WSPC/INSTRUCTION FILEfoschini˙1
International Journal of Modern Physics: Conference Series
c ? World Scientific Publishing Company
POWERFUL RELATIVISTIC JETS IN SPIRAL GALAXIES
LUIGI FOSCHINI
INAF Osservatorio Astronomico di Brera
Via E. Bianchi, 46 – 23807 Merate (LC) Italy
luigi.foschini@brera.inaf.it
Received 14 July 2011
Revised ... ... ...
The discovery of high-energy (E > 100 MeV) γ rays from Narrow-Line Seyfert 1 Galaxies
(γ-NLS1s) has confirmed the presence of powerful relativistic jets in this class of active
galactic nuclei (AGN). Although the jet emission is similar to that of blazars and radio
galaxies, γ-NLS1s have some striking differences: relatively small masses (106−8M⊙),
high accretion rates (0.1−1 times the Eddington limit) and are generally hosted by spiral
galaxies. It is now possible to study a rather unexplored range of mass and accretion rates
of AGN with relativistic jets. Specifically, in this work I present some results obtained
by comparing a sample of blazars and γ-NLS1s with another sample of Galactic binaries
with relativistic jets (stellar mass black holes and neutron stars).
Keywords: galaxies: jets – BL Lacertae objects: general – quasars: general – galaxies:
Seyfert – binaries: general
PACS numbers: 98.54.-h, 97.80.-d
1. Introduction
During the latest decades, we have observed powerful relativistic jets from active
galactic nuclei (AGN), Galactic binaries (stellar mass black holes, neutron stars
and cataclysmic variables), Gamma Ray Bursts (GRBs). Supersonic jets are also
visible in protostellar systems. Despite of the diffusion of jets in almost all types
of accreting astrophysical systems, it seemed that galaxies made some preference:
early observations of the host galaxies of AGN with relativistic jets displayed only
elliptical shapes2,26,29,38. Only a few counterexamples (i.e. jets in spiral galaxies)
were found in the past. The turning point occurred a couple of years ago with the
discovery – by means of the Fermi satellite – of GeV γ rays from narrow-line Seyfert
1 galaxies (γ-NLS1)13,14,15. This class of AGN is composed of sources generally
hosted by spiral (barred) galaxies9,10and, according to the common paradigm,
without jets. Therefore, although some early radio observations had suggested the
presence of some jet emission in a handful of cases23,27, the discovery of GeV γ
rays was somehow a surprise. It confirmed that also this class of AGN can develop
powerful relativistic jets, as like as blazars and radio galaxies (see Ref. 16 for a
review).
1

Page 2

July 15, 20110:16WSPC/INSTRUCTION FILEfoschini˙1
2 Luigi Foschini
Presently, seven γ-NLS1s are known and surely one of them is hosted by a spiral
galaxy40,1,16. Work is in progress to confirm the spiral hosts for the remaining
γ-NLS1s. The fact that no NLS1 is known to be at high redshiftasuggests that it
is very unlikely to find this type of AGN hosted by ellipticals.
The issue of the host is linked to how the mass can affect the generation and
the power of the jet and was studied in the past particularly with the use of the
radio loudness parameter (RL = f5 GHz/f440 nm), which in turn should indicate
the dominance of the synchrotron emission from the jet over the radiation from the
accretion disc (e.g. Ref. 29, 38). Being the blazars and radio galaxies (RL >> 10)
confined in a mass range of 108−10M⊙, while Seyferts and LINERs (RL < 10)
occupy the range 106−8M⊙, it seemed that the jet needed of a large mass to be
generated29,38.
The discovery of powerful relativistic jets in γ-NLS1 is now filling the region of
small masses and high radio loudness (see Ref. 17 for more details). In addition,
it opens also new interesting questions on the comparison of jets in AGN and in
Galactic binaries. This is the topic of the present work and is complementary to
Ref. 17.
2. Sample selection
I have used the same sample of AGN in Ref. 17, made of 9 BL Lac Objects, 30
flat-spectrum radio quasars (FSRQs) and 7 γ-NLS1s. For the present work, I have
added a sample of Galactic objects (X-ray Binaries, XRB), made of three stellar
mass black holes (GX 339−46, H1743−3228, V404 Cyg7) and two neutron stars
(Aql X-134, 4U 1728−3433). For XRB, the X-ray flux measured in different bands
(3 −9 keV in most of the cases), which samples the accretion radiation in Galactic
objects, has been converted into the 2 − 10 keV flux with webPIMMSbby using a
power-law model with photon index equal to 2. Then, the values have been corrected
according to Ref. 33 and scaled to the Eddington luminosity, calculated by assuming
a mass of 1.4M⊙for neutron stars and 5.8M⊙, 10M⊙, and 11.7M⊙for GX 339−419,
H1743−3228, and V404 Cyg19, respectively.
The radio emission has been converted into jet power by assuming the classical
relationship4,28according to which Pjet∝ L12/17
homogeneity with the AGN sample, I have used the Eq. (2) of Ref. 17, taking into
account that the radio emission from jets has a rather flat spectral index (αradio∼
0). The luminosities have been calculated by assuming the distances and the NH
reported in Refs. 8, 19, 33.
core,radio. For the sake of simplicity and
aThe 13th edition of the V´ eron-Cetty & V´ eron catalogue of quasars and AGN39contains only
NLS1s with z ? 1. This is also in agreement with the findings that these sources are very young31.
bhttp://heasarc.gsfc.nasa.gov/Tools/w3pimms.html

Page 3

July 15, 20110:16WSPC/INSTRUCTION FILE foschini˙1
Powerful relativistic jets in spiral galaxies3
−6−4
log(Ldisc/LEdd)
−20
30
32
34
36
38
log(Pjet M−1.4)
Fig. 1.
Squares indicate BL Lac Objects, circles are for FSRQs, and stars are γ-NLS1s. Triangles are
stellar mass black holes (in different states) and asterisks are for neutron stars. See the text for
more details.
Jet power per unit of M1.4vs accretion. Data from AGN are red, while XRB are blue.
3. Jet power and accretion
As already noted in Ref. 17 (Fig. 5), the addition of γ-NLS1s to the blazars makes
this sample of AGN directly comparable with a sample of jets from Galactic compact
objectsc. Blazars can be compared with stellar mass black holes, while γ-NLS1s
occupy the region of neutron stars. Therefore, in term of relativistic jets, γ-NLS1s
are the low mass counterpart of blazars so as neutron stars are the small mass
version of Galactic black holes. The difference of the jet power is about 12-13 orders
of magnitude, more than the mass difference between the two samples. To remove
the mass dependence, I have calculated the jet power per unit mass. Then, I have
searched for correlations in the two samples by using the ASURV v. 1.2 software
package30,12,25. In the case of AGN, I obtain:
AGN : logPjet
M
= (38.3 ± 0.1) + (0.55 ± 0.06)logLdisc
LEdd
(1)
and the results for the statistical tests are: Z = 4.1 (Pchance < 10−4) for the
Kendall’s test, while the Spearman’s method gives ρ = 0.6 (Pchance= 2 × 10−4).
In the case of Galactic binaries:
cIn Ref. 17, I compared Fig. 5 with an analogous figure in Ref. 8 (incidentally, still Fig. 5!).

Page 4

July 15, 2011 0:16WSPC/INSTRUCTION FILE foschini˙1
4 Luigi Foschini
−6−4
log(Ldisc/LEdd)
−20
30
32
34
36
38
log[Pjet M−1.4 (Ldisc/LEdd)−0.5]
Fig. 2.
while XRB are blue. Squares indicate BL Lac Objects, circles are for FSRQs, and stars are γ-
NLS1s. Triangles are stellar mass black holes (in different states) and asterisks are for neutron
stars. See the text for more details.
Jet power per unit of M1.4and (Ldisc/LEdd)0.5vs accretion. Data from AGN are red,
XRB : logPjet
M
= (34.8 ± 0.1) + (0.33 ± 0.03)logLdisc
LEdd
(2)
with Kendall’s Z = 6.7 and Spearman’s ρ = 0.7, both with Pchance< 10−4.
The difference in normalization of the jet power per unit mass between AGN
and XRB is now about three orders of magnitude. This difference, in excess of the
difference of mass ranges, has been already noted by several authors (e. g. Refs. 24,
32). Particularly, Ref. 24 explained it by invoking a non-linear dependence on the
mass, founding that Lradio∝ M17/12, in the hypothesis of the scale invariance of the
jet. Indeed, by dividing the jet power by M1.4, the difference in the normalization
between AGN and XRB is removed (Fig. 1) and the correlation is now:
AGN + XRB :logPjet
M1.4= (34.70 ± 0.07) + (0.44 ± 0.03)logLdisc
LEdd
(3)
with Kendall’s Z = 9.1 and Spearman’s ρ = 0.7, both with Pchance< 10−4.
Interestingly, Fig. 1 shows also that the most efficient engines are the γ-NLS1s,
with the greatest jet power per unit M1.4(also per unit mass): small, but nasty.
This is agreement with the theory that the central engine of AGN with jets is able
to distribute more or less equal power between accretion and ejection37,20and,
therefore, since γ-NLS1s have the greatest accretion, they are at the top end of the
distribution of specific jet power. In addition, the fact that γ-NLS1s and FSRQs have

Page 5

July 15, 2011 0:16 WSPC/INSTRUCTION FILE foschini˙1
Powerful relativistic jets in spiral galaxies5
very high jet powers and accretion rates, and that the Blandford-Znajek mechanism
tends to saturate in a radiation-pressure dominated regime22suggest that the jet
needs of an additional contribution, perhaps like Blandford-Payne or other hybrid
mechanisms17. The factor 0.44 of the accretion luminosity displayed in Eq. (3)
suggests that it is possible to remove the residual dependence shown in Fig. 1 by
dividing the specific jet power by the term (Ldisc/LEdd)0.5. This is indeed what is
shown in Fig. 2. This is a fact. Its interpretation is much more cumbersome.
This term indicates something that is not clear yet. It could point to the effi-
ciency η of a spherical accretion flow18and, thus, could emphasize the importance
of Bondi accretion3,36. Another option is that it is proportional to the efficiency η
of the accretion disc, which in turn would point to the presence of some advection.
Not necessarily an advection-dominated accretion flow (ADAF), but also a slim disc
for the greatest rates. The hypothesis is that AGN could have composite discs, as
like as XRB35, with a standard disc plus an advective part, which could be ADAF
or slim depending on the accretion rate. The possibility that (Ldisc/LEdd)0.5∝ η is
appealing, because the basic concept is that high disc efficiency means high number
of photons that can act both as direct and indirect (e.g. through the ionization of
the broad-line region, BLR) source of seed photons to efficiently cool the relativistic
electrons of the jet. It is known11that the radial scale of the BLR depends on
(Ldisc/LEdd)0.5and, then, the slope factor in Eq. (3) could be a direct indicator
of the seed richness of the environment close to the central black hole (and, this
time, indirectly linked to the radiative efficiency of the disc). Obviously, such an
argument could be of some interest for AGN, but not for XRB that do not have a
BLR.
On the other hand, it is worth noting that the factor 0.44 in Eq. (3) is just the
average of the factors 0.55 in Eq. (1) and 0.33 in Eq. (2). The physical explanation
of these factors could be indeed different, indicating a separated dependence on
the disc luminosity in Eddington units, although the final aim is the same. In the
case of AGN, the factor 0.55 could be linked to the BLR (and indirectly to the
disc), while in the case of XRB, the factor 0.33 could (directly) be the average
for the efficiency of advective discs in different states. It is necessary to take into
account that while FSRQs evolve into BL Lacs through gigayears, Galactic black
holes change their states on human time scales. γ-NLS1s are on a different evolutive
track as like as neutron stars are not linked to stellar mass black holes. Therefore,
if the different slopes indicate different relationships (direct or indirect) with the
radiative efficiency, this could be due to the different time scales.
Hypotheses, just hypotheses. How exactly the accretion disc contributes to the
jet power is still to be understood. However, despite all these differences and un-
knowns, Fig. 2 confirms the idea that the jets developed by all these systems are
basically the same thing.