«Jet-Induced Emission-Line Nebulosity and Star Formation in the High-Redshift Radio Galaxy 4C41.17 Geoﬀrey V. Bicknell1 Ralph S. Sutherland1 Wil J. ...»
September 15, 1999
Jet-Induced Emission-Line Nebulosity and Star Formation in the
High-Redshift Radio Galaxy 4C41.17
Geoﬀrey V. Bicknell1
Ralph S. Sutherland1
Wil J. M. van Breugel2
Michael A. Dopita1
George K. Miley4
1. ANU Astrophysical Theory Centre, Research School of Astronomy & Astrophysics, Australian
National University. Postal address: Mt Stromlo Observatory, Private Bag, Weston PO,
ACT, 2611, Australia. Email addresses: Geoﬀrey Bicknell: Geoﬀ.Bicknell@anu.edu.au;
Ralph Sutherland: email@example.com; Michael Dopita: firstname.lastname@example.org
2. Institute of Geophysics & Planetary Physics, LLNL, Livermore, CA 94550. Email:
3. KPNO/NOAO, 950 N. Cherry Ave., PO Box 26732, Tucson, AZ 85726. Present address:
Dept. of Physics and Astronomy, The Johns Hopkins University, Baltimore, MD 21218.
4. Leiden Observatory, PO Box 9513, 2300 RA, Leiden, The Netherlands. Email:
email@example.com –2– ABSTRACT The high redshift radio galaxy 4C41.17 has been shown in earlier work to consist of a powerful radio source in which there is strong evidence for jet-induced star formation along the radio axis. We argue that nuclear photoionization is not responsible for the excitation of the emission line clouds along the axis of the radio source and we therefore construct a jet-cloud interaction model to explain the major features revealed by the detailed radio, optical and spectroscopic data of 4C41.17. The interaction of a high-powered (∼ 1046 ergs s−1 ) jet with a dense cloud in the halo of 4C41.17 produces shock-excited emission-line nebulosity through ∼ 1000 km s−1 shocks and induces star formation. The C iii] to C iv line ratio and the C iv luminosity emanating from the shock, imply that the pre-shock density in the line-emitting cloud is high enough (hydrogen density ∼ 1 − 10 cm−3 ) that shock initiated star formation could proceed on a timescale (∼ a few × 106 yrs), well within the estimated dynamical age (∼ 3 × 107 yrs) of the radio source. The star formation eﬃciency in the shocked cloud is ∼ 1%. Broad (FWHM ≈ 1100 − 1400 km s−1 ) emission lines are attributed to the disturbance of the gas cloud by a partial bow–shock and narrow emission lines (FWHM ≈ 500 − 650 km s−1 ) (in particular C ivλλ1548, 50) arise in precursor emission in relatively low metallicity gas.
The implied baryonic mass ∼ 8 × 1010 M of the cloud is high and implies that Milky Way size condensations existed in the environments of forming radio galaxies at a redshift of 3.8. Our interpretation of the data provides a physical basis for the alignment of the radio, emission-line and UV continuum images in some of the highest redshift radio galaxies and the analysis presented here may form a basis for the calculation of densities and cloud masses in other high redshift radio galaxies.
One of the most intriguing discoveries in the study of high-redshift radio galaxies (HzRG) has been that the rest-frame UV continuum emission from their parent galaxies is aligned with the non-thermal radio emission (McCarthy et al. 1987; Chambers, Miley, & van Breugel 1987). The nature of this continuum and ‘alignment eﬀect’ has remained unclear. In nearby radio galaxies evidence has been found for jet-induced star formation, scattered light from hidden quasar-like AGN and nebular re-combination continuum (Van Breugel et al. (1985);van Breugel & Dey (1993);Dey et al. (1996);Tadhunter, Dickson, & Shaw (1996);Dickson et al. (1995); di Serego Alighieri et al. (1989);Cimatti et al. (1996)). A good example of radio-aligned UV emission in a very high-redshift radio galaxy is 4C41.17 at z = 3.800, which has been extensively studied at optical and radio wavelengths (Chambers, Miley, & van Breugel 1990; Miley et al. 1992;
Carilli, Owen, & Harris 1994; Chambers et al. 1996). Recent HST observations have shown that the rest-frame UV morphology of 4C41.17 consists of four main regions, the brightest of which (4C41.17-NE) contains an edge-brightened bifurcated feature consisting of several compact knots located between the radio nucleus and a bright radio knot (Van Breugel et al. 1998).
Deep spectropolarimetric observations with the W. M. Keck Telescope by Dey et al. (1997) show that 4C41.17 is unpolarized between λrest ∼ 1400 ˚ − 2000 ˚, implying that scattered light A A does not dominate the aligned UV continuum. Instead, the observations show absorption lines and P-Cygni-like features that are similar to those seen in z ≈ 2 − 3 star forming galaxies and nearby Wolf-Rayet starburst systems. The possibility of jet-induced star formation in 4C41.17 and other HzRGs has been suggested before (De Young 1981; De Young 1989; Rees 1989; Begelman & Cioﬃ 1989; Chambers, Miley, & van Breugel 1990; Daly 1990) but until now has lacked suﬃcient observational basis. In this paper we revisit the jet-induced star formation scenario for 4C41.17 in the light of the new data that are now available, and present a self-consistent model in which interactions between jets and dense clouds in 4C41.17 produce both shock-excited line emission and induce star formation. As we show below, it is fortunate that both phenomena occur since information provided by the former process enables us to better constrain the parameters relating to the latter.
Throughout this paper we assume that H0 =50 km s−1 Mpc−1 and q0 =0.1. The luminosity distance dL, angular size distance dA and linear scale at the redshift of 4C41.17 (z = 3.800) are then 51.6 Gpc, 2.24 Gpc, and 10.8 kpc arcsec−1 respectively. We follow the notation of Chambers, Miley, & van Breugel (1990) and Carilli, Owen, & Harris (1994) in referring to the radio features (components, knots etc.).
2. HST Observations of 4C41.17 and the Relationship to the Radio Emission
emission in order to facilitate the following theoretical discussion.
The montage in Figure 1 shows three HST images in diﬀerent bands with the X-band radio images of Carilli, Owen, & Harris (1994) superimposed in the form of contours. The top image is a deep rest-frame UV image (F702W ﬁlter, λrest ∼ 1430˚; 6.0 hours exposure); the middle image A was acquired through the F569W ﬁlter, which includes Lyα(2.0 hours exposure); the bottom image is a Lyα image (LRF ﬁlter at λc ∼ 5830 ˚; 2.0 hours exposure). All of these images show strongly A aligned non-thermal and thermal components. The direct association of the radio components with both UV continuum and Lyα emission, together with the spectroscopic evidence for young stars from the Keck observations, strongly points to jet-induced star formation. In particular, the radio knot B2 (the second from the left in these images) is associated with the brightest Lyα region and the F702W and F569W images reveal an interesting bifurcated or oval feature (approximated with a 0.8 by 0.24 [ ∼ 8.6 × 2.6 kpc] oval or parabola, shown enlarged on the right of Figure 1, which we interpret as tracing the locus of newly formed stars. See Van Breugel et al.
(1998) for a more detailed discussion of these images.
In Figure 2 a 0.3 smoothed version of the F702W image is displayed. This brings out an additional star forming region to the South of the regions evident in the unsmoothed version.
Van Breugel et al. (1998) have estimated the star-formation rates in these regions from the UV luminosity, using the relationship between ultraviolet ﬂux and star formation rate determined by Conti, Leitherer, & Vacca (1996). The estimated star formation rates in the various regions are given in table 1.
Following Van Breugel et al. (1998) we adopt the following nomenclature for the components in the HST image: The NE component is the region of edge-brightened UV emission located on the core side of the bright radio knot B2; NEE is the more diﬀuse component to the East of this. NW is the UV component along the radio axis on the Western side of the radio core and S represents is the clumpy component to the South, revealed by the smoothed image.
The evidence for jet-induced line emission and star formation in the brightest UV emission region in 4C41.17 (4C41.17NE) is compelling and can be summarized as follows (Van Breugel
et al. 1998; Dey et al. 1997):
• The star formation rate per square kiloparsec in the four UV bright regions mentioned in the introduction is by far the greatest in 4C41.17NE (Van Breugel et al. 1998). (1996). The morphology of 4C41.17NE and its close proximity to the radio knot, B2, strongly indicate that star formation has been induced by the interaction between the northern jet of the radio source and the cloudy medium of the forming parent galaxy as expected in jet-induced star formation models (e.g. De Young (1989)). The random distribution and lower star formation rates in the 4C41.17S knots, which are comparable to hose of ‘Lyman-break’ galaxies (Steidel et al. 1996), suggests that star formation here is unaided by bowshocks from the radio jet.
• The HST Lyα image shows a bright arc-shaped feature near B2 at the apex of the –5– edge-brightened UV structure, suggestive of a strong shock at a location where the jet interacts with dense ambient gas. Such emission-line features near bright radio structures are also often seen in nearby radio galaxies (van Breugel et al. 1985; Tadhunter et al. 1994) and these have a similar interpretation.
• The kinematics of the Lyα emission is very much disturbed in the aligned component with velocities with respect to systematic ∼ 500 − 1400 km s−1 and velocity dispersions σ ∼ 300 − 600 km s−1 (Dey et al. 1997; Chambers, Miley, & van Breugel 1990), suggesting large (projected) velocities and strong turbulence caused by jet/cloud interactions.
It follows from the above three points that the emission lines from this galaxy are probably related either to the star-forming region or to emission from radiative cloud shocks rather than excitation by UV-X-ray emission from the active nucleus.
• The Keck spectra by Dey et al. (1997) show that emission-line gas associated with the components B1, B2 and B3 of the inner radio source 4C41.17 consists of two distinct kinematic components: relatively narrow lines for all species (Lyα, N v, Si ii, Si iv, C iv, He ii, and C iii) with FWHM ≈ 500 − 650 km s−1 (σ ≈ 220 − 270 km s−1 ), and broad Lyα and C iii] with FWHM ≈ 1100 − 1400 km s−1 (σ ≈ 470 − 600 km s−1 ). (Si iv is possibly broad also; however the estimate of the line width is complicated by associated absorption.) We assume that the narrow velocity components are related to the jet-driven radiative shocks, and the broad components by ﬁlaments pulled out of the cloud through the action of the Kelvin-Helmholtz instability at the jet-cloud interface. This is discussed further in § 3.
• The brightest Lyα emission is found on the same side (East) which has the outer hotspot of 4C41.17 closest the nucleus. This agrees with the general radio / EELR morphological asymmetry correlation seen in powerful FR-II radio galaxies (McCarthy, van Breugel, & Kapahi 1991), and suggests that the radio source has been impeded in this direction as a result of its encounter with relatively dense gas.
• As we noted in the introduction, the absence of any evidence for a polarized, scattered AGN continuum supports the notion that, in the case of 4C41.17, the active nucleus is not responsible for the extended UV emission.
4C41.17 is located at the center of a large Lyα halo (Chambers, Miley, & van Breugel 1990).
The passage of relativistic jets through such a halo will inevitably result in substantial jet–cloud interactions. In the case of the jet-cloud interaction evident near the radio knot B2 we suggest that a glancing incidence of the jet on the cloud causes a partial bow-shock to be driven in to the cloud. This is manifest through the associated shock-excited line emission and associated star formation in the bifurcated structure referred to above. The jet deposits much of its momentum –6– at this site and it continues onward to the knot B3 where the decelerated jet plasma accumulates as a radio “lobe”. In this section we estimate physical cloud and jet parameters implied by this interaction model and then consider other emission regions in the HST images.
An important feature of the spectroscopic observations of 4C41.17 is the three kinematically distinct components namely the broad emission lines (σ ∼ 470 − 600 km s−1 ), the narrow emission lines (σ ∼ 220 − 270 km s−1 ) and the narrow absorption lines (σ ∼ 170 − 340 km s−1 ). We suggest that these components arise in the following way (see Figure 4). The narrow emission lines have a velocity dispersion similar to the halo and are formed either by locally induced photoionization of halo gas or in the winds of newly formed stars. The natural location for the narrow absorption lines is in the atmospheres of the young stars and in some cases the narrow absorption and emission lines comprise a typical P-cygni-like proﬁle characteristic of winds from young stars see Dey et al. (1997). We suggest that the broad emission lines arise from shock-excited gas which has been signiﬁcantly disturbed by the jet-cloud bow shock. This phenomenon is also observed in low redshift radio galaxies (van Breugel et al. 1985; Tadhunter 1991).
Many of the observed narrow emission lines could be produced either in the shock or in the photoionized winds of the newly formed stars. An important exception is C iv which is weak in stars older than 3 × 106 yrs (Leitherer, Robert, & Heckman 1995). Moreover, when this line is present in emission in young stars, its strength is comparable to the absorption strength.
In 4C41.17 the C iv emission line strength dwarfs the absorption component and we therefore completely attribute this component of emission to the eﬀects of the radiative cloud shock. The C ivemission is narrow and this is a strong indication that most of the ﬂux from this line originates in the precursor material ahead of the cloud bow shock. As we show below this is consistent with the velocity ∼ 1000 km s−1 that we adopt for the normal component of this shock.