Stellar Dynamics and Star Formation Histories of z ∼ 1 Radio-loud Galaxies

The LEGA-C survey (van der Wel et al. 2016) is an ESO public spectroscopic survey with VLT/VIMOS (LeFevre et al. 2003) with the aim of obtaining high signal-to-noise ratio (S/N ∼ 20 Å⁻¹) continuum spectra of 0.6 < z < 1 galaxies. The full LEGA-C sample will consist of more than 3000 galaxies, K-band selected from the Muzzin et al. (2013) UltraVISTA survey in the 1.62 square degree region within the COSMOS field (Scoville et al. 2007). The spectral resolution is R = 2500, spanning the wavelength range from 6300 Å to 8800 Å. The current paper uses the Data Release II¹⁴ sample of 1989 galaxies observed during the first two years of LEGA-C observations. This sample is representative of the final one, which in turn is representative of the galaxy population at a given K-band flux density. That is, our sample selection is independent of galaxy color and morphology.

In this study, we use redshifts; stellar velocity dispersions; D_n(4000) break and Hδ absorption indices; nebular emission line equivalent widths, as well as physical parameters estimated from broad-band photometry (UV+IR star formation rates (SFR) and stellar masses). UV and IR luminosity based SFRs are estimated following Whitaker et al. (2012). Stellar masses are derived by using an initial mass function (Chabrier 2003), dust extinction (Calzetti et al. 2000), and stellar population libraries and exponentially declining SFR (Bruzual & Charlot 2003). For further details on the data reduction steps and the method used to derive the physical parameters, see van der Wel et al. (2016).

We use the observations at 3 GHz (10 cm) (PI: Vernesa Smolčić) covering the 2 square degree COSMOS field, obtained by the VLA radio interferometer. The observations were conducted between 2012 and 2014 with a total observation time of 384 hours, yielding a final mosaic with the angular resolution of 075 and a median rms of 2.3 μJy beam⁻¹. For further details, see Smolcic et al. (2017).

The current LEGA-C data set contains 1989 galaxies, out of which 322 are found to have a radio counterpart after cross-matching the LEGA-C survey catalog and the VLA 3 GHz 5.5σ catalog (Smolcic et al. 2017). The matching radius between LEGA-C and VLA coordinates is 07. The systematic offset in right ascension and declination is 01 and 003 respectively, while the random offset is 01. We convert radio continuum fluxes for cross-matched sources to luminosities, using VLA 3 GHz flux densities and LEGA-C spectroscopic redshifts, following the Condon (1992) luminosity relation. Implied radio-based star formation rates are then found from luminosities using Bell (2003) calibration of the radio-FIR correlation.

Figure 1 compares the radio luminosity, ${L}_{3\mathrm{GHz}}$ (and the corresponding star formation rate) with the UV+IR-based SFR for this cross-matched sample. The good correspondence between the star formation rate indicators is by construction, as the radio-based SFR is calibrated with the SFR derived from the IR luminosity. We see no evidence for a change in the far-infrared to radio luminosity ratio q, but this is not inconsistent with the evidence for such evolution (Magnelli et al. 2015; Delhaize et al. 2017) obtained from larger samples across a much larger redshift range than z = 0 to z = 1.

Zoom In Zoom Out Reset image size

Of interest for this study are the sources that are outliers from the one-to-one relation (black line); that is, the sources with excessive radio luminosities. We estimate the radio AGN luminosity by subtracting three times the radio luminosity expected on the basis of the UV+IR-based SFR from the observed radio luminosity. That is, we allow for a factor 3 scatter in the radio-SFR relation. Adopting the radio luminosity limit from Best et al. (2005), we select those 58 galaxies with radio AGN luminosities $\gt {10}^{23}$ W Hz⁻¹ as our radio-loud AGN sample (see Figure 1, red dashed curve).

We examine the optical spectra of the 58 radio-loud AGNs to distinguish between jet-mode (low-excitation radio galaxy, LERG) and radiative-mode (high-excitation radio galaxy, HERG) based on the presence or absence of strong high-excitation emission lines (Hine & Longair 1979; Laing 1994). We classify a radio galaxy as a LERG if there are no or only low-excitation (Balmer) emission lines (EW([O iii], [O ii], [Hβ]) > −5 Å). For systems with strong emission lines, we classify those with high [O iii]λ5007/Hβ and/or [O ii]λ3727/Hβ ratios (>1) as HERGs. In our sample of 58 radio-loud AGNs, 35 are classified as LERGs and seven as HERG with high confidence. The classification of 16 objects remains undetermined, as we were not able to discriminate with high confidence between jet/radiative-mode. For the remainder of this paper, we consider these objects as LERGs with 50% probability for the purpose of statistical calculations. In Figure 2, we summarize the results of our classification, which illustrates that almost 60% of our sample are LERGs. We list the properties of the 58 radio AGNs in Table 1.

Zoom In Zoom Out Reset image size

Table 1.  Physical Properties of the Observed Sub-sample of Radio-loud Galaxies

ID	R.A.	Decl.	zspec	L3 GHz [W Hz−1] · 1023	${\alpha }_{3}^{1.4}$ a	d [kpc]b	Flagc
126578	150.09805	2.281367	0.750	2.89 ± 0.18	−0.46 ± 0.14	⋯	1
140050	150.14168	2.446062	0.899	8.04 ± 0.42	−0.03 ± 0.11	⋯	0.5
209637	150.10378	2.507702	0.608	3.21 ± 0.17	−0.93 ± 0.19	17.69 ± 5.10	0
131063	150.35016	2.334642	0.667	12.27 ± 0.63	−0.41 ± 0.09	⋯	1
182797	150.39577	2.481532	0.903	2.22 ± 0.17	<−0.42	⋯	1
184916	150.42589	2.513884	0.679	6.11 ± 0.33	<1.77	50.00 ± 5.36	1
185625	150.42506	2.524558	0.984	1.83 ± 0.18	<−0.89	⋯	0.5
210716	150.17261	2.523343	0.69	12.21 ± 0.60	−0.80 ± 0.08	⋯	1
106926	150.29488	2.034494	0.955	77.68 ± 3.91	1.02 ± 0.09	⋯	0.5
108227	150.26643	2.049850	0.960	7.07 ± 0.39	−0.41 ± 0.31	⋯	0.5
109352	150.10892	2.063952	0.724	4.96 ± 0.28	$-{0.49}_{-0.17}^{+0.16}$	⋯	1
110509	150.26849	2.077003	0.667	6.64 ± 0.34	−0.06 ± 0.14	⋯	1
110805	150.17149	2.084074	0.729	3.77 ± 0.20	$-{0.82}_{-0.17}^{+0.16}$	⋯	0.5
113394	150.35657	2.117532	0.875	2.02 ± 0.16	<−0.48	⋯	1
128311	150.05669	2.301382	0.730	4.49 ± 0.24	−0.50 ± 0.28	19.12 ± 5.52	1
129746	150.02608	2.318864	0.941	3.40 ± 0.24	<0.13	⋯	0.5
205180	150.00731	2.453467	0.730	18.94 ± 0.96	$-{0.43}_{-0.08}^{+0.07}$	⋯	1
209377	150.02267	2.508070	0.746	85.76 ± 4.47	−0.79 ± 0.07	⋯	0
210031	150.02242	2.516584	0.679	5.82 ± 0.29	−0.93 ± 0.13	⋯	0
210739	150.00941	2.526713	0.733	4.09 ± 0.24	−0.49 ± 0.25	⋯	1
234067	149.85017	2.452237	0.714	29.15 ± 1.48	−0.35 ± 0.07	⋯	1
236682	149.87180	2.479084	0.734	20.77 ± 1.04	−0.78 ± 0.07	95.81 ± 5.53	1
236994	149.86151	2.484360	0.730	2.57 ± 0.16	−0.65 ± 0.29	109.57 ± 5.52	1
129631	149.98328	2.317157	0.934	9.60 ± 0.49	−0.61 ± 0.16	⋯	0.5
131657	149.95264	2.341849	0.945	2.33 ± 0.19	−1.06 ± 0.20	⋯	1
169076	149.78040	2.318275	0.677	1.96 ± 0.12	−0.43 ± 0.17	⋯	1
169901	149.79379	2.327209	0.893	2.64 ± 0.19	<−0.23	⋯	0
210564	149.91573	2.521326	0.729	6.55 ± 0.35	−0.47 ± 0.14	⋯	1
235394	149.76112	2.460729	0.671	4.92 ± 0.25	${0.09}_{-0.29}^{+0.28}$	⋯	1
235431	149.78880	2.466439	0.732	1.19 ± 0.11	<−0.61	⋯	0.5
237437	149.79221	2.489063	0.734	1.25 ± 0.11	<−0.56	⋯	1
111543	149.91492	2.094372	0.884	2.09 ± 0.17	<−0.54	⋯	0.5
113852	150.01424	2.123182	0.675	31.92 ± 1.59	−0.47 ± 0.07	35.63 ± 5.34	1
125257	150.06847	2.265479	0.979	3.62 ± 0.25	<−0.04	⋯	0.5
147270	149.87502	2.062635	0.847	12.63 ± 0.65	−1.28 ± 0.07	51.16 ± 5.81	1
151161	149.89481	2.109374	0.666	7.65 ± 0.43	−0.004 ± 0.14	⋯	1
161004	149.83919	2.226176	0.943	4.68 ± 0.28	−0.82 ± 0.32	24.00 ± 6.00	0.5
105328	149.90935	2.013062	0.848	1.84 ± 0.15	<−0.45	⋯	1
117992	149.94199	2.173145	0.688	2.31 ± 0.14	−0.28 ± 0.32	⋯	1
120120	149.99265	2.202235	0.629	2.21 ± 0.13	−0.59${}_{-0.25}^{+0.27}$	⋯	1
157229	149.74300	2.179562	0.631	15.21 ± 0.75	−0.64 ± 0.08	⋯	0.5
212718	150.07712	2.548955	0.890	70.58 ± 0.62	<−1.03	247.90 ± 5.84	1
203666	150.39935	2.794159	0.822	8.33 ± 0.43	−0.95 ± 0.19	⋯	0.5
215835	150.24612	2.585822	0.675	10.21 ± 0.53	0.34 ± 0.15		0.5
217020	150.16193	2.601267	0.893	2.07 ± 0.18	<−0.55	⋯	1
218725	150.04684	2.620396	0.736	2.94 ± 0.18	−1.03 ± 0.11	⋯	0
232020	150.01646	2.784381	0.983	15.48 ± 0.78	−0.69 ± 0.17	⋯	0.5
232196	149.98419	2.787762	0.853	4.63 ± 0.28	−0.96 ± 0.33	⋯	0.5
245325	149.88518	2.581121	0.694	4.47 ± 0.24	−0.13 ± 0.11	⋯	1
94215	150.68156	2.324819	0.978	2.04 ± 0.20	<−0.81	⋯	0
94982	150.63631	2.333361	0.609	11.47 ± 0.63	−0.32 ± 0.09	⋯	1
96860	150.66121	2.364529	0.826	79.72 ± 0.52	<−1.03	150.99 ± 5.77	1
182890	150.61380	2.484840	0.744	1.45 ± 0.11	<−0.41	⋯	1
183927	150.61508	2.500369	0.796	2.66 ± 0.17	<0.26	458.54 ± 5.69	1
225672	149.91795	2.701692	0.892	17.19 ± 0.85	−0.75 ± 0.1	⋯	0
233281	149.94615	2.801806	0.611	12.19 ± 0.61	−1.20 ± 0.23	59.58 ± 5.11	1
250117	149.77776	2.645909	0.737	22.20 ± 1.12	−0.52 ± 0.08	⋯	0.5
27265	150.14487	1.776603	0.733	13.33 ± 0.39	<2.21	⋯	1

Notes.

^aRadio spectral slope α inferred using ${\rm{S}}\propto {\nu }^{\alpha }$ at 1.4 GHz and 3 GHz. ^bLinear size (diameter) of the jet, with the errors estimated from the beam size (075). ^cClassification of galaxies onto LERG (1) or HERG (0). For conflicting indicators, we classify galaxies as 50% LERG (0.5).

Download table as:  ASCIITypeset image

We examine the VLA 3 GHz images of our sample for visual confirmation of radio jets. Forty-six radio-loud galaxies are consistent with point sources, but in 12 cases we identify extended morphologies. We show the VLA 3 GHz and Hubble Space Telescope (HST) F814W images, as well as the LEGA-C spectra of these 12 objects, in Figure 3.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Using the selection criteria defined in Section 2.4, we determine the fraction of jet-mode radio galaxies in the LEGA-C sample, considering both star-forming and quiescent galaxies (see Figure 4). Because we adopt the radio-loud AGN selection criteria from Best et al. (2005), we compare our fraction of jet-mode galaxies at a redshift range 0.6 < z < 1 with their fraction of jet-mode galaxies for present-day galaxies. At fixed mass, the fraction of jet-mode galaxies is on average five to 10 times higher at 0.6 < z < 1 when compared with the present-day universe. Furthermore, we notice the flattening of the power-law mass dependence for the highest mass bin. These findings are consistent with the measurements from Donoso et al. (2009).

Zoom In Zoom Out Reset image size

One possible evolutionary scenario is that galaxies grow in stellar mass by a factor of two from z ∼ 1 to the present, while conserving the fraction of jet-mode galaxies. This would imply that growth in stellar mass exceeds growth in black hole mass if we assume that the black hole mass is the only factor that sets the probability or frequency of becoming a radio galaxy. Alternatively, if BHs and stellar mass grow in lockstep (e.g., Calhau et al. 2017), then more frequent (or longer) radio-loud AGN phases at fixed BH mass could be understood by shorter cooling times at earlier cosmic times. The evolution in the fraction of jet-mode AGN in quiescent galaxies must be even faster than that of the general population, given that the fraction of quiescent galaxies is lower at z ∼ 1 than at the present day (e.g., Bell et al. 2004; Faber et al. 2007). We show the fraction of jet-mode AGN in quiescent galaxies in Figure 4, and see that above 10¹¹ M_⊙ the fraction of jet-mode galaxies reaches 20%.

The fraction of jet-mode AGNs also strongly depends on stellar velocity dispersion σ_*, which can be seen as a proxy for black hole mass. In Figure 5, we show that more than ∼10% of galaxies with σ_* ≳ 200 km s⁻¹ have jet-mode AGNs, reaching 20%–30% at σ_* ∼ 300 km s⁻¹. This behavior is only weakly dependent on star formation activity if at all, which suggests that jet-mode AGNs are not associated with quiescence, but rather with high σ_*. Below σ_* = 175 km s⁻¹, we find three jet-mode AGNs (one 100% and two 50% objects), whereas if the power-law trend seen at high σ_* were to continue to low σ_*, we would expect a total of 10 jet-mode AGNs in the three low-σ_* bins. The detection of only three jet-mode AGNs may suggest a threshold black hole mass of $\sim {10}^{8}\,{M}_{\odot }$ for jet-mode AGNs, as inferred from the local black hole mass–σ_* relation (Gebhardt et al. 2003; Beifiori et al. 2012; van den Bosch 2016). We argue that this threshold is not artificially introduced by our radio luminosity limit, as there is virtually no correlation between σ_* and radio luminosity as shown in Figure 6.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Assuming that the jet-mode fraction of 10%–30% at high σ_* can be interpreted as a duty cycle, we convert these fractions into the period (or frequency) at which galaxies turn on a jet-mode AGN. To make the conversion of the fraction of jet-mode AGNs into the period, we need a lifetime of an AGN jet. Examination of the jet structure morphology of the 12 objects in Figure 3 reveals that they are reminiscent of the classical Fanaroff-Riley I (FRI) type radio galaxy (Fanaroff & Riley 1974; Ledlow & Owen 1996). It has been argued that lifetimes evolve with redshift (Athreya & Kapahi 1998), but among our radio galaxies that are detected in the pre-existing 1.4 GHz VLA data (Schinnerer et al. 2010), we find that the spectral slopes, and therefore presumably the ages, are similar to local counterparts (see Table 1). The typical spectral slopes of local counterparts range between −1.3 < α < −0.5 with the average spectral slope being −0.8 (Condon 1992). Parma et al. (1998) find a correlation for FRI radio galaxies between the linear size of the jet and the synchrotron age of the jet as traced by the spectral slope, implying typical jet ages of about 100 Myr with an uncertainty of at most a factor 2. We therefore assume a lifetime of 2 × 100 Myr and show the resulting periods in Figure 5 on the left. We conclude that galaxies turn on a jet-mode AGN about once every Gyr, provided that it has a stellar velocity dispersion in excess of σ_* = 175 km s⁻¹, corresponding with a black hole mass of 10⁸ M_⊙. Remarkably, this is the same black hole mass threshold that has been shown to separate quiescent and star-forming galaxies in the local universe (Terrazas et al. 2016).

Our z ∼ 1 radio AGN typically live in red, quiescent galaxies (Figure 7) as was shown before by Smolčić et al. (2009), Simpson et al. (2013), and Rees et al. (2016).

Zoom In Zoom Out Reset image size

With the LEGA-C spectra, we can examine, for the first time, the detailed stellar population properties of z ∼ 1 radio galaxies. Figure 8 shows the Balmer absorption line index Hδ as a function of the 4000 Å break D_n(4000) (Bruzual et al. 1983; Balogh et al. 1999; Kauffmann et al. 2003b; P.-F. Wu et al. 2017, in preparation). Both of these parameters trace the recent star formation activity within the galaxy. This shows that the jet-mode AGNs have a strong D_n(4000) and weak Balmer absorption, which implies that these galaxies have been quiescent for more than a Gyr (Bruzual & Charlot 2003). That is, the observed AGNs with a lifetime of ∼100 Myr are not immediately responsible for the quenching of recent star formation. However, they may play a crucial role in maintaining quiescence, at least over the past 7 Gyr.

Zoom In Zoom Out Reset image size