Type 1 low z AGN. I. Emission properties

TL;DR

This work constructs a large, SDSS DR7–based sample of 3{,}579 broad H$ extalpha$–selected type 1 AGN spanning $L_{ m bH extalpha}\nolinebreak[4] \\sim\\ olinebreak[4] 10^{40}-10^{44}$ erg s$^{-1}$ and $M_{ m BH} olinebreak[4] \sim 10^{6}-10^{9.5}$ M$_\\odot$, and builds rest-frame SEDs from 2.2 μm to 2 keV by cross-matching with GALEX, 2MASS, and ROSAT. The analysis shows that (i) the broad H$ extalpha$ FWHM distribution is independent of luminosity, (ii) the mean optical–UV SED matches a fixed luminous-quasar shape scaled linearly with $L_{ m bH extalpha}$ plus a host component, and (iii) a fixed 3% host fraction of inactive galaxies hosts an AGN, with hosts becoming bluer and more star-forming at higher $L_{ m bH extalpha}$. The optical–UV net AGN SED appears uniform with small dispersion consistent with thin-disk emission, while reddening and host contamination explain much of the observed scatter, and the primary driver of the X-ray/UV ratio is luminosity rather than $M_{ m BH}$ or $L/L_{ m Edd}$. These results illuminate how AGN and host light combine across wavelengths and support a unified view of AGN SEDs across a broad luminosity range, with implications for BH growth and host galaxy evolution.

Abstract

We analyze the emission properties of a new sample of 3,579 type 1 AGN, selected from the SDSS DR7 based on the detection of broad H-alpha emission. The sample extends over a broad H-alpha luminosity L_bHa of 10^40 - 10^44 erg s^-1 and a broad H-alpha FWHM of 1,000 - 25,000 km s^-1, which covers the range of black hole mass 10^6<M_BH/M_Sun<10^9.5 and luminosity in Eddington units 10^-3 < L/L_Edd < 1. We combine ROSAT, GALEX and 2MASS observations to form the SED from 2.2 mic to 2 keV. We find the following: 1. The distribution of the H-alpha FWHM values is independent of luminosity. 2. The observed mean optical-UV SED is well matched by a fixed shape SED of luminous quasars, which scales linearly with L_bHa, and a host galaxy contribution. 3. The host galaxy r-band (fibre) luminosity function follows well the luminosity function of inactive non-emission line galaxies (NEG), consistent with a fixed fraction of ~3% of NEG hosting an AGN, regardless of the host luminosity. 4. The hosts of lower luminosity AGN have a mean z band luminosity and u-z colour which are identical to NEG with the same redshift distribution. With increasing L_bHa the AGN hosts become bluer and less luminous than NEG. The implied increasing star formation rate with L_bHa is consistent with the relation for SDSS type 2 AGN of similar bolometric luminosity. 5. The optical-UV SED of the more luminous AGN shows a small dispersion, consistent with dust reddening of a blue SED, as expected for thermal thin accretion disc emission. 6. There is a rather tight relation of nuL_nu(2 keV) and L_bHa, which provides a useful probe for unobscured (true) type 2 AGN. 7. The primary parameter which drives the X-ray to UV emission ratio is the luminosity, rather than M_BH or L/L_Edd.

Figures (23)

• Figure 1: Examples of the range of H$\alpha$ and H$\beta$ profiles observed in the host-subtracted spectra. The black lines are the fits to the broad H$\alpha$ (8$^{\rm th}$ --10$^{\rm th}$ order GH functions) and to the narrow lines (4$^{\rm th}$-order GHs). Dotted lines mark the FWHM of the Broad H$\alpha$. (a) Seyfert 1.0, $F_{\lambda; \text{bH$\alpha$}}/F_{\lambda; \text{nH$\alpha$}}\simeq 3$: note the extended wings of the BL profile compared to a Gaussian, fit by high-order GH components. (b) Seyfert 1.5, $F_{\lambda; \text{bH$\alpha$}}/F_{\lambda; \text{nH$\alpha$}}\simeq 0.5$: the FWHM of the total H$\alpha$ profile is dominated by the narrow H$\alpha$, therefore this object is excluded from the SDSS quasar catalog and the Vanden Berk et al. (2006) sample. (c) Seyfert 1.9, $F_{\lambda; \text{bH$\alpha$}}/F_{\lambda; \text{nH$\alpha$}}\simeq 0.05$: such an object will be defined as a Seyfert 2 if only H$\beta$ is available, despite an EW$_{\rm{b\text{H$\alpha$}}}=28\text{\AA}$, which may indicate this is an unobscured broad line AGN. (d) A type 1 AGN with no clear transition in the profile between $F_{\lambda; \text{bH$\alpha$}}$ and $F_{\lambda; \text{nH$\alpha$}}$. The Balmer NL profiles are based on [O iii], since [N ii], [S ii] and [O i] are undetectable.
• Figure 2: The selection criteria effects on the distribution of the T1 objects (small gray dots) in the broad H$\alpha$ flux vs. FWHM plane. Large coloured markers are used to plot the 4% of the T1 objects which are just above the following three selection thresholds. (Upper panel) The $\Delta F / \sigma_{\rm data} > 2.5$ selection criterion. At $F_{\rmn{bH\alpha}}<10^{-14}\ \rm erg\ s^{-1}\ cm^{-2}$, this selection increases the minimal detectable $\Delta {\rm v}_{1000}$ from 1 to 3. (Middle panel) The $F_{\lambda, \rm{b\text{H$\alpha$}}} / \epsilon_{\lambda} > 2$ selection criterion. At $F_{\rmn{bH\alpha}}<10^{-14}\ \rm erg\ s^{-1}\ cm^{-2}$, this selection decreases the maximal detectable $\Delta {\rm v}_{1000}$ from 10 to 3. The (Lower panel) S/N $>10$ selection criterion. This selection criterion does not form any boundary in the $F_{\rmn{bH\alpha}}$ vs. $\Delta {\rm v}_{1000}$ plane. The T1 objects extend down to $\Delta {\rm v}_{1000}=1$, in particular at $F_{\rmn{bH\alpha}}<10^{-13.5}\ \rm erg\ s^{-1}\ cm^{-2}$, and objects with $\Delta {\rm v}_{1000}<1$ are excluded ($\sim 4\%$, see text). The $\Delta {\rm v}_{1000} < 25$ criterion has practically no effect, as objects with $\Delta {\rm v}_{1000}>15$ are extremely rare. Our $F_{\rmn{bH\alpha}}$ detection limit is lowest for $\Delta {\rm v}_{1000} \sim 3$ objects.
• Figure 3: The detection fractions of the T1 objects in the GALEX (UV, blue), 2MASS (NIR, red) and ROSAT (X-ray, black) surveys, as a function of $F_{\rmn{bH\alpha}}$, for extended and point like sources. Detections in all bands of each survey is required. Solid/dashed lines represent extended/point source objects. The lower panel shows the number of T1 point-like and extended objects per 0.5 decade $F_{\rmn{bH\alpha}}$ bin. The GALEX detection fraction is relative to the 89% of T1 objects observed. Note the nearly complete 2MASS and GALEX detections. When analyzing the X-ray luminosities (§ 3.6 and § 3.9), we utilize only $F_{\rmn{bH\alpha}}>10^{-13.5}\ \rm erg\ s^{-1}\ cm^{-2}$ objects, where the detection fraction is $>50\%$. Note that most of the lowest $F_{\rmn{bH\alpha}}$ T1 AGN reside in extended objects. Their bright host enables them to pass the SDSS flux limit, despite their low $F_{\rmn{bH\alpha}}$.
• Figure 4: The $L_{\rmn{bH\alpha}}$ versus $z$ distribution for the T1 sample at $z<0.2$ (3 077 objects). Empty circles mark the 423 objects (14%) that also appear in the SDSS quasar catalog (QCV). Note that the QCV sample terminates at $L_{\rmn{bH\alpha}}$ of a few $10^{42}\ \rm erg\ s^{-1}$, while the T1 sample extends further down to $10^{40}\ \rm erg\ s^{-1}$.
• Figure 5: Comparison of $L_{\rmn{bH\alpha}}$ ranges of different broad line AGN samples. We include only $z<0.35$ objects of the QCV sample, where H$\alpha$ is observable. Although the T1 sample extends down by two orders of magnitude in $L_{\rmn{bH\alpha}}$, compared to QCV, it does not extend as low as the Ho et al. (1997a) sample, which reaches $L_{\rmn{bH\alpha}}<10^{38.5}\ \rm erg\ s^{-1}$ (magnified ten-fold for clarity).