Context. The most highly accreting quasars are of special interest in studies of the physics of active galactic nuclei (AGNs) and host galaxy evolution. Quasars accreting at high rates (L/LEdd ∼ 1) ...hold promise for use as “standard candles”: distance indicators detectable at very high redshift. However, their observational properties are still largely unknown. Aims. We seek to identify a significant number of extreme accretors. A large sample can clarify the main properties of quasars radiating near L/LEdd ∼ 1 (in this paper they are designated as extreme Population A quasars or simply as extreme accretors) in the Hβ spectral range for redshift ≲0.8. Methods. We use selection criteria derived from four-dimensional Eigenvector 1 (4DE1) studies to identify and analyze spectra for a sample of 334 candidate sources identified from the SDSS DR7 database. The source spectra were chosen to show a ratio RFeII between the FeII emission blend at λ4570 and Hβ, RFeII > 1. Composite spectra were analyzed for systematic trends as a function of Fe II strength, line width, and OIII strength. We introduced tighter constraints on the signal-to-noise ratio (S/N) and RFeII values that allowed us to isolate sources most likely to be extreme accretors. Results. We provide a database of detailed measurements. Analysis of the data allows us to confirm that Hβ shows a Lorentzian function with a full width at half maximum (FWHM) of Hβ ≤ 4000 km s−1. We find no evidence for a discontinuity at 2000 km s−1 in the 4DE1, which could mean that the sources below this FWHM value do not belong to a different AGN class. Systematic OIII blue shifts, as well as a blueshifted component in Hβ are revealed. We interpret the blueshifts as related to the signature of outflowing gas from the quasar central engine. The FWHM of Hβ is still affected by the blueshifted emission; however, the effect is non-negligible if the FWHM Hβ is used as a “virial broadening estimator” (VBE). We emphasize a strong effect of the viewing angle on Hβ broadening, deriving a correction for those sources that shows major disagreement between virial and concordance cosmology luminosity values. Conclusions. The relatively large scatter between concordance cosmology and virial luminosity estimates can be reduced (by an order of magnitude) if a correction for orientation effects is included in the FWHM Hβ value; outflow and sample definition yield relatively minor effects.
Improving our understanding of the nuclear properties of high-Eddington-ratio (
λ
Edd
) active galactic nuclei (AGN) is necessary since at this regime the radiation pressure is expected to affect the ...structure and efficiency of the accretion disc-corona system. This may cause departures from the typical nuclear properties of low-
λ
Edd
AGN, which have been largely studied so far. We present here the X-ray spectral analysis of 14 radio-quiet,
λ
Edd
≳ 1 AGN at 0.4 ≤
z
≤ 0.75, observed with
XMM-Newton
. Optical/UV data from simultaneous Optical Monitor observations have also been considered. These quasars were selected to have relatively high values of black hole mass (
M
BH
∼ 10
8 − 8.5
M
⊙
) and bolometric luminosity (
L
bol
∼ 10
46
erg s
−1
) in order to complement previous studies of high-
λ
Edd
AGN at lower
M
BH
and
L
bol
. We studied the relation between
λ
Edd
and other key X-ray spectral parameters, such as the photon index (Γ) of the power-law continuum, the X-ray bolometric correction (
k
bol, X
), and the optical/UV-to-X-ray spectral index (
α
ox
). Our analysis reveals that, despite the homogeneous optical and supermassive black hole accretion properties, the X-ray properties of these high-
λ
Edd
AGN are quite heterogeneous. We indeed measured values of Γ between 1.3 and 2.5, at odds with the expectations based on previously reported Γ −
λ
Edd
relations, for which Γ ≥ 2 would be a ubiquitous hallmark of AGN with
λ
Edd
∼ 1. Interestingly, we found that ∼30% of the sources are X-ray weak, with an X-ray emission about a factor of ∼10 − 80 fainter than that of typical AGN at similar UV luminosities. The X-ray weakness seems to be intrinsic and not due to the presence of absorption along the line of sight to the nucleus. This result may indicate that high-
λ
Edd
AGN commonly undergo periods of intrinsic X-ray weakness. Furthermore, results from follow-up monitoring with
Swift
of one of these X-ray weak sources suggest that these periods can last for several years.
ABSTRACT We present an analysis of 43 years (1972 to 2015) of spectroscopic observations of the Seyfert 1 galaxy NGC 5548. This includes 12 years of new unpublished observations (2003 to 2015). We ...compiled about 1600 Hβ spectra and analyzed the long-term spectral variations of the 5100 continuum and the Hβ line. Our analysis is based on standard procedures, including the Lomb-Scargle method, which is known to be rather limited to such heterogeneous data sets, and a new method developed specifically for this project that is more robust and reveals a ∼5700 day periodicity in the continuum light curve, the Hβ light curve, and the radial velocity curve of the red wing of the Hβ line. The data are consistent with orbital motion inside the broad emission line region of the source. We discuss several possible mechanisms that can explain this periodicity, including orbiting dusty and dust-free clouds, a binary black hole system, tidal disruption events, and the effect of an orbiting star periodically passing through an accretion disk.
Context. When can an active galactic nucleus (AGN) be considered radio loud (RL)? Following the established view of the AGNs inner workings, an AGN is RL if associated with relativistic ejections ...emitting a radio synchrotron spectrum (i.e., it is a “jetted” AGN). In this paper we exploit the AGN main sequence that offers a powerful tool to contextualize radio properties. Aims. If large samples of optically-selected quasars are considered, AGNs are identified as RL if their Kellermann’s radio loudness ratio RK > 10. Our aims are to characterize the optical properties of different classes based on radio loudness within the main sequence and to test whether the condition RK > 10 is sufficient for the identification of RL AGNs, since the origin of relatively strong radio emission may not be necessarily due to relativistic ejection. Methods. A sample of 355 quasars was selected by cross-correlating the Very Large Array Faint Images of the Radio Sky at Twenty-Centimeters survey (FIRST) with the twelfth release of the Sloan Digital Sky Survey Quasar Catalog published in 2017. We classified the optical spectra according to their spectral types along the main sequence of quasars. For each spectral type, we distinguished compact and extended morphology (providing a FIRST-based atlas of radio maps in the latter case), and three classes of radio loudness: detected ( specific flux ratio in the g band and at 1.4 GHz, R′K < 10 R K ′ <10 $ R_{\rm K}^{\prime} < 10 $ ), intermediate ( 10 ≤ R′K < 70 10≤ R K ′ <70 $ 10 \le R_{\rm K}^{\prime} < 70 $ ), and RL ( R′K ≥ 70 R K ′ ≥70 $ R_{\rm K}^{\prime} \geq 70 $ ). Results. The analysis revealed systematic differences between radio-detected (i.e., radio-quiet), radio-intermediate, and RL classes in each spectral type along the main sequence. We show that spectral bins that contain the extreme Population A sources have radio power compatible with emission by mechanisms ultimately due to star formation processes. RL sources of Population B are characteristically jetted. Their broad Hβ profiles can be interpreted as due to a binary broad-line region. We suggest that RL Population B sources should be preferential targets for the search of black hole binaries, and present a sample of binary black hole AGN candidates. Conclusions. The validity of the Kellermann’s criterion may be dependent on the source location along the quasar main sequence. The consideration of the main sequence trends allowed us to distinguish between sources whose radio emission mechanisms is jetted from the ones where the mechanism is likely to be fundamentally different.
Context.
This is the second work dedicated to the observed parallelism between galaxy clusters (GCs) and early-type galaxies (ETGs). The focus is on the distribution of these systems in the scaling ...relations (SRs) observed when effective radii, effective surface brightness, total luminosities, and velocity dispersions are mutually correlated.
Aims.
Using the data of the Illustris simulation we speculate on the origin of the observed SRs.
Methods.
We compare the observational SRs extracted from the database of the WIde-field Nearby Galaxy-cluster Survey with the relevant parameters coming from the Illustris simulations. Then we use the simulated data at different redshift to infer the evolution of the SRs.
Results.
The comparison demonstrate that GCs at
z
∼ 0 follow the same log(
L
)−log(
σ
) relation of ETGs and that both in the log(⟨
I
⟩
e
)−log(
R
e
) and log(
R
e
)−log(
M
*
) planes the distribution of GCs is along the sequence defined by the brightest and massive early-type galaxies (BCGs). The Illustris simulation reproduces the tails of the massive galaxies visible both in the log(⟨
I
⟩
e
)−log(
R
e
) and log(
R
e
)−log(
M
*
) planes, but fails to give the correct estimate of the effective radii of the dwarf galaxies that appear too large and those of GCs that are too small. The evolution of the SRs up to
z
= 4 permits to reveal the complex evolutionary paths of galaxies in the SRs and indicate that the line marking the zone of exclusion, visible in the log(⟨
I
⟩
e
)−log(
R
e
) and the log(
R
e
)−log(
M
*
) planes, is the trend followed by virialized and passively evolving systems.
Conclusions.
We speculate that the observed SRs originate from the intersection of the virial theorem and a relation
L
=
L
0
′
σ
β
where the luminosities depend on the star formation history.
Classification of the optical spectra of active galactic nuclei (AGN) into different types is currently based on features such as line widths and intensity ratios. Although well founded on AGN ...physics, this approach involves some degree of human oversight and cannot scale to large datasets. Machine learning (ML) tackles this classification problem in a fast and reproducible way, but is often (and not without reason) perceived as a black box. However, ML interpretability and are active research areas in computer science that are providing us with tools to mitigate this issue. We apply ML interpretability tools to a classifier trained to predict AGN types from spectra. Our goal is to demonstrate the use of such tools in this context, obtaining for the first time insight into an otherwise black box AGN classifier. In particular, we want to understand which parts of each spectrum most affect the predictions of our classifier, checking that the results make sense in the light of our theoretical expectations. We trained a support-vector machine on 3346 high-quality, low-redshift AGN spectra from SDSS DR15. We considered either two-class classification (type 1 versus 2) or multiclass (type 1 versus 2 versus intermediate-type). The spectra were previously and independently hand-labeled and divided into types 1 and 2, and intermediate-type (i.e., sources in which the Balmer line profile consists of a sharp narrow component superimposed on a broad component). We performed a train-validation-test split, tuning hyperparameters and independently measuring performance via a variety of metrics. On a selection of test-set spectra, we computed the gradient of the predicted class probability at a given spectrum. Regions of the spectrum were then color-coded based on the direction and the amount by which they influence the predicted class, effectively building a saliency map. We also visualized the high-dimensional space of AGN spectra using t-distributed stochastic neighbor embedding (t-SNE), showing where the spectra for which we computed a saliency map are located. Our best classifier reaches an F-score of 0.942 on our test set (with 0.948 precision and 0.936 recall). We computed saliency maps on all misclassified spectra in the test set and on a sample of randomly selected spectra. Regions that affect the predicted AGN type often coincide with physically relevant features, such as spectral lines. t-SNE visualization shows good separability of type 1 and type 2 spectra. Intermediate-type spectra either lie in-between, as expected, or appear mixed with type 2 spectra. Misclassified spectra are typically found among the latter. Some clustering structure is apparent among type 2 and intermediate-type spectra, though this may be an artifact. Saliency maps show why a given AGN type was predicted by our classifier resulting in a physical interpretation in terms of regions of the spectrum that affected its decision, making it no longer a black box. These regions coincide with those used by human experts, for example relevant spectral lines, and are even used in a similar way; the classifier effectively measures the width of a line by weighing its center and its tails oppositely.
Context.
We test the scenario according to which the broad absorption line (BAL) phenomenon in quasars (QSOs) is not a temporary stage of their life. In this scenario, the BAL effect acts only if the ...line of sight is within a spatially limited and collimated massive outflow cone covering only a fraction of the sky from the point of view of the nucleus.
Aims.
The aim is to understand the theoretical mechanism behind the massive outflow in BAL QSOs, which is important for modelling the impact of quasars on the star formation rate in the host galaxy, and, subsequently, on the galaxy evolution.
Methods.
We applied the specific theoretical model of dust-driven wind that was developed to explain broad emission lines. The model has considerable predictive power. The 2.5D version of the model called failed radiatively accelerated dusty outflow (FRADO) includes the formation of fast funnel-shaped outflow from the disk for a certain range of black hole masses, Eddington ratios, and metallicities. We now interpret BAL QSO as sources that are viewed along the outflowing stream. We calculated the probabilities of seeing the BAL phenomenon as functions of these global parameters, and we compared these probabilities to those seen in the observational data. We included considerations of the presence or absence of obscuring torus.
Results.
Comparing our theoretical results with observational data for a sample of QSOs consisting of two sub-populations of BAL and non-BAL QSOs, we found that in the model and in the data, the BAL phenomenon mostly occurs for sources with black hole masses higher than 10
8
M
⊙
. The effect increases with accretion rate, and high metallicities are also more likely in QSOs showing BAL features if a torus is taken into account.
Conclusions.
The consistency of the model with the data supports the interpretation of the BAL phenomenon as the result of the orientation of the source. It also supports the underlying theoretical model, although more consistency checks should be made in the future.
Context.
We present an analysis of UV spectra of quasars at intermediate redshifts believed to belong to the extreme Population A (xA), aimed to estimate the chemical abundances of the broad line ...emitting gas. We follow the approach described in a previous work, extending the sample to 42 sources.
Aim.
Our aim is to test the robustness of the analysis carried out previously, as well as to confirm the two most intriguing results of this investigation: evidence of very high solar metallicities and a deviation of the relative abundance of elements with respect to solar values.
Methods.
The basis of our analysis are multicomponent fits in three regions of the spectra centered at 1900, 1550, and 1400 Å in order to deblend the broad components of Al
III
λ
1860, C
III
λ
1909, C
IV
λ
1549, He
II
λ
1640, and Si
IV
λ
1397+O
IV
λ
1402 and their blue excess.
Results.
By comparing the observed flux ratios of these components with the same ratios predicted by the photoionization code
CLOUDY
, we found that the virialized gas (broad components) presents a metallicity (
Z
) higher than 10
Z
⊙
. For nonvirialized clouds, we derived a lower limit to the metallicity around ∼5
Z
⊙
under the assumption of chemical composition proportional to the solar one, confirming the previous results. We especially relied on the ratios between metal lines and He
II
λ
1640. This allowed us to confirm systematic differences in the solar-scaled metallicity derived from the lines of Aluminum and Silicon, and of Carbon, with the first being a factor ≈2 higher.
Conclusions.
For luminous quasars accreting at high rates, high
Z
values are likely, but
Z
-scaled values are affected by the possible pollution due to highly-enriched gas associated with the circumnuclear star formation. The high-
Z
values suggest a complex process involving nuclear and circumnuclear star formation and an interaction between nuclear compact objects and an accretion disk, possibly with the formation of accretion-modified stars.
Context. This is the second work dedicated to the observed parallelism between galaxy clusters (GCs) and early-type galaxies (ETGs). The focus is on the distribution of these systems in the scaling ...relations (SRs) observed when effective radii, effective surface brightness, total luminosities, and velocity dispersions are mutually correlated. Aims. Using the data of the Illustris simulation we speculate on the origin of the observed SRs. Methods. We compare the observational SRs extracted from the database of the WIde-field Nearby Galaxy-cluster Survey with the relevant parameters coming from the Illustris simulations. Then we use the simulated data at different redshift to infer the evolution of the SRs. Results. The comparison demonstrate that GCs at z ∼ 0 follow the same log(L)−log(σ) relation of ETGs and that both in the log(⟨I⟩e)−log(Re) and log(Re)−log(M*) planes the distribution of GCs is along the sequence defined by the brightest and massive early-type galaxies (BCGs). The Illustris simulation reproduces the tails of the massive galaxies visible both in the log(⟨I⟩e)−log(Re) and log(Re)−log(M*) planes, but fails to give the correct estimate of the effective radii of the dwarf galaxies that appear too large and those of GCs that are too small. The evolution of the SRs up to z = 4 permits to reveal the complex evolutionary paths of galaxies in the SRs and indicate that the line marking the zone of exclusion, visible in the log(⟨I⟩e)−log(Re) and the log(Re)−log(M*) planes, is the trend followed by virialized and passively evolving systems. Conclusions. We speculate that the observed SRs originate from the intersection of the virial theorem and a relation L = L0′σβ where the luminosities depend on the star formation history.