Resolving the Multiphase Gas Structure and Dynamics of a Nearby Seyfert Galaxy
We are spatially resolving the central kiloparsec of nearby luminous AGN in order to directly probe the impact of feedback on the host galaxy at circumnuclear scales. We undertook an in-depth study of a nearby Seyfert galaxy, NGC 5728, using the combination of three IFU datasets from VLT/SINFONI, VLT/MUSE, and ALMA to probe the circumnuclear multiphase ISM. Through emission line fitting and dynamical modelling, we find a complex interplay between a gas-rich, star-forming circumnuclear ring, inner warped molecular disk, AGN photoionized NLR, and AGN driven outflow. With our rich dataset, we were able to spatially resolve the mass outflow rate, finding a relatively weak outflow extending to only 250 pc. NGC 5728 is an indication that moderate luminosity AGN (LAGN = 1044 erg/s) lack the strength to drive out large quantities of gas and deplete the ISM.
See: Shimizu et al. 2019 (submitted to MNRAS)
Type 1 AGN with Massive Absorbing Columns
As part of our LLAMA study of local luminous active galaxies, we have identified a population of type 1 AGN with modest optical obscuration but remarkably high X-ray absorbing columns. After removing sources in which a prominent outflow masquerades as a fake Broad Line Region (BLR), we find that about 15% of type 1 AGN, mostly classified as Seyfert 1.9, have massive columns of relatively dust-free gas along the line of sight. They are similar to the few objects with highly variable gas columns, which is understood in terms of single BLR clouds passing through the line of sight. But in this more prevalent population, the phenomenon appears to be static and so has implications on the geometry of the gas and dust distributed in the nuclear region close to the AGN.
Constraints on the properties of, and the extinction to, the Broad Line Region in local Seyferts
The instrument VLT/X-SHOOTER allows us to observe medium-resolution spectra from the UV to the near-IR contemporaneously. Such observations are particularly important to study the properties of the broad-line region of AGNs. For the nine broad-line AGNs of our new sample (LLAMA, see Davies et al. 2015 below), we have used these observations and fitted the broad components of multiple Hydrogen transitions (see figure) and simultaneously derived the physical properties as well as the obscuration of the broad line region. By comparing the line ratios with broad line region models, we find typical densities of 1011 cm-3 and show that the HI emitting region is near or at the sublimation radius, consistent with theoretical predictions. We also confirm that in many objects the line ratios are far away from case B. We further find extinctions to the BLR of AV <= 3 mag for Seyfert 1-1.5 and 4-8 for Seyfert 1.8 and 1.9, consistent with our previous estimates using near-IR IFU data (see Burtscher et al. 2016, below).
This study surpasses other similar studies in several ways. Traditionally, spectra covered either a smaller wavelength range (making it hard to fit both the obscuration and the physical properties of the gas at the same time) or were taken at different epochs and carry a considerable uncertainty due to the fast variability of many AGNs. And since our LLAMA sample is complete and volume-limited, selection effects also do not play a role in interpreting our results.
Thick Disks of Dense Gas around the Nuclei of Nearby Seyfert Galaxies
We have been examining the molecular gas in the central regions of local Seyferts unsing the millimetre HCN (1-0) line which traces dense gas. Combining our previous data (Sani et al. 2012) in our new analysis (Lin et al. 2016) we have 8 objects where the kinematics of the line trace a circumnuclear disk. The typical sizes are 30-100pc, and in all cases but one the velocity dispersion needed to match the line width of 20km/s or more. This implies that the disks are geometrically thick. And also, since the warm molecular hydrogen has a much higher dispersion (Hicks et al. 2009) that the gas may be stratified.
The implied column densities are high, exceeding 1022 cm-2 even for low gas fractions. Unexpectedly we have independent confirmation of the colum density from one Seyfert where we observed HCN absorption - with the extreme depth of the absortion impying a very high column density, in the range 1022-23 cm-2 (Lin et al. 2016) both for the rotating disk and in the outflow. One of our goals in the coming years is to link the tracers of the various phases of outflows, as well as piece together the role of outflows across the full range of AGN luminosities and host types.
Far left: absorption profile of H13CN in NGC3079 (black). The two components arise from absorption in the approaching side of the nuclear disk and from the outflow at blue-shifted velocities up to 400 km/s. Centre left: complex profile of the H12CN line. This comprises the sum of the intrinsic symmetric emission profile modified by self-absorption through the edge-on galaxy disk (centre right, dashed and solid blue) and the highly saturated continuum absorption (far right, dashed blue).
The relation between optical obscuration and X-ray absorption
In several studies, differences between the optical obscuration properties and the X-ray absorption properties of AGNs have been claimed. This has even led to the invention of new types of AGNs (e.g. "X-ray absorbed, optically unobscured") and caused quite some confusion regarding the nature and location of the absorbers.
We revisit this topic using a newly collected sample of about 50 local AGNs with very high-resolution VLT/SINFONI observations. After establishing a simple method to estimate the optical extinction towards the hot dust, we compared them with the absorption in X-ray provided by a new and uniform analysis of X-ray absorbing columns derived from fits to the spectra in the very wide range 0.3-150 keV. Additionally, we compiled from the literature ranges for the variability of the X-ray absorbing column. By assembling both sets of measurements, a picture is revealed (see figure) in which the optical obscuration, converted to a gas column using the Galactic gas-to-dust ratio, sets the floor for the X-ray absorption in AGNs. Several AGNs lie close to this value and in those the X-ray absorption is likely dominated by dusty gas that is also probed in our obscuration measure. Several other AGNs, however, show variability in the X-ray absorbing column which is likely connected to dust-free neutral gas within the dust sublimation radius.
Furthermore, a classification system can be found that is consistent between X-rays and optical observations, if the boundaries for absorbed/unabsorbed and obscured/unobscured are properly chosen.
The AGN phenomenon is likely short-lived. Statistical evidence, e.g. from the fraction of X-ray bright, optically normal galaxies (XBONGs), as well as hydrodynamical simulations indicate that a particular AGN phase may only last for as little as 105 years. It is therefore not surprising that little differences between the host galaxy properties of active and inactive galaxies are found when assessing e.g. the host galaxy morphology which will only change on dynamical times much larger than the AGN timescale. We have therefore set up a unique study to compare the nuclear inventory of AGNs and inactive galaxy in a comprehensive way. For this we first selected the twenty most luminous local AGNs based on the sensitive ultra hard X-ray all-sky survey provided by the Swift/BAT satellite. The rationale for this selection is to only select AGNs where coherent inflow mechanisms are required due to the large luminosity or mass accretion. We further require the galaxies to be local in order to resolve the nuclear region, in particular the nuclear stellar cluster on scales of 50-100 pc. And finally we employ the ultra hard X-ray band to be largely unimpeded by obscuration which can severly affect AGN selection based on optical properties. The so selected AGN sample is accompanied by a sample of inactive galaxies (where no ultra hard X-ray emission has been detected) that are matched in host galaxy properties to the active sample. This sample of 39 galaxies in total is currently being observed or observations are planned with VLT/X-SHOOTER, VLT/SINFONI, APEX as well as ALMA and HST.
The inactive galaxy IC 5267 appears regular on large scales. In the nucleus, however, a dusty filament appears to connect material from kpc-scales down to the nucleus. The kinematics of the nuclear region, as seen with SINFONI, reveal a gas disk in the plane of the galaxy (matching the large scale disk) and a ring of gas at a radius of a few hundred parsecs at a different orientation, associated with the dust lane. It is very likely that the gas in this ring will collide with gas disk and lead to a major inflow turning this inactive galaxy into an AGN. Since the gas ring is not associated with the plane of the galaxy, the gas has likely been accreted externally, e.g. from the group environment of the galaxy. In fact, S0 galaxies such as IC 5267 do not have a major gas supply of their own and therefore tend to occur in small groups (where the inter-galactic gas is not necessarily ionized such as in clusters, and hence may be able to accrete). We are currently preparing a follow-up study to investigate the environment of AGN host galaxies depending on their morphology.
NGC 1097 exhibits many classical features of disk galaxies: a large scale bar, a circumnuclear ring and starburst, an inner spiral, and a low luminosity AGN. Its one peculiarity is that the inner spiral has 3 photometric arms. Our integral field observations show that this pattern is due to obscuration of stellar light by gas and dust in the molecular arms. Intriguingly, the gas kinematics reveal a strong non-circular velocity residual in the form of a 2-arm spiral -- as expected from linear theory which indicates that the projected line-of-sight velocity pattern of an m-arm spiral is an (m-1)-arm spiral. The properties of the arms allow us to derive an flow rate along them into the central 20pc of about 1.2Msun/yr. But hydrodynamical simulations of nuclear spirals show that there is also significant outflow between the arms in the low density regions (gas from the arms overshoots the nucleus and continues as a diverging outflow). Applying this correction to NGC 1097 means the net inflow could be as little as 0.06Msun/yr.
Although there is no doubt that a torus of molecular gas exists, and that it obscures the AGN from some viewing angles, what it actually looks like is very uncertain. Our SINFONI and OSIRIS data of molecular gas may have revealed the large scale structures associated with the torus. We have been able to show that it fits the necessary criteria: the size of tens of parsecs means it is compact; the huge column density means it is optically thick; the high dispersion, which traces bulk motions, means it is vertically extended. Intriguingly, the distribution and kinematics of this molecular gas is very similar to that of the nuclear stellar population. This can only mean the gas and star are mixed, and that the torus is a star forming torus. The implication is that the torus is a dynamical entity, the properties of which are determined by a combination of gas inflow and the effects of star formation.
Although a prototypical Seyfert 2, NGC 1068 is in fact rather unusual. The H2emission in the central 250pc originates in an expanding off-centre shell with particularly bright and massive clumps around the north east side. Filaments of gas extend from the ring at a radius of about 30pc to the AGN on both sides. Modelling the morphology and kinematics of the filaments has shown that the only way to simultaneously account for both constraints is if they trace gas that is falling almost directly in towards the AGN. These models indicate that the infall timescale is about 1.3Myr. One of the filaments lies across the front of the AGN, suggesting that inelastic collisions may allow the gas to settle on scales of a few parsecs.
There appears to be a relation between the characteristic age of the most recent episode of star formation and the luminosity of, or accretion rate onto, the AGN. This figure suggests that there is a delay of 50-100Myr between the short duration starburst and the onset of accretion onto the AGN. This timescale is interesting because it corresponds to the age at which, for a short burst, one expects the type IIsupernovae phase to be finishing. On the other hand, slower outflows, perhaps from AGB stars, remain bound and so in principle can be accreted inwards to smaller scales.
Using the spectral information to separate out the non-stellar continuum (hot dust emission) associated with the AGN, has shown that there are starbursts around nearby AGN with typical size scales <50pc. These are distinct stellar populations: photometrically, from an excess in the stellar continuum; and kinematically, from a drop in the stellar dispersion.
Stellar population synthesis models indicate that the starburst ages are young, in the range 10-300Myr. And the remarkably low equivalent width of the Brγ line shows unambiguously that the star formation has ceased: i.e. although still young, these are post-starbursts. Intriguingly, starburst models imply that during the short star forming phase, the luminosity of the nuclear starburst would have been an order of magnitude greater than it is now -- with a star formation rate per unit area comparable to that in ULIRGs.
Stellar Kinematical Black Hole Mass in a Seyfert 1
One of the most reliable ways to measure the mass of the black hole at the centre of a galaxy is to use spatially resolved stellar kinematics. We have done this for the first time in a Seyfert 1 galaxy, NGC 3227. Using adaptive optics to reach the resolution necessary to resolve the sphere of influence of the black hole, we applied Schwarzschild superposition modelling to the full 2D kinematics measured by SINFONI. The black hole mass lies in the range (0.7-2.0)x107Msun. This is a robust result, although the range is rather large because of the limited signal-to-noise and also from degeneracies between the mass of the various components in the model.