Key results from our GRAVITY-AGN programme. Left: Resolved kinematic structure of the BLR in 3C 273 revealing a dominant rotation component Right: Spatially resolved image of the hot dust in NGC 1068 revealing a thin ring

The AGN broad emission line region and obscuring structure

Two critical components of active galactic nuclei (AGN), the broad line region (BLR) and the surrounding dusty obscuring structure are thought to be inherently linked together through the unified model of AGN whereby a large obscuring structure blocks the BLR for some viewing angles and splits AGN into two subsamples: Type 1 (unobstructed view of the BLR) and Type 2 (obstructed view). Thus, it is of great importance to understand in detail both structures, yet the physical origin of both and their connection to the accretion disk and larger host galaxy remains unclear. Of equal importance, the strong emission lines from the BLR due to the high velocity motion of clouds near the black hole can be used to measure the black hole mass if the motions are due to gravity.

The central engines of the brightest AGN, however, only span milli-arcseconds on the sky, but are accessible with the Very Large Telescope Interferometer (VLTI) when using the UT telescopes. Early results in the mid-infrared found that the warm dust emission (few 100 K) has an extended polar component that might be part of an outflow. Keck Interferometer and VLTI/AMBER observations in the near-infrared have found a compact size for the hot dust emission (1000 K) region, compatible with the location where dust is no longer destroyed by the powerful ionizing AGN continuum (the “sublimation radius”). Our GRAVITY observations can reach 10-100x higher sensitivity due to fringe tracking and greatly improve the uv-coverage by combining all 4 UTs.

Broad line region structure and black hole mass estimates

Studies of BLR structure have necessarily relied mostly on reverberation mapping (RM). RM techniques use the time lag between the variable AGN continuum and the emission lines to measure the BLR radius. These RM programs established an R-L relation (RBLR = L1/2) which allows for black hole mass estimates from a single AGN spectrum. This is the only available method for measuring black hole mass in large surveys and out to high redshift and plays a key role in our understanding of black hole growth over cosmic time. However, the R-L relation is not well determined due to large scatter in the data, questioning the reliability of earlier mass determinations. Moreover, velocity-resolved RM studies indicate a variety of BLR geometries and a previously unknown dependence on the Eddington ratio. GRAVITY spectro-astrometry provides a new, direct probe of the BLR spatial and velocity structure which can test reverberation methods and measure an independent R-L relation. The brightest AGN on sky are expected to have comparable angular sizes, making interferometry particularly well suited to studying BLR properties over a wide range of AGN luminosity. Establishing a new, GRAVITY-based BLR sample will provide a more accurate R-L relation and improved AGN black hole mass determinations.

Key Publications

The spatially resolved broad line region of IRAS 09149-6206 (Gravity Collaboration submitted to A&A)

Spatially resolved rotation of the broad-line region of a quasar at sub-parsec scale (Gravity Collaboration 2018)

The near-infrared size of the hot dust

A long-standing issue of AGN unification models is the size and structure of the obscuring, dust-emitting region: is it a torus or disk, inflowing or outflowing? The near and mid infrared luminosity originates in dust surrounding the AGN and heated by it. However, like the BLR, circumnuclear dust in AGNs is unresolved in single-dish images. In the past decade, infrared interferometry has begun to shed light on the physical structure of this component. Tens of AGN have been observed in the mid-IR with MIDI. Detailed results from Circinus and NGC 1068 show evidence for an inner disk, but have also revealed dust in the polar regions possibly indicative of outflow. The presence of multiple components could have a severe impact on models of AGN unification. The NIR is thought to trace hot dust just beyond the sublimation radius. Previous K band interferometric observations with Keck and VLTI AMBER have detected partially resolved structures with sizes comparable to the sublimation, but are limited in uv coverage and sensitivity. GRAVITY observations provide the first resolved view of the shape and structure of the hot dust emission region whose orientation can be compared directly with that of the BLR.

Key Publications

An image of the dust sublimation region in the nucleus of NGC 1068 (Gravity Collaboration 2020a)

The resolved size and structure of hot dust in the immediate vicinity of AGN (Gravity Collaboration 2020b)



We have begun a large program to use deep VLTI/GRAVITY spectro-astrometry data of a sample of the brightest AGN (K < 10 mag) in the southern hemisphere to measure the size of their BLRs and establish a new radius-luminosity (RL) relation based on spatially resolved data at previously inaccessible micro-arcsecond scales. Our aim is to simultaneously measure the BLR kinematics and the BLR and hot dust size and structure. Our observations will address these key science questions:

  • Does the BLR radius scale as RBLR = La, a finding from reverberation studies on which practically all current black hole mass estimates are based? What is the value of a and how large is the intrinsic scatter in the relation?
  • Are BLR kinematics always dominated by rotation and well described by gravitationally bound gas clouds? Or is there evidence for outflow (disk wind) or inflow?
  • Is the hot dust emission region concentrated in a torus with an inner edge at the sublimation radius? Are the torus and BLR part of one continuous structure feeding or outflowing from the central accretion disk?

While the previous questions and topics are achievable with the current capabilities of GRAVITY, with its upgrade to GRAVITY+, we will be able to push our BLR and dust structure studies to higher redshift (z~2). Further, we will explore exciting new areas related to AGN such as supermassive black hole binaries, tidal disruption events, and super-Eddington accretion.

Team and Collaborations

Yann Clenet (LESIA, France), 
Ric Davies (MPE),
 Jason Dexter (University of Colorado, Boulder, US), 
Andreas Eckart (MPIfR),
 Frank Eisenhauer (MPE), Reinhard Genzel (MPE), Damien Gratadour (LESIA, France), Sebastian Hönig (University of Southhampton, UK),
 Makoto Kishimoto (Kyoto Sangyo University, Japan), Sylvestre Lacour (ESO), Dieter Lutz (MPE)
F. Millour (Université Côte d’Azur, France), Hagai Netzer (Tel Aviv University, Israel), Guy Perrin (LESIA, France), Brad Peterson (Ohio State University, US), P.O. Petrucci (Univ. Grenoble Alpes), Oliver Pfuhl (ESO), Almudena Prieto (Instituto de Astrofísica de Canarias, Spain), Daniel Rouan (LESIA, France), Jinyi Shangguan (MPE), Taro Shimizu (MPE), Eckhard Sturm (MPE), Julien Woillez (ESO)



1. Spatially resolved rotation of the broad-line region of a quasar at sub-parsec scale

     GRAVITY Collaboration et al. 2018, Nature, 563, 7733

2. The resolved size and structure of hot dust in the immediate vicinity of AGN

     GRAVITY Collaboration et al. 2020, A&A, 635, 92

3. An image of the dust sublimation region in the nucleus of NGC 1068

     GRAVITY Collaboration et al. 2020, A&A, 634, 1

4. The spatially resolved broad line region of IRAS 09149-6206

     GRAVITY Collaboration et al. 2020, submitted to A&A

Go to Editor View