Upgrade to Pro — share decks privately, control downloads, hide ads and more …

The central parsec of NGC 3783: a rotating broad emission line region, asymmetric hot dust structure, and compact coronal line region / 2021-09-01

Akio Taniguchi
September 01, 2021

The central parsec of NGC 3783: a rotating broad emission line region, asymmetric hot dust structure, and compact coronal line region / 2021-09-01

Akio Taniguchi

September 01, 2021
Tweet

More Decks by Akio Taniguchi

Other Decks in Science

Transcript

  1. The central parsec of NGC 3783: a rotating broad emission

    line region, asymmetric hot dust structure, and compact coronal line region GRAVITY collaboration 2021, A&A, 648, A117 Akio Taniguchi / 2021-09-01 / A-lab Journal Club
  2. Overview • Purpose: building a comprehensive picture of the 0.01-100

    pc region of an AGN • Observations: Brγ (→BLR), [Ca VIII] (→CLR), and 2μm continuum (→hot dust) at the center of NGC 3783 (type 1 AGN) by VLT(I) • Discoveries: • BLR: r~0.013 pc, rotating thick disk • Hot dust: r~0.14 pc central source + offset cloud • CLR: r~0.4 pc, between BLR and NLR CLR NLR BLR Hot Dust M olecular Gas Disk The proposed view of NGC 3783 built from the paper results
  3. Introduction: The Uni fi ed Model of AGN • Active

    galactic nucleus (AGN) • One of the most energetic source in a galaxy that hosts a supermassive black hole (SMBH) • Radiation from an accretion disk • Mass transfer by a jet and/or an out fl ow • The Uni fi ed Model of AGN • AGN types (presence of BLR, etc) can be explained by a thick gas and dust torus surrounding a SMBH (Antonucci 1993) • Compactness makes it challenging to con fi rm REVIEW ARTICLE NATURE ASTRONOM understanding of AGN obscuration, showing that ~70% of all local AGNs are obscured21,22. While nuclear obscuration is mostly associ- ated with dust in the torus at IR wavelengths, it can also be related to dust-free gas in the case of X-rays. Indeed, it is likely that X-ray scenario40–42. From both IR and X-ray observations it has bee shown that the nuclear dust is distributed in clumps25,43, and furth constraints on the torus size and geometry have been provided b MIR interferometry13,42. The MIR-emitting dust is compact an Corona BLR Torus Ionization cone NLR Polar dust Outflow log( ) z pc –5 –5 –4 –3 –2 –1 0 1 2 3 –4 –3 –2 –1 0 1 2 log( ) r pc SMBH Disk Fig. 1 | Sketch of the main AGN structures seen along the equatorial and polar directions. From the centre to host-galaxy scales: SMBH, accretion disk and corona, BLR, torus and NLR. Different colours indicate different compositions or densities. Sketch of the main AGN structures seen along the equatorial and polar direction (Almeida and Ricci 2017) ?
  4. Introduction: NGC 3783 (type 1 AGN) • NGC 3783: a

    nearby Seyfert galaxy that hosts one of the most luminous type 1 AGNs • d ~ 38.5 Mpc (1 arcsec = 187 pc) • LAGN ~ 1044.5 erg/s • MSMBH ~ 3 x 107 Msun • Features • Notable ionized out fl ows (101-2 pc scale) • Size of BLR is estimated (~10 light days) by UV-to-optical reverberation mapping (RM) • Issues: innermost geometry is not con fi rmed yet © CGS
  5. This paper: observations of the innermost region by VLT(I) •

    Purpose: building a comprehensive picture of the 0.01-100 pc region of NGC 3873 • Observations: • VLTI/GRAVITY: interferometer • VLT/SINFONI: imaging spectrograph • Observables: • Brγ (n=7→4; 2.17 μm): BLR tracer • [Ca VIII] (2.32 μm): CLR tracer • K-band continuum: hot dust tracer © ESO/S. Brunier SINFONI GRAVITY
  6. Observations: VLT/SINFONI (imaging spectrograph) • SINFONI • NIR imaging spectrograph

    with AO • Mounted at the cassegrain focus of UT3 • High (R ~ 4000) spectral resolution • K-band observations • One night on April 20, 2019 • Total on-source time of 10 minutes (?) • PSF (effective resolution): 2D Gaussian with 75 mas (~14 pc) FWHM • K-band continuum image (right) showing on/o ff nuclear regions • Integrated spectra extracted from the SINFONI cube (top) nuclear o ff -nuclear 40 pc
  7. Observations: VLTI/GRAVITY (interferometer) • Very Large Telescope Interferometer (VLTI) •

    Four 8.2 m Unit Telescopes (UT1 - UT4) • GRAVITY • Four-beam interferometric combiner • Medium (R~500) spectral resolution • K-band observations • Six nights in 2018 - 2020 • Total on-source time of 8.1 hours • Three binned visibilities for each baseline uv coverage of the observations (60 Mλ ~ 3.4 mas)
  8. Results: overview of BLR, hot dust, CLR analyses • GRAVITY

    data: visibility amplitude and phase normalized to the continuum level • SINFONI data: pro fi les of each emission line • Analyses: • BLR: physical model fi tting to derive the size, the thickness, etc of the Brγ emission at μas scale • Hot dust: imaging to derive the distribution and the size of continuum at mas scale (i.e. superresolution) • CLR: 2D Gaussian model fi tting to derive the size of the [Ca VIII] emission at mas scale CLR NLR BLR Hot Dust M olecular Gas Disk The proposed view of NGC 3783 built from the paper results
  9. Results (BLR): di ff erential amplitude and phase A bump

    (amp) and an S-shape (phase) seen in the longest baseline indicate the compact BLR (shorter baseline) (longer baseline) Di ff . visibility amplitude Di ff . visibility phase in deg 1.00 1.01 1.02 0.0 +0.5 -0.5 2.15 2.20 (μm) 2.15 2.20 2.15 2.20 Colored curves: observed spectra Black curves: physical model Gray curves: line pro fi le of Brγ spectra: GRAVITY (amp) SINFONI (phase)
  10. Results (BLR): physical model fi tting to the data A

    physical model (a large number of clouds trapped by a SMBH) well explains the observed visibilities (both amp and phase) → a rotating thick disk (r ~ 0.013 pc, angular thickness ~ 24 deg) (a) (b) (c) Cloud distribution and velocity Amp and phase (averaged) Amp and phase (each baseline)
  11. Results (BLR): physical model fi tting to the data A

    physical model (a large number of clouds trapped by a SMBH) well explains the observed visibilities (both amp and phase) → a rotating thick disk (r ~ 0.013 pc, angular thickness ~ 24 deg) Single point-like source model fi tting to each frequency ch (so called "photocenter") Two point-like sources model fi tting (black point is the null model to be rejected) Physical model described in the previous slide SMBH SMBH SMBH Δφ ∝ ux + vy
  12. Results (hot dust): superresolution imaging • Continuum image reconstruction: •

    MiRA algorithm (Thiébaut 2008) • Least square + regularization optimization • Regularization: • Positivity prior ( fl ux must be >0) • Hyperbolic regularization • fl ux should be smooth inside • but sharp edges are allowed • Bright central source + faint offset source: • FWHM of 0.14 pc for the central source 10 5 0 5 10 R.A. (m a s) 10 5 0 5 10 Dec. (m a s) 2 pc 10 2 10 1 100 N orm a lized Flux Ionization Cone BLR Polar Axis Polarim etry MIR O ff set cloud (0.6 pc away from center)
  13. Results (hot dust): superresolution imaging The robustness of the detection

    of (sub)structures was checked in multiple images reconstructed by a different prior or a different algorithm MiRA (positivity + smoothness prior) MiRA (compactness prior) SQUEEZE (Baron+ 2010)
  14. Results (CLR): 2D Gaussian fi tting to the data •

    Coronal line region (CLR): • Highly ionized atoms (like solar corona) • high ionization potentials (> 100 eV) • high critical densities (107-10 cm-3) • CLR may be located between BLR and NLR • 2D Gaussian fi tting to [Ca VIII] spectra: • constraint of the CLR size for the fi rst time • FWHM ~ 0.4 pc > BLR and hot dust sizes • (cannot rule out the origin of inner regions) Colored curves: observed di ff . spectra (each baseline) Black curves: Spectra extracted from the 2D Gaussian Gray curves: line pro fi le of [Ca VIII] of GRAVITY
  15. Discussion: connecting the circum/nuclear regions [Si VI] (another coronal line)

    maps show non-circular features at the north, suggesting an out fl ow BLR axis BLR axis BLR axis 10 5 0 5 10 R.A. (m a s) 10 5 0 5 10 Dec. (m a s) 2 pc 10 2 10 1 100 N orm a lized Flux Ionization Cone BLR Polar Axis Polarim etry MIR
  16. Discussion: connecting the circum/nuclear regions Ro-vibrational H2 (hot molecular gas)

    map suggest thick disk with low inclination BLR axis BLR axis 10 5 0 5 10 R.A. (m a s) 10 5 0 5 10 Dec. (m a s) 2 pc 10 2 10 1 100 N orm a lized Flux Ionization Cone BLR Polar Axis Polarim etry MIR BLR axis
  17. Discussion: connecting the circum/nuclear regions • Differences of axes: •

    Molecular gas: PA ~ -20 deg • BLR: PA ~ -65 deg • CLR: PA ~ -18 - +10 deg • Warping of the gas disc • to connect large-scale kinematic axis • Curve of the out fl ow from the AGN • to connect large-scale out fl ow to north CLR NLR BLR Hot Dust M olecular Gas Disk The proposed view of NGC 3783 built from the paper results Large-scale gas disk ??
  18. Summary • Discoveries: • BLR: r~0.013 pc, rotating thick disk

    • Hot dust: r~0.14 pc central source + offset cloud • CLR: r~0.4 pc, between BLR and NLR • Opinions: • Good: how they derive many information (e.g. spatial distribution) from very limited visibility datasets • To be improved: Various model fi ttings have many (>10) parameters and χ2 are sometimes less than unity, implying over fi t? CLR NLR BLR Hot Dust M olecular Gas Disk The proposed view of NGC 3783 built from the paper results