Clumpy AGN Tori in a 3D geometry

This is the official website of the CAT3D clumpy torus models as presented in Hönig & Kishimoto (2010).
Contact: Any questions, suggestions, comments are welcome:

1. Summary and purpose

CAT3D aims at providing model SEDs and images for clumpy dust emission in a torus around the AGN accretion disk by combining Monte Carlo radiative transfer simulations and ray-tracing techniques. This allows for recovering the dust cloud distribution in a 3D geometry accounting for all the statistical effects that can happen when clouds are randomly distributed according to a pre-specified set of model parameters. The model has been applied to a number of case and sample studies (see (4) CAT3D in action and (6) Work using CAT3D). On this website we want to make the model results available to the community.

2. The method

In a first step, the phaseangle-dependent emission for each cloud is simulated by Monte Carlo radiative transfer simulations. For that, we apply the non-iterative method delineated by Bjorkman & Wood (2001), based on work by Lucy (1999). In our previous model, we used the code mcSim (Ohnaka et al. 2006) which is able to simulate a variety of geometries. However, in order to obtain fast results for different dust compositions, we created a new Monte Carlo code which is optimized for the AGN-cloud-configuration. For each given dust composition, we determine the sublimation radius rsub=r(Tsub=1500K) from the source, and simulate several clouds at different distances (normalized for rsub). To finally simulate the torus emission, dust clouds are randomly distributed around an AGN according to some physical and geometrical parameters. Each cloud is associated with a pre-simulated model cloud, after accounting for the individual cloud’s direct and indirect heating balance. The final torus image and SED is calculated via raytracing along the line-of-sight from each cloud to the observer. This method accounts for the actual 3-dimensional distribution of clouds in the torus involving all statistical variations of randomly distributed clouds in a time-efficient way.

3. Model downloads

Simulating model grids is a work-in-progress. At the moment, we are providing some standard model SEDs which cover a range of model parameters as presented in Hönig & Kishimoto (2010).

3.1 Model parameters
Before using the model SEDs, it is important to understand the model parameters, so please take the time and read Sect. 2.2 in the paper. There are 7 parameters in total which characterize the distribution of dust clouds and the properties of the individual clouds. However only 5 of them have a direct influence on the SEDs. These are:
  • the index a of the radial dust cloud distribution power law
  • the half-opening angle θ0
  • the number of clouds N0 along an equatorial line-of-sight
  • the outer radius of the torus Rout (but see Sect. 4.1.4 in the paper)
  • the optical depth τV of the individual clouds

3.2 Currently available models
Varieties of model SED grids based on these parameters are available and more will be added:
  • CAT3D.model_standard.tar.gz: model grid covering a=0.0...-2.0 (in steps of 0.5); θ0 = 5°, 30°, 45°, 60°; N0 =2.5...10 (in steps of 2.5); τV =30, 50, 80; Rout = 150 based on ISM dust distribution (sizes 0.025 μm to 0.25 μm) using 47% graphite (Draine 2003) and 53% silicates (Ossenkopf et al. 1994); dust sublimation radius rsub;0 = 0.9pc at 1046 erg/s
(Note section (5) How to cite CAT3D for terms of use of these models.)
In each model grid .tar.gz file you will find a number of ascii files. The ascii file names are referring to the model parameters used. Each file consists of 12 header lines (indicated by leading ‘#’) followed by 8 columns of data: col. 1 frequency (Hz); col. 2 wavelength (μm); col. 3-9 model νFν (W/m2) for inclination 0, 15, 30, 45, 60, 75, 90°.

3.3 Model scaling
The models are scaled in a way that they can be used for all kind of AGN. Nature does us a favor by setting an intrinsic scale for the torus: the dust sublimation radius which is depending on the luminosity. At this radius, the luminosity per unit area is always the same, no matter how powerful the AGN is. If we scale all of our sizes involved in the models for the dust sublimation radius, we are independent of the actual luminosity. Therefore the model fluxes are representing the emission that an observer at the distance of the sublimation radius would see. The conversion between model fluxes and observed fluxes is simply
where rsub is the sublimation radius of the AGN you observe and DL is the luminosity distance to the AGN. If you do not want to deal with the sublimation radius, it is also possible to use the AGN (optical/UV) luminosity instead because
rsub;0 is depending on the dust which is used in the models and provided in the model description.

Examples:
  1. We have a type 1 AGN at a distance of 45 Mpc with a sublimation radius (e.g. based on interferometry of IR reverberation mapping) is 0.06 pc. Our favorite model has a 12 μm model flux of 1.0 x 105 W/m2. Thus, the prediction for the observed flux would be 1.78 x 10-13 W/m2 or 0.71 Jy.
  2. Now, we have a type 2 AGN at a distance of 14.5 Mpc. It has a 12 μm model flux of 2.0 x 104 W/m2. In type 2s there’s no reverberation mapping which can measure rsub. However we can infer a UV luminosity by other means, let’s say log L = 44.5. Using the standard model (rsub;0 = 1.1 pc), we will obtain 3.64 x 10-12 W/m2 or 14.6 Jy.

4. CAT3D in action

False color images of dust tori at 3 different wavelengths.
A relation between the mid-IR spectral index and the radial dust distribution power law index in type 1 AGN. This opens an easy way to get some physical parameter from your observations.
The nominal relation between power law index and mid-IR slope is
Modeling mid-IR interferometry of NGC3783:

5. How to cite CAT3D

Whenever you use the models provided here, the correct way of citing CAT3D is

H\”onig, S. F., & Kishimoto, M. 2010, A&A, 523, 27

and if you want you can refer in addition to the original paper introducing the concept:

H\”onig, S. F., Beckert, T., Ohnaka, K., & Weigelt, G. 2006, A&A, 452, 459.