A statistical investigation of clumpy torus model predictions in the mid-infrared

<p>The dusty torus is the key ingredient of the unification scheme of AGN. It is known that clumpy dust configurations are more realistic than homogeneous ones and reproduce better the observations. Nevertheless, it is important to improve the physical hypotheses of current clumpy torus. To do so, we compare Hönig & Kishimoto (2010) clumpy torus models with a more physically motivated version of the same models. We investigate the mid-infrared spectral index and the strength of the silicate feature as they provide information about the radial distribution of the clouds and the number of clouds in the torus, respectively. We also compiled mid-infrared spectra for 55 Seyfert galaxies obtained with instruments on 8-10 m telescopes (CanariCam, VISIR, T-ReCS, Michelle). Before we compare the data with the clumpy torus model predictions, we use the spectral decomposition tool deblendIRS to separate the star-formation component from the emission of the dusty torus. The main goal of this work is to constrain some of the parameters of the Hönig & Kishimoto models from the statistical point of view using the mid-IR AGN spectral indices and silicate feature strengths measured for our sample of Seyfert galaxies.<br>