This paper is the second part of a work investigating the properties of dusty tori in active galactic nuclei ( AGN ) by means of multi-component spectral energy distribution ( SED ) fitting .
It focuses on low luminosity , low redshift ( z \leq 0.25 ) AGN selected among emission line galaxies in the overlapping regions between SWIRE and SDSS Data Release 4 as well as X-ray , radio and mid-infrared selected type 2 AGN samples from the literature .
The available multi-band photometry covers the spectral range from the u -band up to 160µm .
Using a standard \chi ^ { 2 } minimisation , the observed SED of each object is fit to a set of multi-component models comprising a stellar component , a high optical depth ( \tau _ { 9.7 } \geq 1.0 ) torus and cold emission from a starburst ( SB ) .
The torus components assigned to the majority of the objects were those of the highest optical depth of our grid of models ( \tau _ { 9.7 } = 10.0 ) .
The contribution of the various components ( stars , torus , SB ) is reflected in the position of the objects on the IRAC colour diagram , with star- , torus- and starburst-dominated objects occupying specific areas of the diagrams and composite objects lying in between .
The comparison of type 1 ( as derived from Paper 1 , Hatziminaoglou et al . 22 ) and type 2 AGN properties is broadly consistent with the Unified Scheme .
The estimated ratio between type 2 and type 1 objects is about 2-2.5:1 .
The AGN accretion-to-infrared luminosity ratio is an indicator of the obscuration of the AGN since it scales down with the covering factor .
We find evidence supporting the receding torus paradigm , with the estimated fraction of obscured AGN , derived from the distribution of the covering factor , decreasing with increasing optical luminosity ( \lambda L _ { 5100 } ) over four orders of magnitude .
The average star formation rates are of \sim 10 M _ { \odot } / yr for the low- z sample , \sim 40 M _ { \odot } / yr for the other type 2 AGN and \sim 115 M _ { \odot } / yr for the quasars ; this result however , might simply reflect observational biases , as the quasars under study were one to two orders of magnitude more luminous than the various type 2 AGN .
For the large majority of objects with 70 and/or 160 µm detections an SB component was needed in order to reproduce the data points , implying that the far-infrared emission in AGN arises mostly from star formation ; moreover , the starburst-to-AGN luminosity ratio shows a slight trend with increasing luminosity .