Galactic nuclei: dust structure & evolution, accretion & jet structure, spectral energy distributions, gas kinematics & black hole masses
Infrared observations: high-spatial resolution, mid-infrared, ground-based methods, dust & PAH emission
The subarcsecond MIR atlas of local AGN is a collection of all available N- and Q-band images obtained at ground-based 8-meter class telescopes with public archives (Gemini/Michelle, Gemini/T-ReCS, Subaru/COMICS, and VLT/VISIR). It includes in total 895 images, corresponding to 253 local AGN with a median redshift of 0.016 (Asmus et al. 2014) and is accessible in the Virtual Observatory. The atlas contains the uniformly processed and calibrated images and nuclear photometry obtained through Gauss and PSF fitting for all objects and filters. This also includes measurements of the nuclear extensions. In comparison to the arcsecond-scale MIR emission as probed by Spitzer, the continuum emission is much lower on subarcsecond scales in many cases. The silicate feature strength is similar on both scales and generally appears in emission (absorption) in type I (II) AGN. However, the polycyclic aromatic hydrocarbon emission appears weaker or absent on subarcsecond scales. The differences of the MIR SEDs on both scales are particularly large for AGN/starburst composites and close-by (and weak) AGN. The nucleus dominates over the total emission of the galaxy only at luminosities ≳1044 erg s-1. The AGN MIR atlas is well suited not only for detailed investigation of individual sources but also for statistical studies of AGN unification.
The mid-infrared -- X-ray correlation for AGN:
We used SASMIRALA to re-investigate the well-known correlation between the AGN emission at MIR and X-ray wavelengths (Asmus et al. 2015). This correlation is surprising for many reasons: 1. It extends over the full range of probed luminosities with a slope of unity not showing any change of slope as expected for example by the receding torus scenario.
2. It is very tight with a scatter of two, applicable to all AGN types (except maybe narrow-line Seyfert 1s), which implies very small differences in the emission of the different AGN types, in particular unobscured versus obscured AGN. Canonical torus models predict larger differences between these two.
3. Even radio-loud AGN seem to follow the correlation rather well, despite the expected additional contribution of jet synchrotron emission to the MIR. Looking very closely, it seems that the slope od this correlation is different for the radio-loud objects at least but with a normalisation such that the population overlap in the MIR--X-ray plane.
A possible explanation for at least the first two points, is given by the results from the next project:
The mid-infrared emission from polar dust:
Up to know the nuclear MIR emission has been associated directly with the dusty obscurer, and could well be fit with clumpy torus models. However, the structures at parsec sales resolved with MIR interferometry (Hönig et al. 2012, 2013, Tristram et al. 2014, and López-Gonzaga et al. 2014) show a surprising polar elongation which is somewhat at odds with the canonical clumpy torus.
This polar emission is also found on larger (10 to 100pc) scales as it turns out! We could show this statistically for the first time using the extended AGN in SASMIRALA, which have no significant star formation on nuclear scales (Asmus et al. 2016). The results indicate that the MIR emission of AGN is in fact dominated by dust on the edges of the ionisation cones (a dusty wind?) instead of a clumpy torus. This in turn implies that the real obscurer must be more compact and hotter than previously thought...