Simulating AGN

Since there was so much interest our quest to understand how the presence of an AGN influences the classification of the host galaxy morphologies, I’m going to expand on that topic here.

AGN in nature come with a large variety of luminosities. Some are so bright that we cannot even tell they are inside of a large galaxy (so-called Quasars) and some are so faint that we have no way of detecting their presence in our data. The AGN that I have been studying are somewhere in between these two extreme cases. So we were wondering how much a bright-ish point source might alter our ability to determine the host galaxies classifications. (Note: this is not an issue of the ability of Zooites to classify galaxies, but a question of when the brightness of the central AGN begins to hide or distort the visible features of the galaxy.)

Here are a few of the actual AGN from the HST images: I would not be able to identify these as AGN just from the images and for some of them – not even the spectra reveal the presence of the AGN.

We teamed up with Brooke Simmons, a PhD student at Yale and an expert image analyzer to create these images. We selected 116 normal galaxies (with no AGN) and then added a central point source of varying brightness to each. We selected the brightness of point sources to cover the range of the types of AGN that are found in the deep Hubble surveys. All together using the 116 normal galaxies with the sampled range of AGN luminosities, we ended up with 1740 total simulated AGN galaxies.

Stay tuned: we’ll be posting the results from these classifications shortly.