Speaker
Description
A complete census of SMBH increases our understanding of the role of AGN evolution over cosmic time. As AGN detection is less affected by obscuration effects in the X-ray window, eROSITA offers increased likelihood and purity in detecting these objects. That being said, a substantial fraction of spectroscopic redshifts for AGN identified by eROSITA will be available only in 2-3 years from now at best. In the meantime, we must rely on photometric redshifts (photo-z), where for wide-area surveys, the quality of current estimates for AGN using broad-band photometry is poor. The limited number of photometric bands is insufficient to disentangle complex convolved AGN/host-galaxy contributions, resulting in a high fraction of outliers, as relevant parameters established by the source detection and flux estimate algorithms, are usually fine-tuned for galaxies and not AGN.
More recent efforts to compute photo-z for AGN utilizing the radial light distribution with aperture photometry (provided by the Legacy Survey) via ML, have shown promising improvements, as its shape changes with redshift given a fixed resolution. For this reason, we further extend our novel single-survey Deep Learning algorithm by raising the spatial light distribution resolution through images, alleviating previous empirical approaches by decreasing the fraction of outliers. In my talk, I will show how our works "CircleZ" and "PICZL" outperform previous results while uncovering various sources of contaminants.
