Affiliation: Instituto de Astrofísica e Ciências do Espaço – U. de Lisboa

Contribution: Oral

Title: Radio luminosity functions from ML analysis

Abstract: Determining the distribution of Active Galactic Nuclei (AGN) across cosmic time is crucial for understanding the evolution of Supermassive Black Holes (SMBHs) and their connection to galaxy formation. A central tool for this analysis is the Luminosity Function (LF), that quantifies the number density of objects within a specific luminosity and redshift range. However, the limited number of radio-detected AGN, particularly at high redshifts (EoR, z>6), makes the construction of robust RLFs a difficult task. While future observatories promise to improve detection rates, efficient methods are needed to analyse the massive datasets they will generate and select relevant sources. 

To address this challenge, Carvajal et al (2023) developed a machine learning (ML) based pipeline that, using multi-band, multi-instrument optical and infrared (IR) photometry, can efficiently identify large numbers of Radio Galaxy candidates and estimate their redshift. This pipeline is also versatile, allowing its application in different regions of the sky with moderately different photometric coverage. 

We will present the results of the application of the ML-based pipeline to IR-detected sources in the area of the Pilot Survey of the Evolutionary Map of the Universe (EMU). Leveraging the expanded number of radio-detectable AGN and SFGs identified with our technique, we have constructed a robust RLF incorporating necessary correction. This RLF has the potential to constrain the density of sources in a broader redshift range than previous results, leading to a deeper understanding of AGN evolution, particularly in the early Universe.

This contribution can be found here (pdf).