Variability selected AGNs in the Chandra Deep Field South region

The Reionization of the Universe is a landmark event in the history of cosmological structure formation and marks the end of the Dark Ages and the beginning of the so called Cosmic Dawn. Measurements of Gunn-Peterson effects on luminous z > 6 quasar spectra (Fan et al. 2006; Becker et al. 2015; McGreer, Mesinger, and D’Odorico 2015) and of the Thomson optical depth τ = 0.054 ± 0.007 from the CMB (Planck Collaboration et al. 2020b), set the boundaries of Reionization Era in the redshift interval z ≈ 6 − 8.5. It is generally thought that the Intergalactic Medium (IGM) is kept ionized by the integrated UV emission from Active Galactic Nuclei (AGNs) and Star Forming Galaxies (SFG), but the relative contribution of these sources as a function of cosmic epoch are poorly known at present. The numerous population of high redshift SFGs suggested them as main candidates for the emission of the UV ionizing radiation. However, measures of low Lyman Continuum escape fraction fesc(LyC) (e.g. Grazian et al. 2017) for these sources questioned their capability to ionize the Universe. AGNs have, indeed, a very high ionizing power, but their space density shows a steep decrease for z > 3. The sough for high-z AGNs at z > 4 has been carried out mainly through selections based on X-ray or UV-optical colors, but recent works (Schindler et al. 2019b; Schindler et al. 2019a) have shown that these criteria are not 100% complete, and they could be complemented by other selection criteria, e.g. based on variability. I developed a variability-selection method, based on the statistical parameter Median Absolute Deviation (MAD), to search for a population of faint AGNs that have not been detected by other selection techniques. I used all available observations of Chandra Deep Field South (CDF-S) obtained with the VIMOS/VLT instrument in R-band to create a set of stacked images, one for each year, with a time coverage extending from year 2007 to 2015. The R-band dropout limits the exploitation of this method to the z < 5 redshift boundary, after Reionization has been completed, but it shows its potential in a well known and well studied field as the CDF-S. I extracted a catalogue from each of the single epoch images and compared them to each other. I could then study the magnitude variability of each source among different epochs. The method described here selected a sample of 73 variable sources in the CDF-S brighter than a 90% completeness limit of R ≤ 26 magnitude. Five of these variable sources were previously identified by other four variability-based catalogues from the recent literature. Small overlaps were also found by comparing reference catalogues to each other. This confirms a strong dependency of variability-selection results from the method characteristics such as the photometric bands observed, the sampling time-scale and the statistical methods adopted. Therefore, results obtained via a single instrument/dataset cannot be considered exhaustive for variable objects selection. A small overlap were also found between the catalogue of variable sources discussed in this work with X-ray surveys from Chandra and XMM. Only 10 of the 73 variable sources show a detectable X-ray emission. The source characterization throughout a cross-match with the multiwavelength surveys Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) and Multiwavelength Survey by Yale-Chile (MUSYC) provides a redshift value (spectroscopic or photometric) for the majority of variable sources of this work. The distribution of source magnitudes with redshift shows that X-ray emitter sources are just the tip of the iceberg of a more numerous AGN population, selected through variability, that is typically fainter in the optical band. The optical variability is able to identify a large number of lowluminosity AGNs, especially at high redshifts, including those that remain undetectable even at the deepest observations in the X-rays. New instruments for surveys are in their final phase of construction, as the Vera Rubin Observatory and Euclid. Variability will be one of the most powerful tool to detect AGNs in their data. They will provide multi-epoch observations in the optical and near infrared bands, that are needed to explore up to z ∼ 7, at the Reionization epoch.