Why is colour special? Understanding the differences between red and blue quasars

Abstract

The majority of quasars have very blue ultraviolet-optical colours due to the unobscured view of the accretion disc. However, recent work has uncovered a significant population of quasars with much redder colours ("red quasars") which are prime candidates for a transitional phase in the evolution of quasars. In this thesis, I build on the work from Klindt et al. (2019), who discovered fundamental differences in the 1.4 GHz radio properties of a sample of Sloan Digital Sky Survey (SDSS) optically selected red quasars, compared to typical blue quasars, adding weight to this evolutionary paradigm. I aim to further investigate these differences, using higher spatial resolution and deeper radio data, and utilise optical-near-infrared spectroscopy in order to understand the physical origin of the red colours and mechanisms driving the radio emission in red quasars.

Firstly, I investigate the radio properties of red quasars across a broad range of frequencies (0.144-3GHz) and sensitivities (~20-750Jy/bm at 1.4GHz). I build on the results from Klindt et al. (2019), finding that SDSS-selected red quasars display enhanced radio emission compared to typical blue quasars, which peaks around the radio-quiet threshold across all of the radio datasets utilised, but at a higher statistical significance than previous studies (Chapters 3 and 4). Using the high spatial resolution of VLA Stripe 82, I find that this enhanced radio emission is compact/unresolved at ~16 kpc (host-galaxy) scales (at z~1.5). Utilising the radio-far-infrared correlation, I show that the origin of this radio emission is likely due to AGN processes, rather than star formation (Chapter 3).

To explore the origin of the red colours and the enhanced radio emission in red quasars, I utilise ultraviolet-near-infrared VLT/X-shooter medium resolution spectroscopy of 40 red and blue quasars (Chapter 5). From fitting a dust-reddened control composite to the spectra, I find that the majority of red quasars are indeed red due to the line-of-sight dust, also consistent with the larger Balmer decrements calculated for the red quasars. Only one Fermi-detected red quasar cannot be fit well by any dust extinction curve, and is instead likely to be synchrotron dominated. Fitting a simple thin accretion disc model to both the red and blue quasars, I identify no significant differences in the accretion properties, suggesting that this is not driving the enhanced radio emission. Finally, by fitting the [OIII]5007 and CIV emission lines, I explore the incidence of outflows in the red and blue quasars. I detect outflows in both the red and blue quasars, but do not find any evidence to suggest outflows are more powerful or prevalent in the red quasar sample.

Extending this work to more obscured systems, I present the first eight months of our program within the Dark Energy Spectroscopic Instrument (DESI), which selects dust-reddened quasars predominantly based on their mid-infrared colours (Chapter 6). So far, 3038 candidates have been observed. From visual inspection of their spectra, I find that 74% of the high quality objects are quasars, consistent with the main DESI quasar survey. Defining red quasars based on their optical colour, and “Extremely Red Quasars” (ERQs) based on their optical–mid-infrared colour, I discover that 76% and 38% of the quasars in our sample are red quasars and ERQs, respectively. From fitting dust extinction curves, I find that my sample doubles the number of quasars with an mags, compared to the nominally selected quasars. Exploring the 144MHz radio properties, I identify a striking relationship between the line-of-sight dust extinction towards a quasar and the radio detection fraction, that is not driven by redshift or luminosity effects.

Overall, my thesis demonstrates an intrinsic connection between opacity and the production of radio emission in quasars. Future work will help to uncover the physical mechanisms behind the enhanced radio emission in dusty quasars.