Broadband spectral energy distributions of SDSS-selected quasars and of their host galaxies: intense activity at the onset of AGN feedback

TL;DR

This study constructs the average spectral energy distribution of high-redshift quasars using multi-wavelength data, revealing features indicative of intense star formation and AGN activity, and suggests far-IR detection as a marker for feedback onset.

Contribution

It provides the first comprehensive mean SED of SDSS quasars across a broad wavelength range, linking IR features to AGN feedback and star formation coevolution.

Findings

Detected IR excess and prominent mid-IR bump in quasar SEDs.

Estimated typical AGN luminosities and star formation rates.

Identified IR-based criteria to select quasars at feedback onset.

Abstract

We present the mean spectral energy distribution (SED) of a sample of optically selected quasars (QSOs) at redshifts of $1 \le z \le 5$. To derive it, we exploit photometric information from SDSS, UKIDSS, and WISE surveys in combination with a stacking analysis of \textit{Herschel}, \textit{AKARI}, and \textit{Planck} maps at the location of the QSOs. The near-UV and optical parts of the reconstructed mean rest-frame SED are similar to those found in other studies. However, the SED shows an excess at 1-2 $\mu$m (when compared to the aforementioned SEDs normalized in the near-UV) and a prominent bump around 4-6 $\mu$m, followed by a decrease out to $\sim 20 \,\mu$m and a subsequent far-IR bump. From the fitted SEDs we estimate the average active galactic nuclei (AGN) luminosity $L_{\rm AGN}$ and star formation rate (SFR) as function of cosmic time, finding typical $L_{\rm AGN} \sim 10^{46} - 10^{47}$ erg/s and SFR $\sim 50 - 1000\, M_{\odot}/$yr. We develop mid-IR based criteria to split the QSO sample, finding that these allow us to move along the average relationship in the SFR vs. $L_{\rm AGN}$ diagram toward increasing AGN luminosities. When interpreted in the context of the in-situ coevolution scenario presented by Lapi et al. 2014, our results suggest that the detection in the far-IR band is an effective criterion to select objects where the star formation is on the verge of being affected by energy/momentum feedback from the central AGN.