Constraining thermal dust emission in distant galaxies with number counts and angular power spectra

TL;DR

This study combines galaxy counts and infrared background anisotropy measurements across a wide frequency range to constrain the evolution of dust properties and luminosity functions in distant galaxies, revealing significant redshift evolution in dust temperature.

Contribution

It provides the first joint analysis of counts and power spectra across multiple instruments to constrain dust evolution and clustering without complex halo models.

Findings

Redshift evolution in thermal dust SED is strongly supported by data.

Dust temperature increases at high redshift, affecting galaxy emission models.

Power spectra can be fitted without complex halo modeling, highlighting potential biases in other analyses.

Abstract

We perform a joint fit to differential number counts from Spitzer's MIPS and Herschel's SPIRE instruments, and angular power spectra of cosmic infrared background (CIB) anisotropies from SPIRE, Planck, the Atacama Cosmology Telescope, and the South Pole Telescope, which together span 220 < \nu / GHz < 4300 (70 < \lambda / \mu m < 1400). We simultaneously constrain the dust luminosity function, thermal dust spectral energy distribution (SED) and clustering properties of CIB sources, and the evolution of these quantities over cosmic time. We find that the data strongly require redshift evolution in the thermal dust SED. In our adopted parametrization, this evolution takes the form of an increase in graybody dust temperature at high redshift, but it may also be related to a temperature - dust luminosity correlation or evolution in dust opacity. The counts and spectra together constrain the evolution of the thermal dust luminosity function up to z ~ 2.5-3, complementing approaches relying on rest-frame mid-infrared observations of the rarest bright objects. We are able to fit the power spectra without requiring a complex halo model approach, and show that neglecting scale-dependent halo bias may be impairing analyses that do use this framework.