Far-Infrared Spectral Energy Distributions and Photometric Redshifts of Dusty Galaxies

TL;DR

This paper presents a method to analyze dusty galaxies' spectral energy distributions using radiative transfer, enabling accurate photometric redshift estimation from far-infrared data, especially useful for upcoming space observatories.

Contribution

It introduces an analytic radiative transfer approach for modeling dusty galaxy SEDs and develops a new method for estimating photometric redshifts from millimeter to far-infrared data.

Findings

Millimeter to FIR SEDs fit well with a dust opacity index of 2.

Photometric redshifts can be estimated with about 10% accuracy.

The method is applicable to compact, centrally heated dusty sources.

Abstract

We infer the large-scale source parameters of dusty galaxies from their observed spectral energy distributions (SEDs) using the analytic radiative transfer methodology presented in Chakrabarti & McKee (2005). For local ultra-luminous infrared galaxies (ULIRGs), we show that the millimeter to far-infrared (FIR) SEDs can be well fit using the standard dust opacity index of 2 when self-consistent radiative transfer solutions are employed, indicating that the cold dust in local ULIRGs can be described by a single grain model. We develop a method for determining photometric redshifts of ULIRGs and sub-mm galaxies from the millimeter-FIR SED; the resulting value of $1+z$ is typically accurate to about 10%. As such, it is comparable to the accuracy of near-IR photometric redshifts and provides a complementary means of deriving redshifts from far-IR data, such as that from the upcoming $\it{Herschel Space Observatory}$. Since our analytic radiative transfer solution is developed for homogeneous, spherically symmetric, centrally heated, dusty sources, it is relevant for infrared bright galaxies that are primarily powered by compact sources of luminosity that are embedded in a dusty envelope. We discuss how deviations from spherical symmetry may affect the applicability of our solution, and we contrast our self-consistent analytic solution with standard approximations to demonstrate the main differences.