Strong New Constraints on the Extragalactic Background Light in the Near- to Mid-IR

TL;DR

This study uses gamma-ray observations from Fermi and ground-based telescopes to set new upper limits on the extragalactic background light in the near- to mid-infrared, challenging several existing models.

Contribution

It provides the first combined gamma-ray constraints on the EBL in the near- to mid-IR, ruling out multiple theoretical models with high confidence.

Findings

EBL intensity at 15 microns constrained to 1.36 +/- 0.58 nW/m^2/sr

Several existing EBL models are inconsistent with the new constraints at >3 sigma

Mid-IR EBL predictions are significantly higher than observational limits

Abstract

Direct measurements of the extragalactic background light (EBL) in the near-IR to mid-IR waveband are extremely difficult due to an overwhelming foreground from the zodiacal light that outshines the faint cosmological diffuse radiation field by more than an order of magnitude. Indirect constraints on the EBL are provided by gamma-ray observations of AGN. Using the combination of the Fermi Gamma-Ray Space Telescope together with the current generation of ground-based air Cherenkov telescopes (H.E.S.S., MAGIC, and VERITAS) provides unprecedented sensitivity and spectral coverage for constraining the EBL in the near- to mid-IR. In this paper we present new limits on the EBL based on the analysis of the broad-band spectra of a select set of gamma-ray blazars covering 200 MeV to several TeV. The EBL intensity at 15 microns is constrained to be 1.36 +/- 0.58 nW m^-2 sr^-1. We find that the fast evolution and baseline EBL models of Stecker et al. (2006), as well as the model of Kneiske et al. (2004), predict significantly higher EBL intensities in the mid-IR (15 microns) than is allowed by the constraints derived here. In addition, the model of Franceschini et al. (2008) and the fiducial model of Dominguez et al. (2011) predict near- to mid-IR ratios smaller than that predicted by our analysis. Namely, their intensities in the near-IR are too low while their intensities in the mid-IR are marginally too high. All of the aforementioned models are inconsistent with our analysis at the >3 sigma level.