[Paper Review] Insights into Galaxy Evolution from Mid-infrared Wavelengths

Ranga‐Ram Chary|arXiv (Cornell University)|Dec 26, 2006

Astronomy and Astrophysical Research3 citations

TL;DR

This paper uses mid-infrared surveys from IRAS, ISO, and Spitzer to show that ~70% of star formation at 0.5 < z < 3 is dust-obscured, with AGN contributing less than 20% of the bolometric luminosity density. It demonstrates that dust extinction in high-redshift galaxies (z > 5) inferred from redshifted Hα emission in IRAC 4.5 μm bands indicates dust in ~20% of star-forming galaxies, implying higher true star-formation rates than UV measurements suggest, with implications for reionization requiring either brief, inhomogeneous processes or higher escape fractions than canonical estimates.

ABSTRACT

In this paper, I have attempted to highlight key results from deep extragalactic surveys at mid-infrared wavelengths. I discuss advances in our understanding of dust enshrouded star-formation and AGN activity at 03 will become possible only with future facilities like ALMA. Currently, the presence of dust can only be assessed in a small fraction of the youngest starbursts at z&gt;5 by looking for redshifted large equivalent width H_alpha emission in broadband filters like the IRAC 4.5 micron passband. H_alpha to UV ratios in these objects are a tracer of dust extinction and measuring this ratio in GOODS galaxies indicate dust in ~20% of star-forming galaxies at z&gt;5. Finally, implications for reionization based on the measured stellar mass density and star-formation rates of galaxies at these redshifts are discussed.

Motivation & Objective

• To understand the role of dust-enshrouded star formation and AGN activity in galaxy evolution across cosmic time (0 < z < 3).
• To assess the reliability of mid-infrared emission as a tracer of obscured star formation and its implications for bolometric luminosity density and reionization.
• To evaluate the impact of dust extinction on UV-based star-formation rate measurements at high redshift (z > 5), particularly through redshifted Hα emission.
• To constrain the past star-formation history of high-redshift galaxies using multiband photometry and evolving starburst models.
• To examine the implications of observed stellar mass density and star-formation rates at z ~ 6 for the reionization of the intergalactic medium.

Proposed method

• Utilizes deep extragalactic surveys at mid-infrared wavelengths using IRAS, ISO/ISOCAM, and Spitzer/MIPS to detect dust-enshrouded sources.
• Applies broadband photometry in IRAC 3.6 μm and 4.5 μm bands to identify redshifted Hα emission as a tracer of dust extinction in high-redshift galaxies (z > 5).
• Employs spectral energy distribution (SED) fitting and evolving starburst models to extrapolate past star-formation rates from observed stellar mass and SFR densities at z ~ 6.
• Compares mid-infrared (7 μm) and far-infrared (70 μm) luminosities to calibrate the tight correlation between mid- and far-infrared emission as a star-formation tracer.
• Uses k-correction techniques to estimate rest-frame luminosities, accounting for redshift uncertainties that affect mid-infrared spectral energy distribution interpretation.
• Applies the e-folding time model (Equation 1) to correct for dust extinction and starburst timescale effects in estimating past star-formation rates.

Experimental results

Research questions

• RQ1 What fraction of star formation at 0.5 < z < 3 is obscured by dust, and how does this compare to unobscured star formation?
• RQ2 To what extent do AGN, including obscured ones, contribute to the total bolometric luminosity density at z < 3?
• RQ3 How reliable is the 4.5 μm/3.6 μm flux ratio in identifying dust extinction via redshifted Hα emission in galaxies at z > 5?
• RQ4 What is the implied past star-formation rate history of z ~ 6 galaxies when corrected for dust extinction and e-folding timescales?
• RQ5 What constraints do observed stellar mass and star-formation rate densities at z ~ 6 place on the reionization of the intergalactic medium?

Key findings

• Approximately 70% of the co-moving star-formation rate density at 0.5 < z < 3 is obscured by dust, indicating that mid-infrared surveys are essential for accurate star-formation rate measurements.
• AGN, including obscured sources, contribute less than 20% to the co-moving bolometric luminosity density at z < 3, suggesting star formation dominates the energy output.
• At z ~ 2, star formation is predominantly in massive, ultraluminous infrared galaxies (ULIRGs), while at z ~ 1, it occurs in less luminous LIRGs, indicating a shift in the dominant mode of star formation with redshift.
• About 10–20% of spectroscopically confirmed galaxies at z > 5 show enhanced 4.5 μm flux due to redshifted Hα emission, indicating dust extinction in this fraction of high-redshift starbursts.
• Hα-derived star-formation rates are on average ~3 times higher than UV-based estimates in z > 5 galaxies, implying significant dust extinction and underestimation of true SFRs from UV data alone.
• The measured star-formation rate density at z ~ 6 falls more than a factor of 10 short of the minimum required to reionize the IGM under canonical estimates of escape fraction (fesc ~ 0.15) and clumping factor (C = 30), suggesting either brief reionization or higher escape fractions than assumed.