Table of Contents
Fetching ...

Mid-IR luminosity functions: inferred dusty cosmic star-formation and black hole accretion histories from the JWST SMILES

Chih-Teng Ling, Tomotsugu Goto, Seong Jin Kim, Cossas K. -W. Wu, Amos Y. -A. Chen, Ece Kilerci, Tetsuya Hashimoto, Terry Long Phan

TL;DR

This study uses the JWST SMILES survey, complemented by deep JADES photometry, to construct comprehensive mid-IR galaxy luminosity functions across $0.5<z<6$ in eight MIRI bands, plus $L_{ m IR}$ and AGN LFs. Through a two-step photometric redshift approach and SED fitting with CIGALE, the authors identify 2607 reliable sources and derive LFs via a $V_{ m max}$ method, anchoring the bright end with literature and performing MCMC fits to obtain evolving faint-end slopes. They find that the faint-end slope of the $L_{ m IR}$ LF steepens with redshift and that the AGN faint-end is relatively flat, revealing sustained obscured AGN activity up to $z\sim5$ and constraining the dust-enshrouded cosmic star-formation rate density and black hole accretion density histories. The results align with ALMA and other JWST studies, highlighting the significant role of obscured star formation and AGN in galaxy evolution and the importance of wide-area JWST surveys to mitigate cosmic variance. Overall, the work provides new, tighter constraints on dusty galaxy evolution and AGN growth across cosmic time, informing models of co-evolution and early black hole assembly.

Abstract

Mid-infrared (mid-IR) observations are crucial for understanding galaxy evolution, tracing star formation, and active galactic nuclei (AGN) activity via dust emission. This work presents mid-IR galaxy luminosity functions (LFs) at $0.5 < z < 6$, derived from the JWST Systematic Mid-infrared Instrument Legacy Extragalactic Survey (SMILES) program. We combine 8 MIRI bands ($5-25$ $μ$m) of SMILES and archival 23-band HST+JWST NIRCam photometry to construct an extensive catalog containing 2,813 galaxies with sub-$μ$Jy level completeness in the mid-IR. We obtain monochromatic (in 5.6, 7.7, 10, 12.8, 15, 18, 21, and 25.5 $μ$m), $L_{\rm IR}$, and AGN LFs, with a limiting luminosity down to $10^{9.5}$ $L_\odot$ at $z=0.5-1.0$, $\sim10^{10.5}$ $L_\odot$ at $z=2.0-4.0$, and to $\sim10^{11}$ $L_\odot$ at $z=4.0-6.0$. With the unprecedented sensitivity and resolution of JWST, here we better constrain the faint-end slope and its evolution of the mid-IR LFs, quantifying the dusty cosmic star formation and black hole accretion histories out to $z \sim 5$. These results provide essential insights to refine our understanding of the obscured star formation and galaxy-AGN co-evolution over cosmic time.

Mid-IR luminosity functions: inferred dusty cosmic star-formation and black hole accretion histories from the JWST SMILES

TL;DR

This study uses the JWST SMILES survey, complemented by deep JADES photometry, to construct comprehensive mid-IR galaxy luminosity functions across in eight MIRI bands, plus and AGN LFs. Through a two-step photometric redshift approach and SED fitting with CIGALE, the authors identify 2607 reliable sources and derive LFs via a method, anchoring the bright end with literature and performing MCMC fits to obtain evolving faint-end slopes. They find that the faint-end slope of the LF steepens with redshift and that the AGN faint-end is relatively flat, revealing sustained obscured AGN activity up to and constraining the dust-enshrouded cosmic star-formation rate density and black hole accretion density histories. The results align with ALMA and other JWST studies, highlighting the significant role of obscured star formation and AGN in galaxy evolution and the importance of wide-area JWST surveys to mitigate cosmic variance. Overall, the work provides new, tighter constraints on dusty galaxy evolution and AGN growth across cosmic time, informing models of co-evolution and early black hole assembly.

Abstract

Mid-infrared (mid-IR) observations are crucial for understanding galaxy evolution, tracing star formation, and active galactic nuclei (AGN) activity via dust emission. This work presents mid-IR galaxy luminosity functions (LFs) at , derived from the JWST Systematic Mid-infrared Instrument Legacy Extragalactic Survey (SMILES) program. We combine 8 MIRI bands ( m) of SMILES and archival 23-band HST+JWST NIRCam photometry to construct an extensive catalog containing 2,813 galaxies with sub-Jy level completeness in the mid-IR. We obtain monochromatic (in 5.6, 7.7, 10, 12.8, 15, 18, 21, and 25.5 m), , and AGN LFs, with a limiting luminosity down to at , at , and to at . With the unprecedented sensitivity and resolution of JWST, here we better constrain the faint-end slope and its evolution of the mid-IR LFs, quantifying the dusty cosmic star formation and black hole accretion histories out to . These results provide essential insights to refine our understanding of the obscured star formation and galaxy-AGN co-evolution over cosmic time.
Paper Structure (17 sections, 11 equations, 21 figures)

This paper contains 17 sections, 11 equations, 21 figures.

Figures (21)

  • Figure 1: Magnitude histograms of our SMILES sample. Each panel displays the distribution for one of the eight MIRI bands. The black dashed line represents the 80% completeness limit of each band, converted to AB magnitude.
  • Figure 2: Comparison between spec-$z$ ($z_\mathrm{spec}$) and EAZY photo-$z$ ($z_\mathrm{phot}$) for 276 sources with available NIRSpec data. The left panel shows the results obtained using only NIRCam photometry (NIR only), while the right panel includes both NIRCam and MIRI photometry (NIR+MIR). Red data points indicate catastrophic outliers. The lower subpanels show the distribution of $\Delta z / (1+z)$, where $\Delta z = z_\mathrm{phot} - z_\mathrm{spec}$. The number of sources (N), the number of outliers and their fraction ($\eta$), and $\sigma_\mathrm{NMAD}$ (with and without outliers) are shown at the top.
  • Figure 3: The stacked redshift (photo-$z$) probability distribution function from EAZY, normalized to 1.
  • Figure 4: Distribution of the reduced $\chi^2$ values from the CIGALE SED fitting for our full sample. The black vertical line indicates the median and the red vertical line marks our criteria of $\chi^2 = 5$. Above this, we consider the SED fit to be unreliable.
  • Figure 5: Photometric redshift distribution of the final 2607 sources. Filled-colored histograms represent the distribution within each of the seven redshift bins used for the luminosity function analysis.
  • ...and 16 more figures