J/A+A/707/A352 Escape fraction estimates of EoR galaxies (Giovinazzo+, 2026)
Breaking through the cosmic fog:
JWST/NIRSpec constraints on ionizing photon escape in reionization-era galaxies.
Giovinazzo E., Oesch P.A., Weibel A., Meyer R.A., Witten C., Bhagwat A.,
Brammer G., Chisholm J., de Graaff A., Gottumukkala R., Jecmen M., Katz H.,
Leja J., Marques-Chaves R., Maseda M., Shivaei I., Trabitsch M.,
Verhamme A.
<Astron. Astrophys. 707, A352 (2026)>
=2026A&A...707A.352G 2026A&A...707A.352G (SIMBAD/NED BibCode)
ADC_Keywords: Galaxies ; Redshifts ; Ultraviolet ; Equivalent widths
Keywords: galaxies: high redshift - - early Universe -
dark ages, reionization, first stars
Abstract:
The escape fraction of Lyman continuum photons fesc(LyC) is the last
key unknown in our understanding of cosmic reionization. Directly
estimating the escape fraction (fesc) of ionizing photons in the epoch
of reionization (EoR) is impossible, due to the opacity of the
intergalactic medium (IGM). However, a high fesc leaves clear imprints
in the spectrum of a galaxy, due to reduced nebular line and continuum
emission, which also leads to bluer UV continuum slopes. Here, we
exploit the large archive of deep JWST/NIRSpec spectra from the DAWN
JWST Archive to analyze over 1400 galaxies at 5<zspec<10 and
constrain their fesc based on SED fitting enhanced with a picket fence
model.
We identify 71 high-confidence sources with significant fesc based on
Bayes factor analysis strongly favouring fesc>0 over fesc=0
solutions. We compare the characteristics of this high-escape subset
against both the parent sample and established diagnostics including
UVbeta slope, O32, and SFR surface density.
For the overall sample, we find that most sources have a low escape
fraction (<1%), however, a small subset of sources seems to emit a
large number of their ionizing photons into the IGM, such that the
average fesc is found to be ∼10%, as needed for galaxies to drive
reionization.
Although uncertainties remain regarding recent burstiness and the
intrinsic stellar ionizing photon output at low metallicities, our
results demonstrate the unique capability of JWST/NIRSpec to identify
individual LyC leakers, measure average fesc and thus constrain the
drivers of cosmic reionization.
Description:
In this work, we present the addition of a picket-fence model to
bagpipes as a new tool to recover fesc from the spectra of EoR
galaxies. This allows us to estimate this important parameter for
reionization directly, without the need for indirect tracers
calibrated at lower redshift.
Using public NIRSpec/prism spectra and photometry available in the
DJA, we compile 1428 galaxies at redshifts 5<zspec<10 from a
variety of public programs. We derive the physical properties and fesc
of each source through SED fitting on the spectra, using the
picket-fence model with bagpipes. We perform a second bagpipes run
with fesc=0 and quantify the best fit model with the use of the Bayes
factor. Doing this, we identify 71 galaxies as high-confidence Lyman
continuum emitters, based on their Bayes factor (B12>100).
Table A1: Properties of the sources with high-confidence escape
fraction estimates.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
tablea1.dat 163 71 Properties of the sources with high-confidence
escape fraction estimates
--------------------------------------------------------------------------------
Byte-by-byte Description of file: tablea1.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 47 A47 --- Name Name
49- 52 F4.2 --- z Spectroscopic redshift
54- 65 F12.8 deg RAdeg Right ascension (J2000)
67- 78 F12.8 deg DEdeg Reclination (J2000)
80- 83 F4.2 --- fesc Escape fraction
85- 88 F4.2 --- e_fesc Lower uncertainty on escape fraction
90- 93 F4.2 --- E_fesc Upper uncertainty on escape fraction
95- 98 F4.2 [Msun] logM log10 stellar mass
100-103 F4.2 [Msun] e_logM Lower uncertainty on log stellar mass
105-108 F4.2 [Msun] E_logM Upper uncertainty on log stellar mass
110-114 F5.2 [Msun/yr] logSFR log10 star formation rate
116-119 F4.2 [Msun/yr] e_logSFR Lower uncertainty on log star formation rate
121-124 F4.2 [Msun/yr] E_logSFR Upper uncertainty on log star formation rate
126-131 F6.2 mag UVMAG Absolute UV magnitude (MUV)
133-136 F4.2 mag e_UVMAG Absolute UV magnitude uncertainty (e_MUV)
138-142 F5.2 --- betaUV Beta slope
144-147 F4.2 --- e_betaUV Beta slope uncertainty
149-155 F7.2 0.1nm EWHb Hbeta equivalent width
157-163 F7.2 0.1nm e_EWHb Hbeta equivalent width uncertainty
--------------------------------------------------------------------------------
Acknowledgements:
Emma Giovinazzo, emma.giovinazzo(at)unige.ch
(End) Patricia Vannier [CDS] 30-Jan-2026