The Lyman-alpha and Continuum Origins Survey II: the connection between the escape of ionizing radiation and Lyman-alpha halos in star-forming galaxies

Abstract

One of the current challenges in galaxy evolution studies is to establish the mechanisms that govern the escape of ionizing radiation from galaxies. Here, we investigate the connection between Lyman Continuum (LyC) escape and the conditions of the Circumgalactic Medium (CGM), as probed by Ly$α$ halos (LAHs) in emission. We use Ly$α$ and UV continuum imaging data from the Lyman alpha and Continuum Origins Survey (LaCOS), targeting 42 nearby ($z \simeq 0.3$), star-forming galaxies with LyC observations (escape fractions of $f_{\rm esc}^{\rm LyC} \simeq 0.01-0.49$). LaCOS galaxies show extended Ly$α$ emission ubiquitously, with LyC emitters (LCEs) having more compact Ly$α$ morphologies than non-LCEs, and Ly$α$ spatial offsets that do not exceed the extent of the UV continuum. We model the diffuse LAHs using a combined Sérsic plus exponential 2D profile, and find that the characteristic scale length of the Ly$α$ halo is ten times larger than the UV, on average. We unveil a significant anti-correlation between $f_{\rm esc}^{\rm LyC}$ and the Ly$α$ Halo Fraction (HF, or contribution of the halo to the total Ly$α$ luminosity), that we propose as a new LyC indicator. Our observations show that halo scale lengths and HFs both scale positively with the optical depth of the neutral gas in the ISM, revealing a picture in which Ly$α$ and LyC photons in LCEs either emerge directly from the central starbursts or escape isotropically and, in the case of Ly$α$, minimize the number of scattering interactions in a less-extended CGM.

Figures (12)

• Figure 1: LyC-to- Ly$\alpha$ properties of the LaCOS sample. Ionizing escape fraction (_ esc^ LyC $f_{\rm esc}^{\rm LyC}$) versus the Ly$\alpha$ equivalent width in the rest-frame (_ Lyα $W_{\rm Ly\alpha}$, left), and the Ly$\alpha$ escape fraction (_ esc^ Lyα $f_{\rm esc}^{\rm Ly\alpha}$, right). Solid circles and downward triangles show LaCOS detections and upper limits, while shaded symbols in the background display other low-$z$ samples in the literature: Flury2022a, Izotov2016aIzotov2016bIzotov2018aIzotov2018bWang2019Izotov2021 at $z \simeq 0.3$, and the lensed LAEs at $z \simeq 2.3$ by Citro2025. The solid lines draw the empirical relations from Pahl2021 at $z \simeq 3$ and Izotov2024 at $z \simeq 0.3$. The _ Lyα$W_{\rm Ly\alpha}$ and _ esc^ Lyα$f_{\rm esc}^{\rm Ly\alpha}$ are among the most robust one-dimensional _ esc^ LyC$f_{\rm esc}^{\rm LyC}$ indicators. Yet, the scatter in these relations is large, and the calibrations disagree between different redshifts.
• Figure 2: Extended Ly$\alpha$ emission in LaCOS galaxies (${\rm 15~kpc \times 15~kpc}$ cutouts). A smoothed version of the Ly$\alpha$ emission is shown in blue (${\rm arcsinh}$ scale), with the orange contours depicting the more compact, UV continuum counterpart (applying a 5 pix Gaussian filter). White labels show the galaxy ID, measured LyC escape fraction for every object (_ esc^ LyC $f_{\rm esc}^{\rm LyC}$, including upper limits), and the estimated Ly$\alpha$ Halo Fraction (HF) (see Sect. \ref{['sec:results_modeling']}). Panels are sorted by ascending _ esc^ LyC $f_{\rm esc}^{\rm LyC}$.
• Figure 3: Comparison between the extent of Ly$\alpha$ and UV emission in LaCOS, as measured by the half-light radius ($r_{50}$). Filled and open circles show LaCOS LyC detections and upper limits, respectively. For reference, similar measurements from the eLARS survey are plotted in pink diamonds Melinder2023, with the solid line indicating the one-to-one relation. Histograms show the size distributions projected on each axis, with the error bars encompassing the interquartile range.Error bars represent the characteristic (median) uncertainty on each axis. With respect to their UV counterpart, LaCOS galaxies show extended Ly$\alpha$ emission almost ubiquitously.
• Figure 4: The relation between the ionizing escape fraction (_ esc^ LyC$f_{\rm esc}^{\rm LyC}$) and the extent of the Ly$\alpha$ emission (left), and the Ly$\alpha$-to-UV size ratio (right).Filled circles and downward triangles show LaCOS LyC detections and upper limits, respectively. The LCE detection fraction is also shown through squared open symbols in the right vertical axis. The results from the survival Kendall correlation test, including censored data, can be found in the inset. Linear fits to the decline in _ esc^ LyC$f_{\rm esc}^{\rm LyC}$ with the size of both the UV continuum and Ly$\alpha$ are plotted in blue and gray lines (Eq. \ref{['eq:fesc_rUV']} and \ref{['eq:fesc_rLya']}). While these linear fits indicate that LCEs may have more compact Ly$\alpha$ than non-LCEs respect to the UV continuum (left), individual data points do not reflect this behavior (right), highlighting the lack of ability of simple size measurements to fully reproduce the morphology of the Ly$\alpha$ emission, and the need of a more sophisticated modeling (see Sect. \ref{['sec:results_modeling']}).
• Figure 5: Ionizing escape fraction (_ esc^ LyC$f_{\rm esc}^{\rm LyC}$) versus the 20%- and 90%-light radius, and the concentration of the Ly$\alpha$ emission, defined as $C_{\rm \rm Ly\alpha Ly$α$} = r_{90}^{\rm \rm Ly\alpha Ly$α$}/r_{20}^{\rm \rm Ly\alpha Ly$α$}$. The LCE fraction (tentatively) increases towards more compact galaxies in Ly$\alpha$, due to the underlying correlation between _ esc^ LyC$f_{\rm esc}^{\rm LyC}$ and Ly$\alpha$$r_{20}$.