Lyman-$α$ Escape through Anisotropic Media

TL;DR

This study investigates Ly$\alpha$ escape through anisotropic, porous neutral gas by performing Monte Carlo radiative transfer in channelized slab geometries that include dust, outflows, and lognormal column density distributions. It shows that Ly$\alpha$ photons do not escape predominantly through the lowest-density channels; instead, they traverse moderately optically thick pathways, producing central flux suppression and non-beamed spectra. Analytical expressions for hole-to-slab flux and spectrum construction—valid for doors, hallways, and filled channels—together with a treatment of outflows and dust, quantify how geometry and physical processes shape observables. Extending to lognormal $N_{\rm HI}$ fields, Ly$\alpha$ probes a broad range of optical depths, with emergent spectra reflecting median-to-mode properties of the density distribution, implying caution when linking Ly$\alpha$ observables to LyC leakage and suggesting that Ly$\alpha$ traces global gas properties rather than just line-of-sight extremes.

Abstract

The escape of Lyman-$α$ (Ly$α$) radiation encodes valuable information on the neutral interstellar medium and is often used as a proxy for the escape of ionizing photons. Yet, the theory of Ly$α$ transfer through anisotropic gas distributions remains underdeveloped. We present Monte Carlo radiative transfer simulations of Ly$α$ propagation through porous, inhomogeneous neutral gas, systematically exploring the effects of channel geometry, outflows, dust, and lognormally distributed column densities. We find that Ly$α$ photons do not preferentially escape through the lowest-column-density pathways, but instead traverse channels of substantial optical depth, leading to suppressed central flux and the absence of strongly beamed escape. Subdividing channels has little impact, indicating that geometry and covering fraction are more important than porosity. Channels containing moderate amounts of neutral hydrogen alter escape in characteristic ways, including the appearance of quadruple-peaked spectra, which can be captured by a simple flux-channel relation. Outflows reshape the spectra by facilitating escape through dense media, redshifting photons and blending central features, while dust modulates the visibility of small channels by suppressing flux at line center; in both cases, we develop an analytical model that predicts the resulting central fluxes. Extending to lognormal column density fields, we show that Ly$α$ photons probe a broad range of optical depths, producing skewed spectra that can be approximated by weighted sums of homogeneous models. Our results have direct implications for using Ly$α$ as a tracer of gas properties and ionizing photon escape; for instance, spectra suggestive of high column densities may nonetheless allow LyC leakage through narrow channels.

Figures (16)

• Figure 1: Simulation setup to study the effect of non-uniform gas distributions on the Ly$\alpha$ profile. The box has periodic boundaries along the $\parallel$-axes and contains a slab with a thin layer of neutral gas and dust. Anisotropy is introduced by piercing the gas layer with a square hole of side length $s$ along the non-periodic axis. The red arrow traces the trajectory of a photon escaping from the slab, while the blue arrow marks a photon exiting through the hole at the maximum angle $\theta$, set by the width $d$.
• Figure 2: Hole-slab flux ratio $\tilde{f}$ as a function of the hole size $\tilde{s}$ to tunnel depth $\rm d$.Points on the left correspond to the “hallway” regime, gradually transitioning to the “door” regime toward the right. The column density is fixed at $N_{\rm HI} = 10^{19}\ \mathrm{cm^{-2}}$. Simulated results closely follow the analytical prediction for $\tilde{f}$ when the geometry factor $\mathrm{g}$ is included (solid lines). For comparison, we also show $\tilde{f}$ without applying the geometry correction (dashed lines).
• Figure 3: Emergent Ly$\alpha$ spectra from slab systems with varying hole geometric configurations. In each panel, the ratio between individual hole size $\bar{s_i}$ and tunnel depth $\rm d$ is held constant, while the total hole area is fixed at $\tilde{s} = 0.04$ and distributed equally across different numbers of holes. Different line colors and thickness indicate the number of holes. The spectra show that as long as the ratio $\bar{s_i}/d$ remains constant, the number of holes does not significantly affect the Ly$\alpha$ line profile.
• Figure 4: Hole-slab flux ratio $\tilde{f}$ as a function of total hole size $\tilde{s}$ keeping a constant ratio of the sub-hole individual size $\bar{s_i}$ and the tunnel depth $\rm d$. The number of holes, marked in the figure with different marker sizes and colors, does not change the behavior of $\tilde{f}$ significantly. Predicted fluxes are marked with different line styles.
• Figure 5: Emergent Ly$\alpha$ spectra from slab systems with optically thick hole ($N_{\rm HI,hole} > 10^{14}\, \rm cm^{-2}$). Left & Center: Effect of varying the column density inside a filled hole on the emergent Ly$\alpha$ spectra. Ly$\alpha$ photons simultaneously trace the high column density of the slab and the lower column density within the hole. When the contrast between the two is significant (more than two orders of magnitude), a distinct central peak emerges. For more minor differences, the inner emission blends with the broader slab features, often resulting in asymmetric red and blue peaks. Right: Example of a Ly$\alpha$ spectrum from a slab with a filled hole. The hole is filled with gas at a column density of $N_{\rm HI_{Hole}} = 10^{16}\ \mathrm{cm^{-2}}$, while the surrounding slab has $N_{\rm HI_{Slab}} = 10^{19}\ \mathrm{cm^{-2}}$. The solid line shows the emergent spectrum from the slab with the filled hole. The black dashed line represents the $\tilde{f}$ scaled combination of two separate slabs with the same respective column densities.