Radiation-driven dusty outflows from early galaxies

TL;DR

This work tackles the tension between JWST’s discovery of numerous UV-bright galaxies at $z \gtrsim 10$ and pre-JWST models by proposing radiation-driven dusty outflows as a dust-clearing mechanism. It develops a modified Eddington framework, introducing a boost factor $A$ that encapsulates dust-opacity and gas-gravity effects, and computes $A$ as a function of eight physical parameters (stellar mass, gas fraction, size, age, metallicity, dust type, dust-to-gas ratio, and SED). Applying the model to 20 spectroscopically confirmed $z \gtrsim 10$ galaxies, the authors find three systems likely in an outflow phase with $v_{\infty} \sim 60$–$100$ km s$^{-1}$, and argue that 15 of the remaining galaxies could have experienced prior dusty outflows that effectively displaced dust to larger radii, enabling brighter UV emission. The results imply that radiation-driven dusty outflows can play a significant role in shaping early galaxy evolution and the observed UV luminosity function at cosmic dawn, with predictions testable by ALMA and JWST follow-ups. The study provides a concrete, parameter-dependent map $A(N_H,Z)$ that delineates when dusty outflows become viable, emphasizing metallicity and column density as primary controls.

Abstract

The James Webb Space Telescope (JWST) has discovered an overabundance of UV-bright ($M_{\rm UV} \lesssim -20$), massive galaxies at $z \gtrsim 10$ in comparison to pre-JWST theoretical predictions. Among the proposed interpretations, such excess has been explained by negligible dust attenuation conditions following radiation-driven outflows launched by young stars when a galaxy goes through a super-Eddington phase. Dust opacity decreases the classical Eddington luminosity by a (boost) factor $A$, thus favoring the driving of outflows by stellar radiation in compact, initially dusty galaxies. Here, we compute $A$ as a function of the galaxy stellar mass, gas fraction, galaxy size, and metallicity (a total of 8 parameters). We find that the main dependence is on metallicity and, for the fiducial model, $A \sim 1800(Z/Z_\odot)/(1+N_{\rm H}/10^{23.5}\, {\rm cm^2})$. We apply such results to 20 spectroscopically confirmed galaxies at $z \gtrsim 10$ and evaluate their modified Eddington ratio. We predict that three galaxies are in the outflow phase. Their outflows have relatively low velocities ($60 -100 \,{\rm km\ s^{-1}}$), implying that they are unlikely to escape from the system. For the remaining 17 galaxies that are not currently in the outflow phase, we calculate the past evolution of the modified Eddington ratio from their star formation history. We find that 15 of them experienced an outflow phase prior to observation during which they effectively displaced their dust to larger radii. Thus, dusty outflows driven by stellar radiation appear to contribute to the observed bright UV galaxies at $z > 10$.

Figures (8)

• Figure 1: Schematic illustration of our modeling approach. We assume a spherically distributed gas configuration prior to the outflow (left panel), and calculate the boost factor $A$ to evaluate whether radiation pressure can overcome gravity. For systems with a modified Eddington ratio $\lambda_{\rm E,mod} > 1$ (or above the threshold value $\lambda_{\rm thres}$), we assume they enter the outflow phase and compute the corresponding outflow velocity (right panel).
• Figure 2: Force acting on one proton as a function of radius. We set the fiducal values written in the Table \ref{['table:parameter']}; $r_{\rm e} = 100\, {\rm pc}, \, f_{\rm gas} = 2/3, M_* = 10^8 \, M_\odot$. For the shell geometry, we adopt $r_{\rm in} = 0.7 r_{\rm e}$. The black solid line shows the gravity from the central stellar components. The blue dashed (solid) line represents the gravity from the surrounding gas with uniform (shell) geometry.
• Figure 3: Boost factor as a function of distance from the center. The blue solid (dashed) line represents the case of a uniform gas geometry with MW (SMC)-like dust type. The red solid and dashed lines are the cases of shell gas geometry with the same dust type (MW) but with a different inner radius of $r_{\rm in} = 0.1 r_{\rm e}$ and $0.7 r_{\rm e}$, respectively. The gray lines show the radially dependent optical depth in the uniform-sphere case (right y-axis), $\tau_{\rm ext}(r) =N_{\rm H}(r) \sigma_{\rm ext}$. The other parameter settings ($M_*, f_{\rm gas}, r_{\rm e}, \texttt{age}, Z$) are set as the fiducial ones in Table \ref{['table:parameter']}.
• Figure 4: Boost factor as a function of the galaxy effective radius. The other parameters (gas fraction, age, metallicity, and dust type) are fixed to their fiducial values shown in Table \ref{['table:parameter']}.
• Figure 5: Boost factor as a function of stellar mass. Purple, blue, and green lines represent different gas geometries. The dotted, dashed, solid, and dash-dot lines show $f_{\rm gas}=0.33, 0.50, 0.67(\text{default}), 0.75$, respectively.