Geometry effects on dust attenuation curves with different grain sources at high redshift

TL;DR

This study explores how dust grain sources and galaxy geometry influence dust attenuation curves in high-redshift galaxies, revealing that attenuation curves alone cannot distinguish dust origins, but IRX–β relations can provide clues.

Contribution

It demonstrates that dust source differences have minimal impact on attenuation curves but significantly affect IRX–β relations, aiding in identifying dust origins in early galaxies.

Findings

Attenuation curves are sensitive to scattering and geometry effects.

Different dust sources produce similar attenuation curves, complicating source identification.

IRX–β relations differ between dust production mechanisms, offering a diagnostic tool.

Abstract

Dust has been detected in high-redshift ($z>5$) galaxies but its origin is still being debated. Dust production in high-redshift galaxies could be dominated by stellar production or by accretion (dust growth) in the interstellar medium. Previous studies have shown that these two dust sources predict different grain size distributions, which lead to significantly different extinction curves. In this paper, we investigate how the difference in the extinction curves affects the dust attenuation properties of galaxies by performing radiative transfer calculations. To examine the major effects of the dust--stars distribution geometry, we adopt two representative cases in spherical symmetry: the well-mixed geometry (stars and dust are homogeneously mixed) and the two-layer geometry (young stars are more concentrated in the centre). In both cases, we confirm that the attenuation curve can be drastically steepened by scattering and by different optical depths between young and old stellar populations, and can be flattened by the existence of unobscured stellar populations. We can reproduce similar attenuation curves even with very different extinction curves. Thus, we conclude that it is difficult to distinguish the dust sources only with attenuation curves. However, if we include information on dust emission and plot the IRX (infrared excess)--$\beta$ (ultraviolet spectral slope) relation, different dust sources predict different positions in the IRX--$\beta$ diagram. A larger $\beta$ is preferred under a similar IRX if dust growth is the dominant dust source.