The IRX-$\beta$ relation: Insights from simulations

TL;DR

This study uses galaxy simulations to explore the IRX-$eta$ relation, revealing how dust properties, galaxy type, and redshift influence UV slopes and infrared excess, challenging some previous assumptions about high-redshift dusty galaxies.

Contribution

It provides new insights into the IRX-$eta$ relation by analyzing dust effects, galaxy evolution, and star formation activity through detailed simulations across different redshifts.

Findings

Dust type significantly affects UV slope predictions.

High-redshift galaxies often lie above the local IRX-$eta$ relation.

Simulated high-z galaxies resemble observed dusty star-forming galaxies.

Abstract

We study the relationship between the UV continuum slope and infrared excess (IRX$\equiv L_{\rm IR}/L_{\rm FUV}$) predicted by performing dust radiative transfer on a suite of hydrodynamical simulations of galaxies. Our suite includes both isolated disk galaxies and mergers intended to be representative of galaxies at both $z \sim 0$ and $z \sim 2-3$. Our low-redshift isolated disks and mergers often populate a region around the the locally calibrated \citet[][M99]{M99} relation but move well above the relation during merger-induced starbursts. Our high-redshift simulated galaxies are blue and IR-luminous, which makes them lie above the M99 relation. The value of UV continuum slope strongly depends on the dust type used in the radiative transfer calculations: Milky Way-type dust leads to significantly more negative (bluer) slopes compared with Small Magellanic Cloud-type dust. The effect on $\beta$ due to variations in the dust composition with galaxy properties or redshift can dominate over other sources of $\beta$ variations and is the dominant model uncertainty. The dispersion in $\beta$ is anticorrelated with specific star formation rate and tends to be higher for the $z \sim 2-3$ simulations. In the actively star-forming $z \sim 2-3$ simulated galaxies, dust attenuation dominates the dispersion in $\beta$, whereas in the $z \sim 0$ simulations, the contributions of SFH variations and dust are similar. For low-SSFR systems at both redshifts, SFH variations dominate the dispersion. Finally, the simulated $z \sim 2-3$ isolated disks and mergers both occupy a region in the \irxbeta\ plane consistent with observed $z \sim 2-3$ dusty star-forming galaxies (DSFGs). Thus, contrary to some claims in the literature, the blue colors of high-z DSFGs do not imply that they are short-lived starbursts.