Dust contribution to the panchromatic galaxy emission

TL;DR

The paper introduces extit{Mentari}, a pipeline that generates galaxy spectral energy distributions from a semi-analytic model, incorporating a physically motivated dust attenuation model and comparing predictions with observations across multiple wavelengths and redshifts.

Contribution

It presents a new pipeline, extit{Mentari}, that self-consistently models dust effects in galaxy formation simulations and improves multi-wavelength galaxy emission predictions.

Findings

Physically motivated attenuation model outperforms constant attenuation.

Model agrees well with observed luminosity functions and cosmic SED at various redshifts.

Underproduces UV emission at high redshift, indicating areas for model improvement.

Abstract

We have developed a pipeline called \mentari to generate the far-ultraviolet to far-infrared spectral energy distribution (SED) of galaxies from the \dustysage semi-analytic galaxy formation model (SAM). \dustysage incorporates dust-related processes directly on top of the basic ingredients of galaxy formation like gas infall, cooling, star formation, feedback, and mergers. We derive a physically motivated attenuation model from the computed dust properties in \dustysage, so each galaxy has a self-consistent set of attenuation parameters based on the complicated dust physics that occurred across the galaxy's assembly history. Then, we explore several dust emission templates to produce infrared spectra. Our results show that a physically-motivated attenuation model is better for obtaining a consistent multi-wavelength description of galaxy formation and evolution, compared to using a constant attenuation. We compare our predictions with a compilation of observations and find that the fiducial model is in reasonable agreement with: (i) the observed $z=0$ luminosity functions from the far-ultraviolet to far-infrared simultaneously, and hence (ii) the local cosmic SED in the same range, (iii) the rest-frame K-band luminosity function across $0 < z < 3$, and (iv) the rest-frame far-ultraviolet luminosity function across $0 < z < 1$. Our model underproduces the far-ultraviolet emission at $z=2$ and $z=3$, which can be improved by altering the AGN feedback and dust processes in \dustysage. However, this combination thus worses the agreement at $z=0$, which suggests that more detailed treatment of such processes is required.