Poster Presentation

A population-level inference of dust attenuation in local galaxies

Author(s): Suchetha Cooray (Stanford), Peter Behroozi (Arizona, NAOJ)

Presenter: Suchetha Cooray (Stanford University)

Dust is critical in inferring the physical parameters of observed galaxy populations from their spectral energy distributions (SEDs). However, our understanding of dust attenuation is limited due to known degeneracies between galaxy properties and dust in SED fitting of single galaxies (dust–metallicity–age degeneracy). However, some degeneracy could be broken by incorporating population-level number density evolution from cosmological structure formation. Thus, we introduce a novel approach to infer population-level dust attenuation properties by statistically matching observed galaxies (SDSS Main Galaxy Sample) with simulated ones from cosmological simulations. We use the star formation histories from UniverseMachine (Behroozi et al. 2019) and star formation-stellar mass-metallicity relations to derive unattenuated SEDs with stellar population synthesis (e.g., Conroy & Gunn 2010). With the matching between attenuated observed SEDs and unattenuated simulated SEDs, we can then use flexible parameterizations to derive the connection between dust and galaxy physical properties. The outcome will be essential for an unbiased understanding of galaxy formation and is part of our efforts towards population-level inference of galaxy SEDs.