Dust attenuation in disk-dominated galaxies: evidence for the 2175A dust feature

TL;DR

This study investigates dust attenuation in disk galaxies, revealing that the 2175A dust feature significantly influences UV colors and complicates traditional attenuation models, suggesting the need for revised dust modeling approaches.

Contribution

It provides evidence that the 2175A dust feature affects UV attenuation in disk galaxies, challenging existing models and highlighting the complexity of dust effects in UV observations.

Findings

UV colors are weakly affected by dust in face-on galaxies

Traditional attenuation curves poorly fit UV colors in inclined galaxies

Presence of the 2175A feature likely explains observed trends

Abstract

The attenuation of starlight by interstellar dust is investigated in a sample of low redshift, disk-dominated star-forming galaxies using photometry from GALEX and SDSS. By considering broadband colors as a function of galaxy inclination we are able to confidently separate trends arising from increasing dust opacity from possible differences in stellar populations, since stellar populations do not correlate with inclination. All commonly employed dust attenuation curves (such as the Calzetti curve for starbursts, or a power-law curve) provide poor fits to the ultraviolet colors for moderately and highly inclined galaxies. This conclusion rests on the fact that the average FUV-NUV color varies little from face-on to edge-on galaxies, while other colors such as NUV-u and u-r vary strongly with inclination. After considering a number of model variations, we are led to speculate that the presence of the strong dust extinction feature at 2175A seen in the Milky Way (MW) extinction curve is responsible for the observed trends. Independent of our interpretation, these results imply that the modeling of dust attenuation in the ultraviolet is significantly more complicated than traditionally assumed. These results also imply a very weak dependence of the FUV-NUV color on total FUV attenuation, and we conclude from this that it is extremely difficult to use only the observed UV spectral slope to infer the total UV dust attenuation, as is commonly done. We propose several simple tests that might finally identify the grain population responsible for the 2175A feature.