An extended dust disk in a spiral galaxy; An occulting galaxy pair in ANGST

TL;DR

This study analyzes an occulting galaxy pair in ANGST data, revealing an extended dusty disk with extinction properties similar to the Milky Way, and discusses methods for modeling overlapping galaxies.

Contribution

First detailed mapping of dust extinction out to 1.5 R extsubscript{25} in a spiral galaxy using HST data, highlighting dust distribution diversity and modeling approaches.

Findings

Dust disk extends to 1.5 R extsubscript{25} with extinction similar to Milky Way

Sampling effects flatten the observed extinction law

Isophotal models are preferred for high-quality overlapping galaxy data

Abstract

We present an analysis of an occulting galaxy pair, serendipitously discovered in ACS Nearby Galaxy Survey Treasury (ANGST) observations of NGC 253 taken with Hubble Space Telescope's Advanced Camera for Survey in F475W, F606W$ and F814W (SDSS-g, broad V and I). The foreground disk system (at z < 0.06) shows a dusty disk much more extended than the starlight, with spiral lanes seen in extinction out to 1.5 R\_25, approximately six half-light radii. This pair is the first where extinction can be mapped reliably out to this distance from the center. The spiral arms of the extended dust disk show typical extinction values of A\_F475W ~ 0.25, A\_F606W ~ 0.25, and A\_F814W ~ 0.15. The extinction law inferred from these measures is similar to the local Milky Way one, and we show that the smoothing effects of sampling at limited spatial resolution (<57 pc, in these data) flattens the observed function through mixing of regions with different extinction. This galaxy illustrates the diversity of dust distributions in spirals, and the limitations of adopting a single dust model for optically similar galaxies. The ideal geometry of this pair of overlapping galaxies and the high sampling of HST data make this dataset ideal to analyze this pair with three separate approaches to overlapping galaxies: (A) a combined fit, rotating copies of both galaxies, (B) a simple flip of the background image and (C) an estimate of the original fluxes for the individual galaxies based on reconstructions of their proper isophotes. We conclude that in the case of high quality data such as these, isophotal models are to be preferred.