Dust is pervasive in astronomy. Dust provides a substantial source of cooling in molecular clouds, catalyzes the formation of molecular hydrogen, and shields molecular hydrogen from dissociation by ultraviolet light. Dust absorbs and scatters starlight, reemitting that light in the infrared, modifying the interstellar radiation field and complicating many kinds of astronomical observations.
The amount of light absorbed and scattered by dust as a function of wavelength is known as the extinction curve. The extinction curve is an important probe of the properties of dust grains, for instance their size and chemical composition. The extinction curve must also be known to accurately account for the effect of dust on observations.
In this work, we use the new extinction curve catalog of Schlafly et al. (2016) (shown here) in concert with the 3D dust map of Green et al. (2015) to map out the variation of the extinction curve in three dimensions. The resulting map is intended to test our understanding of the physics of dust in the Galaxy, as well as enable astronomers to better account for the effect of dust on observations.
Kiloparsec-scale Variation in Dust Properties
A major result of this work is that the extinction curve varies substantially on kiloparsec scales throughout the Milky Way Galaxy. The image shows how we measure the steepness of the extinction curve (R(V), represented by color) and the amount of dust (represented by darkness) to vary throughout the Milky Way. The sun is located at the center of the image, and a circle one kiloparsec in size is shown. Clearly the dust is much more clumpily distributed than its properties, which vary smoothly throughout the image!
Many models predict that the extinction curve shape is set largely by the growth of grains in dense molecular clouds. Naively, then, we expect to find one extinction curve in diffuse clouds, and another in denser regions of molecular clouds. What we actually observe, however, is that most of the variation in the extinction curve is correlated over scales of kiloparsecs. These are much larger distances than the roughly 10 parsec scale of dense molecular clouds.
We conclude that our understanding of what drives variation in the extinction curve is incomplete. A new picture for how dust evolves in the Milky Way is necessary, and this picture must be able to accommodate the observed large-scale variations in the dust properties. In particular, it must feature extremely uniform dust properties within molecular clouds in regions with column densities of up to two magnitudes of reddening, while simultaneously having larger variations in dust properties from one kiloparsec-size chunk of the Galaxy to another.