The scattering properties of dust provide a unique view into the physical properties of dust grains. This view is different from other probes of dust grains (eg., extinction, polarization due to differential extinction, abundances) as it allows their absorption and scattering properties to be separated. The scattering properties of dust grains are indicators of the size and composition of dust grains.
The scattering of photons by dust grains can be described by a single scattering albedo, a, and a scattering phase function, (), where is the scattering angle. In most studies, () is approximated by the single parameter Henyey-Green (1941) phase function
The g parameter is referred to as the scattering phase function asymmetry and varies from -1 (complete back scattering) to 0 (isotropic scattering) to 1 (complete forward scattering). This Henyey-Greenstein phase function is a good approximation for dust grains, except possible in the far-ultraviolet (Witt 1989). Other analytic forms of the scattering phase function have been proposed (e.g., Cornette & Shanks 1992; Draine 2003b), but have yet to be used in dust scattering studies.
There are three astrophysical objects which are suited for studies of Milky Way interstellar dust scattering properties. These are reflection nebulae, dark clouds, and the Diffuse Galactic Light (DGL). These objects consist of either high-density or diffuse dust illuminated by sufficiently strong sources to permit the detection of scattered light. Reflection nebulae usually consist of a single or small number of stars illuminating their natal cloud. Dark clouds are isolated dust clouds which are usually illuminated by the interstellar radiation field. The DGL is the sum of the light of the stars in the Milky Way scattered by dust in the Galaxy. Other objects which do show scattering light, but which are not suitable for such studies are circumstellar disks, HII regions, and external galaxies. These objects have dust which is newly formed, processed, or non-Milky Way dust.
Each of these three astrophysical objects have different strengths and weaknesses with respect to determining dust scattering properties. The strengths of reflection nebulae are that they are bright, have simple, wavelength independent illuminator geometries, and are illuminated by a single or handful of stars. Their weaknesses are that they often have strong emissions in the red (Extended Red Emission [ERE]) and near-infrared (non-equilibrium, thermal small particle emission). In addition, they usually probe high density regions (RV > 3.1) making the results harder to compare to most dust grain models. The strengths of dark clouds are that they have simple, wavelength independent illuminator geometries and are externally illuminated by the interstellar radiation field and/or a small number of stars. Dark clouds share the same weakness as reflection nebulae as they also probe high density regions (RV > 3.1) and, in addition, they are difficult to observe as they are faint. The strength of the DGL is that it probes the diffuse interstellar medium (RV = 3.1). The weaknesses of the DGL are that the illuminator geometry is wavelength dependent and the dust geometry is complex having a disk density distribution with embedded dust clouds. In addition, DGL observations are challenging as observations of the faint DGL signal are needed over a fairly large region of the sky, especially a range of galactic latitudes, to allow for accurate measurements of a and g. The combination of the results from all three astrophysical objects should give a good view of dust scattering properties without undue uncertainties resulting from each object type's weaknesses.
This review will concentrate on published studies of interstellar dust scattering properties in the ultraviolet, visible, and near-infrared. The topic of X-ray scattering by dust is covered elsewhere in these proceedings. A great deal of work has taken place in this area since the review of dust scattering properties at the last dust meeting (Witt 1989). At that time, only a handful of studies existed which made quantitative determinations of dust scattering properties (ie., actually determined values for a and g). Since the last dust meeting, a number of studies have specifically been addressed to this issue.
Brief summaries of these studies, usually in the form of plots of a and g as a function of wavelength, have been presented in the literature (eg., Gordon et al. 1994; Mathis 1996; Li & Greenberg 1997; Witt & Gordon 2000; Draine 2003a, 2003b). One drawback to these reviews has been the inclusion of all literature results without consideration of the quality of each study. This has lead to a misplaced notion that our knowledge of dust scattering properties is quite uncertain in some wavelength regions. This has been especially pronounced in studies of the DGL in the far-ultraviolet (around 1500 Å). In a handful of these studies, preliminary analyses have been revised in later papers resulting in significantly higher determinations of both a and g. Thus, it is important to apply a uniform set of criteria when reviewing dust scattering studies to ensure the basic quality of the results. This has been done in this review.