return to star
formation pagereturn to B.
Whitney's homepage
In order to model the dust scattering properties from astronomical sources, we need to know the scattering geometries (to separate the combined effects of scattering phase function vs. albedo) and the amount of scattering material (to determine relative amounts of scattered vs. direct emission). Fortunately, high resolution imagery, such as obtained by the Hubble Space Telescope, allows us to determine the geometry through modeling. You'll notice in the images linked below that the image morphology does not change much between the grain types, though the colors and fluxes do. Thus, by modeling high-resolution imagery, we can determine geometry without knowing dust properties exactly.
The dust grains that we consider here are essentially those of the classical MRN model: spherical and homogeneous. The first two have been derived by fitting the average diffuse interstellar extinction curve (Rv = Av / E(B-V) = 3.1). The distinguishing feature of each is the form of the size distribution. One employs the standard power-law of MRN (Mathis, Rumpl, & Nordsieck 1977; see also Draine & Lee 1984), while the other allows for an arbitrary function (via the Maximum Entropy Method [MEM]; cf. Kim, Martin, & Hendry 1994 [KMH]). Finally, we also include an MEM-derived size distribution for an Rv=5.5 extinction curve, which is consistent with a dark cloud environment.
We have modified our Monte Carlo scattering code (Whitney & Hartmann 1992, 1993) to compute the dust scattering phase functions exactly by sampling from lookup tables. Thus we can conceivably model scattering and polarization of any dust population that can be calculated. Our suite of dust calculation codes currently includes spherical grains (including coated and multilayer), aligned spheroids, randomly-oriented axisymmetric shapes (e.g., spheroids, cylinders, etc.), and the Discrete Dipole Approximation.
Our initial study considers a standard model of a protostar formed from an initially rotating, collapsing cloud (Terebey, Shu, & Cassen 1984 [TSC]). The model parameters are: centrifugal radius = 50 AU, a flared disk with the same radius, an infall rate = 5 x 10-6 Mo/yr, and a curved cavity (z ~ v1.5) presumably carved by outflows, with some dust in it. The Av's through this envelope are 1.3 towards the polar axis and 200 near the equator. The following links show color images made from this model for the two different grain types. We calculated images at VRIJHKL wavelengths (visual through near-IR). Shown here are combinations of the four mid-IR wavelengths, J (1.25 microns), H (1.65 microns), K (2.2 microns) and L (3.45 microns). Notice that combinations with L in them show more striking differences between the grain types. This is because the grain properties diverge more at this wavelength. This shows that L-band images are useful for distinguishing between grain theories.
return to star
formation page
return to B.
Whitney's homepage