Supernovae in 3-D

		Supernovae are important in nearly every branch of astronomy: they tell us what happens to stars at the ends of their lives, they give us clues to the chemical and star-forming evolution of galaxies, and they help us determine distances to faraway galaxies and thereby study the expansion history of the Universe. In recent years, astronomers have begun to probe the three-dimensional characteristics of supernovae, both with high-quality spectropolarimetric studies at large telescopes and with 3-D hydrodynamical modeling on powerful supercomputers. The combination of these cutting-edge developments creates the potential for a major breakthrough in our understanding of supernova explosions.
For such a breakthrough to happen, there must be communication between observational and theoretical lines of research.  I am collaborating with Alex Filippenko's group at UC Berkeley and Peter Nugent at LBNL to develop a Monte Carlo numerical radiative transfer code that will enable this communication. This code, called SLIP, predicts the spectropolarimetric consequences of hydrodynamical models and allow interpretation of observed spectropolarimetric phenomena in light of these models.
In my initial code development and testing, I have focused on Type IIn supernovae. The most luminous stellar explosions, sometimes reaching MV ~ –21, belong to this subclass, and spectra of these objects consistently reveal intense interaction between the supernova ejecta and dense circumstellar matter configurations, presumably created by pre-explosion mass loss episodes from the supernova progenitor. The enormous optical output of the brightest Type IIn supernovae is powered by this interaction rather than by a process intrinsic to the explosion mechanism. Application of a radiative transfer code like SLIP can reveal the characteristics of the circumstellar material surrounding Type IIn supernovae and help us understand the interactions that drive these powerful engines. I am also interested in the ways that spectropolarimetric modeling may help us resolve some of the apparent heterogeneity of this SN subclass and clarify the relation of SNe IIn to other supernovae.
Figure 1		At the recent meeting on Circumstellar Media and Late Stages of Stellar Evolution (Ensenada, September 2006), I presented an invited talk highlighting some of my model results for line polarization in Type IIn supernovae. The code treats not only electron and resonance-line scattering, but also hydrogen absorption and continuous and line emission in the circumstellar shell. Model photons arise from the SN photosphere, the warm circumstellar material, and a hot shock region interior to the CSM. I find that models without emission from the shock region can reproduce the observed compound Hα line profile (consisting of broad and narrow components) seen in SNe IIn. However, in order to reproduce the narrow peak in the polarized flux profile seen in SNe IIn such as SN 2000P and SN 1997eg, emission from the shock region is necessary. Figure 1 shows Hα polarization spectra from a representative model with a toroidal circumstellar envelope at various viewing angles compared with the Hα polarization feature of the Type IIn supernova 1997eg (Leonard et al. 2000; see also below). I am currently preparing a paper investigating how the relative strengths of the narrow Hα line in the flux and polarized flux spectra may serve as a potential diagnostic for the shell's geometry, temperature, and density.
I recently submitted a paper describing an observational spectropolarimetric investigation of the Type IIn SN 1997eg in collaboration with Doug Leonard at SDSU and the Filippenko group. Three separate epochs of Keck spectropolarimetry provide us with a detailed view of the circumstellar shell surrounding this supernova. In the paper, we develop a model for this system consisting of an aspherical broad-line region surrounded by an aspherical (and misaligned) circumstellar scattering region. This picture explains both the striking blue line wings of the Balmer lines in the polarized flux spectrum (Figure 2) and the Q–U loops across Hα, Hβ, and He i λ5876 line profiles. It also supports the hypothesis that some SN IIn progenitors are LBV's and implies that progenitor rotation is not the sole cause of asymmetric supernova explosions.
Figure 2
I will post future results on this page as they become available. For earlier model results, see my AAS posters from January 2005 and January 2006; for a summary of the SN 1997eg results, see my poster from the 2006 Keck science conference.

Hubble Space Telescope image of SN 2004dj by A.V. Filippenko, P. Challis, et al., courtesy NASA, ESA, and STSCI. This material is based upon work supported by the National Science Foundation (NSF) under Grant AST-0302123. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the NSF. For more information, please send email to Jennifer Hoffman. September 20, 2007