Observational and theoretical studies of SiO maser polarization toward late-type evolved stars: Insights from EVPA reversal features

The detailed mechanisms responsible for the polarization of SiO masers toward late-type evolved stars continue to be a subject of debate. Possible primary sources include the local magnetic field or anisotropic pumping, while additional polarization may arise due to conversion from linear to circular modes through scattering or Faraday rotation. Some maser features display an internal rotation in their electric vector position angle (EVPA) by $\triangle \chi \sim \pi/2$ across their extent; such features can provide robust constraints on SiO maser polarization theories as they allow inference of the angle $\theta$ between the magnetic field and the line of sight. Masers remain an important high-resolution probe of the near-circumstellar environment of these objects and reducing uncertainties in maser polarization theory is critical to our understanding of the astrophysics of these regions. In this thesis, I review the current state of SiO maser polarization theory as well as the typical conditions in which maser emission is excited toward Asymptotic Giant Branch stars. I then describe two research studies of SiO EVPA reversal features. The first of these \citep{tobin18,tobin19} concerns the analysis of a single SiO \vs{1}, \Jt{1}{0} maser feature displaying an EVPA reversal of $\triangle \chi > \pi/2$ toward the Mira variable, TX Cam, as observed by the Very Long Baseline Array in five epochs forming part of a prior larger imaging sequence. While we find that the fractional linear polarization profile, $m_l(\theta)$, across the EVPA rotation feature is consistent with the asymptotic theoretical maser polarization solution derived by \citet{GKK} for saturated masers in the limit of small Zeeman splitting, with $\theta$ passing through the critical Van Vleck angle, the corresponding EVPA profile $\chi(\theta)$ rotates too smoothly to arise from this mechanism alone. Beam smoothing at the scale present in the observations is not found to explain this effect. A first-order, quantitative estimate of the effect of variations in $\theta$ along the line of sight, finds however a possible explanation if $\theta$ varies along the line of sight at the level $\triangle \theta \sim 9^{\circ}$. The observed fractional circular polarization, $m_c(\theta)$, across the feature appears to be most consistent with the theoretical prediction from \citet{gray12} that $m_c(\theta) \approxprop \cos \theta$, though significant scatter in $m_c$ occurs at large values of $\cos \theta$. The second research study concerns development of a new theoretical formalism for maser polarization radiative transport more general than several previous approaches and specifically including optional Faraday rotation. This formalism uses the radiative transfer solution method of \citet{LDI87} to integrate the Stokes $\{I,Q,U,V\}$ parameters. This new formalism allows a more complete analysis of the observational results discussed above. Initial results are presented from a one-dimensional numerical implementation of this maser polarization radiative transfer code.These include the finding that Faraday rotation at modest levels cannot smooth out an otherwise instantaneous $\triangle \chi = \pi/2$ EVPA flip expected from the \citet{GKK} solution; however, it can increase the net EVPA change across the feature. We conclude that Faraday rotation may therefore play a second-order but still significant role in reproducing the observed EVPA rotation in the TX Cam feature, as it may explain the measured EVPA rotation $\triangle \chi$ exceeding $\pi/2$. In addition, we find that the resulting circular polarization is consistent with both the prediction for $V/I$ from \citet{gray12} and the prediction for $V / (\partial I / \partial \nu)$ from \citet{watwyld01}, where $\nu$ is the frequency. Further, our derived theoretical profiles for $m_c(\cos \theta)$ are consistent with the observed $m_c(\theta)$ over the TX Cam EVPA reversal feature, with the scatter at large $\cos \theta$ possibly arising from varying optical depth or the sampling of different frequencies with respect to line center along the sampled lines of sight. The two studies described here provide important new constraints on maser polarization theory and open new observational and theoretical avenues for further exploration of this area of research.LimitedAuthor requested closed access (OA after 2yrs) in Vireo ETD syste