All Figures

$\begin{figure} \par\includegraphics[width=8.8cm,clip]{7168fig1.ps}\par\vspace*{2mm} \includegraphics[width=8.8cm,clip]{7168fig2.ps}\end{figure}$ Figure 1: Visual lightcurve of S Ori as a function of Julian Date and stellar cycle/phase. Data are from the AAVSO (Henden et al. 2006) and AFOEV (CDS) databases. We adopt a period of $P=430~{\rm days}$ and JD of last maximum brightness T₀=2453190. The arrows indicate the dates of our VLTI and VLBA observations. a) The most recent 5-6 cycles illustrating the correspondence of the data with our adopted P and T₀ values. The indicated dates of observations are the epochs from BW05 (JD between 2600 and 2800) and the epochs of the present paper (JD between 3300 and 3800). b) Enlarged view covering our VLTI/MIDI and VLBA observations presented in this paper. The dates of observations are combined into 4 epochs A, B, C, D.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.6cm,clip]{7168fig3.eps} \end{figure}$ Figure 2: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch A, stellar phase 0.42. Shown are our MIDI data compared to the best-fitting model with parameters listed in Table 3. In each panel, the x-symbols with error bars denote the measured values, the solid lines the synthetic values of the overall model consisting of the stellar model atmosphere and the dust shell. The dotted lines denote the contribution from the attenuated atmosphere model alone, and the dashed lines the contribution of only the dust shell (attenuated input radiation subtracted). a) The calibrated MIDI flux spectrum. b) The calibrated MIDI visibility values as a function of wavelength. c) The calibrated MIDI visibility values as a function of spatial frequency for the example of three spectral channels at wavelengths 8.5 $\mu$ m, 10.5 $\mu$ m, and 12.5 $\mu$ m. d) The equivalent UD diameter values. Also indicated are the photospheric angular diameter $\Theta _{\rm Phot}$ and the inner dust shell boundary $\Theta _{\rm in}$ as obtained by our model fit, as well as the mean 42.8 GHz and 43.1 GHz SiO ring angular diameters (Sect. 4). e) The model intensity profiles at spectral bands 8-9 $\mu$ m, 10-11 $\mu$ m, and 12-13 $\mu$ m. The bottom abscissa shows the radius as a function of the photospheric radius $R_{\rm Phot}$ , and the top abscissa the corresponding angular scale obtained from our model fit. The arrows indicate the location of the inner boundary radius of the dust shell $R_{\rm in}$ , as well as of the mean 43.1 GHz and 42.8 GHz SiO maser spot radii R_43.1 and R_42.8.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig4.eps}\end{figure}$ Figure 3: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch B, stellar phase 0.55. As Fig. 2.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig5.eps}\end{figure}$ Figure 4: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch C, stellar phase 1.16. As Fig. 2.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig6.eps}\end{figure}$ Figure 5: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch D, stellar phase 1.27. As Fig. 2.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.6cm,clip]{7168fig7.eps}\end{figure}$ Figure 6: Density profiles of radiatively driven wind models calculated with the DUSTY code (Ivezic & Elitzur 1997; Ivezic et al. 1999) for input parameters from Table 3, together with $\rho \propto r^{-2}$ and $\rho \propto r^{-3}$ curves.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=16cm,clip]{7168fig8.eps}\end{figure}$ Figure 7: The (red) v=2, J=1-0 (42.8 GHz) and (green) v=1, J=1-0 (43.1 GHz) maser images overlaid onto pseudo-color representations of the infrared intensity. The continuum photosphere (in fact mostly hidden behind the molecular atmosphere in the N-band) is enhanced to a light blue color. The darker blue shades represent our model intensity profile as in panels (e) of Figs. 2-5, and the green shades represent the location of the Al₂O₃ dust shell on top of the low-intensity extended wings of the molecular atmosphere. The true location of the star relative to the maser images is unknown. Here, we assume that the center of the star coincides with the center of the maser spot distribution. Synthesized beam sizes for epoch A are $0.59\times 0.40$ mas and $0.60\times 0.37$ mas for the v=1 and v=2 transitions, respectively. For epoch B these sizes are $0.52\times 0.21$ mas and $0.51\times 0.20$ mas, and for epoch C $0.46\times 0.16$ mas and $0.44\times 0.16$ mas, respectively. Epoch D is a MIDI-only epoch and SiO maser observations have not been obtained at this epoch. The size of each panel is $30\times 30$ mas, corresponding to $14.4\times 14.4$ AU.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{7168figa.eps}\hspace*{2mm}... ...eps}\hspace*{2mm} \includegraphics[width=5.6cm,clip]{7168figf.eps} \end{figure}$ Figure 8: LOS velocity structure of the ( top) v=1, J=1-0 (43.1 GHz) and ( bottom) v=2, J=1-0 (42.8 GHz) SiO maser emission toward S Ori. The epochs of observations are from left to right Epoch A ( $\Phi _{\rm vis}=0.46$ ), Epoch B ( $\Phi _{\rm vis}=0.56$ ), and Epoch C ( $\Phi _{\rm vis}=1.14$ ). The top sub-panels of each epoch and transition show the spectra formed by plotting maser intensity versus velocity, color-coded in 1.7 km s^-1 velocity increments from redward to blueward. The solid lines in the top sub-panels represent the scalar-averaged cross-power spectra averaged over all of the VLBA antennas. The bottom (main) sub-panels plot the spatial and velocity distributions of the masers. The color of each point represents the corresponding velocity bin in the spectrum, and the size of each point is proportional to the logarithm of the flux density. Errors in the positions of the features are smaller than the data points. The dashed circles are based on the mean angular distances of the SiO masers from the centers of the distributions. The light blue filled circles in the centers illustrate for comparison the angular sizes of the continuum photosphere derived in Sect. 3.4. The true location of the star relative to the masers is unknown and here assumed to coincide with the center of the maser spot distribution. We use also panels (e) of Figs. 2-5, as well as Fig. 7, for a comparison of the SiO maser ring radii to the extension of the molecular atmosphere and the inner dust shell boundary.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168figg.eps}\end{figure}$ Figure 9: Histogram of the LOS velocities of our SiO maser spots combining VLBA epochs A-C. Panel a) shows the histogram for the v=2 42.8 GHz masers and panel b) for the v=1 43.1 GHz masers. Also indicated are the mean values $\bar{V}_{v=1}$ and $\bar{V}_{v=2}$ . $\bar{V}_{v=1}$ is slightly redder than $\bar{V}_{v=2}$ .
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168figh.eps}\end{figure}$ Figure 10: Distances of the 43.1 GHz maser components from the common center of the distribution (radius) versus their velocity relative to the average LOS velocity for the 43.1 GHz transition ( $V_{\rm LOS} - \bar V_{\rm LOS}$ ). Maser components of the different epochs are denoted by different symbols and colors. The plot indicates that higher-velocity masers are found closer to the star. This trend can be explained by means of expanding maser shells. The curves represent shells at the inner and outer boundaries of the SiO maser shell using (solid lines) a logarithmic velocity gradient and (dashed lines) constant velocity expansion. See text for details.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168figi.eps}\end{figure}$ Figure 11: As Fig. 10, but for the 42.8 GHz transition. The axes' scales are identical for the two figures. Compared to Fig. 10, it is evident that the 42.8 GHz maser spots lie at systematically closer distances to the center of the common maser distribution than the 43.1 GHz maser spots.
Open with DEXTER

In the text

$\begin{figure} \par\resizebox{8cm}{!}{\includegraphics{7168figj.eps}}\hspace*{1cm} \resizebox{8cm}{!}{\includegraphics{7168figk.eps}}\end{figure}$ Figure 12: Sketch of the radial structure of S Ori's CSE at ( left) near-minimum and ( right) post-maximum visual phase as derived in this work. Shown are the locations of the continuum photosphere (dark gray), the at N-band optically thick molecular atmosphere (medium dark gray), the at N-band optically thin molecular atmosphere (light gray), the Al₂O₃ dust shell (dashed arcs), and the 42.8 GHz and 43.1 GHz maser spots (circles/triangles). The numbers below and beside the panels are the mean values of ( left) epochs A & B and ( right) epochs C & D from Table 7.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig4.eps}\end{figure}$	Figure 3: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch B, stellar phase 0.55. As Fig. 2.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig5.eps}\end{figure}$	Figure 4: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch C, stellar phase 1.16. As Fig. 2.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168fig6.eps}\end{figure}$	Figure 5: Results of our VLTI/MIDI 8-13 $\mu$ m interferometry of S Ori at epoch D, stellar phase 1.27. As Fig. 2.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8.6cm,clip]{7168fig7.eps}\end{figure}$	Figure 6: Density profiles of radiatively driven wind models calculated with the `DUSTY` code (Ivezic & Elitzur 1997; Ivezic et al. 1999) for input parameters from Table 3, together with $\rho \propto r^{-2}$ and $\rho \propto r^{-3}$ curves.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168figg.eps}\end{figure}$	Figure 9: Histogram of the LOS velocities of our SiO maser spots combining VLBA epochs A-C. Panel a) shows the histogram for the v=2 42.8 GHz masers and panel b) for the v=1 43.1 GHz masers. Also indicated are the mean values $\bar{V}_{v=1}$ and $\bar{V}_{v=2}$ . $\bar{V}_{v=1}$ is slightly redder than $\bar{V}_{v=2}$ .
Open with DEXTER

$\begin{figure} \par\includegraphics[width=8.5cm,clip]{7168figi.eps}\end{figure}$	Figure 11: As Fig. 10, but for the 42.8 GHz transition. The axes' scales are identical for the two figures. Compared to Fig. 10, it is evident that the 42.8 GHz maser spots lie at systematically closer distances to the center of the common maser distribution than the 43.1 GHz maser spots.
Open with DEXTER