All Figures

$\begin{figure} \par\includegraphics[width=7.2cm,clip]{4986f1}\end{figure}$ Figure 1: Approximate optical depths of core formation for the different CO line lists used in Sect. 5 (strong ¹²C¹⁶O) and Sect. 6 (all others). The temperature structure shown is that of the standard 3D model atmosphere (Asplund et al. 2000b).
Open with DEXTER

In the text

$\begin{figure*} \par\includegraphics[width=18cm]{4986f2.eps} \par\end{figure*}$ Figure 2: Example spatially and temporally averaged, disk-centre synthesised strong CO line profiles and bisectors (solid lines without error bars), shown in comparison to ATMOS profiles (diamonds) and bisectors (solid lines with error bars). The profile agreement is quite good, although as discussed in the text this requires an unreasonably high C and/or O abundance; the bisectors based on the new 3D model (see text) show better agreement with observation than those of the original 3D model version (dashed lines). The solar gravitational redshift was removed from the ATMOS spectrum, and synthesised profiles have been apodized and fitted in abundance and Doppler shift.
Open with DEXTER

In the text

$\begin{figure*} \mbox{\includegraphics[width=8cm]{4986f3a}\hspace{5mm} \includeg... ...{4986f3b}\hspace{5mm} \includegraphics[width=8.2cm]{4986f3d} } \par\end{figure*}$ Figure 3: Solar carbon abundances indicated by the weak ¹²C¹⁶O lines, displayed according to equivalent width ( top) and excitation potential ( bottom). On the left, filled circles indicate 3D results and open circles 1DAV results, whilst on the right filled circles are HM values and open circles MARCS results. Trendlines are produced as linear fits to data sets using a minimised $\chi ^2$ method placing equal weight on each point, with solid lines corresponding to filled circles and dashed lines to open circles. No significant trends can be seen in the output of any model.
Open with DEXTER

In the text

$\begin{figure*} \par\mbox{\includegraphics[height=52.5mm]{4986f4a}\hspace{5mm}\i... ...4986f4b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f4d} } \par\end{figure*}$ Figure 4: The same as Fig. 3, but using the LE ¹²C¹⁶O lines. Definite trends can be seen with equivalent width in the output of the 3D and MARCS models. Significantly, the ¹²C abundances implied by the weakest (i.e. lowest formation height) lines in 3D are consistent with the abundances derived using the weak and $\Delta v = 2$ ¹²C¹⁶O lines. The agreement deteriorates with increased LE ¹²C¹⁶O line strength and therefore formation height. Less prominent trends are also evident in excitation potential for all models.
Open with DEXTER

In the text

$\begin{figure*} \par\mbox{\includegraphics[height=52.5mm]{4986f5a}\hspace{5mm}\i... ...4986f5b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f5d} } \par\end{figure*}$ Figure 5: The same as Fig. 3, but using the $\Delta v = 2$ ¹²C¹⁶O lines. Significant trends can be seen with excitation potential in the 1D results, though none in the case of the 3D model.
Open with DEXTER

In the text

$\begin{figure*} \mbox{\includegraphics[height=52.5mm]{4986f6a}\hspace{5mm}\incl... ...4986f6b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f6d} } \par\end{figure*}$ Figure 6: The same as Fig. 3, but indicating ¹³C through the use of the ¹³C¹⁶O lines. No significant trends can be seen in the output of any model.
Open with DEXTER

In the text

$\begin{figure*} \mbox{\includegraphics[height=52.5mm]{4986f7a}\hspace{5mm}\incl... ...4986f7b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f7d} } \par\end{figure*}$ Figure 7: The same as Fig. 3, but indicating ¹⁸O through the use of the ¹²C¹⁸O lines. Significant trends can be seen in equivalent width and excitation potential in the HM and MARCS results.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.6cm,clip]{4986fA1}\end{figure}$ Figure A.1: The effects of medium BNA upon the same 4752.2 nm CO line as shown in Fig. 2. The dotted profile has had no instrumental profile applied to it, whereas the solid line illustrates the effects of medium BNA characterised by $\Delta \sigma$ = 1.5 km s^-1. Convolution with a simple sinc function of $\Delta \sigma$ = 1.5 km s^-1 leaves the line profile virtually identical to its unconvolved counterpart. In the interests of clarity this curve is not shown. Note the far better agreement with the ATMOS spectrum of the apodized than the unapodized profile (reduced $\chi ^2$ measures indicate an order of magnitude difference), vividly demonstrating the importance of correctly emulating instrumental effects when working with this data.

In the text

$\begin{figure} \par\includegraphics[width=7.2cm,clip]{4986f1}\end{figure}$	Figure 1: Approximate optical depths of core formation for the different CO line lists used in Sect. 5 (strong ¹²C¹⁶O) and Sect. 6 (all others). The temperature structure shown is that of the standard 3D model atmosphere (Asplund et al. 2000b).
Open with DEXTER

$\begin{figure} \par\mbox{\includegraphics[height=52.5mm]{4986f5a}\hspace{5mm}\i... ...4986f5b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f5d} } \par\end{figure}$	Figure 5: The same as Fig. 3, but using the $\Delta v = 2$ ¹²C¹⁶O lines. Significant trends can be seen with excitation potential in the 1D results, though none in the case of the 3D model.
Open with DEXTER

$\begin{figure} \mbox{\includegraphics[height=52.5mm]{4986f6a}\hspace{5mm}\incl... ...4986f6b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f6d} } \par\end{figure}$	Figure 6: The same as Fig. 3, but indicating ¹³C through the use of the ¹³C¹⁶O lines. No significant trends can be seen in the output of any model.
Open with DEXTER

$\begin{figure} \mbox{\includegraphics[height=52.5mm]{4986f7a}\hspace{5mm}\incl... ...4986f7b}\hspace{5mm}\includegraphics[height=52.5mm]{4986f7d} } \par\end{figure}$	Figure 7: The same as Fig. 3, but indicating ¹⁸O through the use of the ¹²C¹⁸O lines. Significant trends can be seen in equivalent width and excitation potential in the HM and MARCS results.
Open with DEXTER