\begin{table}%t2 \caption{\label{tab:modelfitting}Model-fitting results.} \small%\centerline { \begin{tabular}{l|ccccccccc|cc|ccc} \hline\hline &&&&&&&&&&&&&& \\[-8pt] & \multicolumn{9}{c|}{Compact component} & \multicolumn{2}{c|}{Extended component} & \multicolumn{3}{c}{Goodness-of-fit} \\ Model & $\theta_{1}$ & $f$ & $\rho$ & $\phi^{(a)}$ & $i$ & $s$ & $R$ &$H/R$ & $\epsilon$ & $\theta_{\rm 2}$& $I_{\rm 2}/I_{\rm 1}$ & $\chi^{2}_{r,V}$ & $\chi^{2}_{r,\Phi}$ & $\chi^{2}_{r}$ \\ & [mas] & & [mas] & [$\deg$] & [$\deg$] & & [AU] & & & [mas] & & & & \\ \hline & \multicolumn{9}{c|}{\bf{Point-symmetric models:}} & & & & & \\ UD & 13.86 & & & & & & & & & & & 24.82 & $29.32^{(c)}$ & 25.92 \\ RING & 8.92 & $0.25^{(b)}$ & & & & & & & & & & 26.38 & $29.32^{(c)}$ & 25.21 \\ GAUSS & 8.55 & & & & & & & & & & & 18.16 & $29.32^{(c)}$ & 20.88 \\ 2-GAUSS & 5.77 & & & & & & & & & 25.78 & 0.59 & 2.35 & $29.32^{(c)}$ & 8.92 \\ & \multicolumn{9}{c|}{\bf{Asymmetric models:}} & & & & & \\ BINARY & $4.8$ & & $6.2$ & $34$ & & & & & & $14.8$ & $1.04$ & 3.47 & 5.32 & 3.93 \\ SKEWED RING & & $0.8$ & & $190$ & $14$ & $0.64$& $0.44$ & & & $27$ & $0.58$ & 1.58 & 3.13 & 1.96 \\ VERTICAL RIM & & & & $132$ & $16$ & & $0.60$ & $0.35$ & & $27$ & $0.68$ & 2.30 & 6.16 & 3.25 \\ CURVED RIM & & & & $180$ & $35$ & & & & $\epsilon_{\rm cr}$ & $32$ & $0.50$ & 1.77 & 3.28 & 2.14 \\ \hline \end{tabular}} \medskip Notes. $^{(a)}$~This column gives the model orientation, measured East of North. For the BINARY~model, $\phi$ gives the PA of the separation vector, while for the SKEWED RING, VERTICAL RIM, and CURVED RIM~models, the orientation of the ellipse/rim major axis is given. Due to a lack of closure phase calibration observations on the E0-G0-H0 array configuration, we are unfortunately not able to unambiguously define the CP~sign, resulting in a~180$\deg$-ambiguity in the derived position angles. $^{(b)}$~In the fitting process, this parameter was fixed. $^{(c)}$~This model is point-symmetric, resulting in a CP which is identical zero. \vspace*{3mm} \end{table}