Image Details
Caption: Figure 16.
Relative Bayesian evidence and median posterior mG-ring model diameter vs. m-order, from the full-track geometric modeling applied to the Sgr A* data; error bars on the diameter measurements show 1σ credible intervals. Because the absolute values of ﹩\mathrm{ln}{ \mathcal Z }﹩ can be substantially different for each data set (i.e., each combination of frequency band and calibration pipeline), and because only the relative values carry information about model specification preferences, we reference all ﹩\mathrm{ln}{ \mathcal Z }﹩ values to the maximum value attained at any m for that data set.
© 2022. The Author(s). Published by the American Astronomical Society.