2022ApJ...924..125C

We present a reanalysis of the Event Horizon Telescope 228 GHz observations of M87. We apply traditional hybrid mapping techniques to the publicly available network-calibrated data. We explore the impact on the final image of different starting models, including the following: a point source, a disk, an annulus, a Gaussian, and an asymmetric double Gaussian. The images converge to an extended source with a size ∼44 µas. Starting with the annulus and disk models leads to images with the lowest noise, smallest off-source artifacts, and better closure residuals. The source appears as a ring, or edge-brightened disk, with higher surface brightness in the southern half, consistent with previous results. Starting with the other models leads to a surface brightness distribution with a similar size, and an internal depression, but not as clearly ringlike. A consideration of visibility amplitudes versus UV-distance argues for a roughly circularly symmetric structure of ∼50 µas scale, with a sharp edge, based on a prominent minimum in the UV-distribution, and the amplitude of the secondary peak in the UV-plot is more consistent with an annular model than a flat disk model. With further processing, we find a possible modest extension from the ring toward the southwest, in a direction consistent with the southern limb of the jet seen on 3mm VLBI images on a factor of few larger scales. However, this extension appears along the direction of one of the principle sidelobes of the synthesized beam, and hence requires testing with better UV-coverage.