PRIMO algorithm provides improved reconstruction of M87 black hole image

Scientists have presented a new reconstruction of the famous image of the supermassive black hole at the center of the galaxy M87 using a novel algorithm called PRIMO. The Event Horizon Telescope (EHT) captured the original image in 2017, but the sparsity of the data made image reconstruction challenging. PRIMO uses a large library of synthetic images from general relativistic magnetohydrodynamic simulations of accreting black holes as a training set to overcome this challenge. By learning correlations between different regions of the space of interferometric data, PRIMO produces high-fidelity images even in the presence of sparse coverage.

The new reconstructed image shows a thin, bright ring with a diameter of 41.5 ± 0.6 micro-arcseconds, which is at least two times smaller than previously reported. This improvement has important implications for measuring the mass of the central black hole in M87 based on the EHT images. The ring-like shape of the image is due to the observed dependence of the interferometric visibility amplitudes on baseline length, which has the characteristic shape of a Bessel function. The width of the ring-like image is another important diagnostic that can be used to constrain physical characteristics of the accretion flow, such as the accretion rate.

Fig: The PRIMO image offers a superior utilization of the resolution and dynamical range of the EHT array.

Previous imaging algorithms have struggled to accurately determine the width of the ring due to the sparsity of interferometric data at large baseline lengths. However, PRIMO's training with synthetic images eliminates the need for regularizers and produces reconstructions that reach the nominal resolution of the EHT array. As such, the reconstructed image has a ring width of 9.6 ± 0.5 micro-arcseconds, which corresponds to a fractional width of ≲0.23 that is a substantial improvement compared to earlier inferences. This improvement will lead to reduced errors in the inferred mass of the M87 black hole based on the reconstructed image.

Future EHT observations with additional telescope locations and higher bandwidth will allow for even further improvements in the effective resolution of the image, allowing for better image morphology constraints.

DOI: 10.3847/2041-8213/acc32d