Presentation #353.12 in the session “Computational Augmentation to Observations”.
The Event Horizon Telescope (EHT) has recently provided the first horizon-scale images of a black hole, opening a new era of black hole astrophysics. Planned enhancements to the EHT will make it more sensitive to both finer-scale and more extended emission, furthering our understanding of black holes. However, the current imaging techniques may not be optimal when it comes to handling such multi-scale emission structures, since they were designed to handle single-scale emission. In this work, we explored a multi-scale imaging technique for radio interferometry, by utilizing the sparsity in the wavelet domain. The wavelet transform decomposes natural images into sparse sets of coefficients across multi scales, allowing the sparse regularization to more effectively constrain both compact and extended emission. We first demonstrated that multi-scale emission structures in various astronomical images can be efficiently denoised by enforcing the sparsity on the wavelet domain. Then, extending the algorithm to radio interferometric data, we applied it to simulated observations of general relativistic magnetohydrodynamic (GRMHD) simulations of M87 with the expanded EHT array that is anticipated to be fielded in 2025. We found the new technique reconstructs the photon ring, the surrounding accretion flow with spiral arms and extended jet structures with a dynamic range of ~1000. The results indicate that this is a viable method worthy of further investigation. This work was supported by grants (AST-1440254, AST-1614868, AST-1950348, AST-2034306) from the National Science Foundation