Black holes would seem impossible to visualize: they emit no light and reflect none. They have an event horizon, which, in a rigorous sense, marks the edge of the universe. Despite this, in 2019, the Event Horizon Telescope (EHT) directly imaged the billion-degree gas orbiting the supermassive black hole M87*. Now, a team from that effort is working to build a space telescope (Black Hole Explorer, BHEX) to join terrestrial telescopes to observe the photon ring, a shell of pure light that captures everything there is to know about a black hole: mass and spin.
However, creating an image of a photon ring boils down to solving a complex Bayesian inverse problem. Solving this inverse problem will push statistical methods to the brink. The imaging model has millions of parameters that are non-linearly related to the data, a hierarchical model structure, and requires analyzing petabytes of data. In this talk, we will present efforts to jointly leverage novel computational hardware with AWS, along with new statistical techniques that scale to highly parallel computing environments, to unveil the chaotic environment around black holes, pushing our scientific knowledge to the horizon.
Around the horizon of a black hole, an edge of the universe, light is captured, spun into orbit by the black hole’s powerful gravitational pull. Lying within the orange donut in the famous first image of a black hole, this “photon ring” would be a prize...
My world was jolted by two shadow images: one, thrilling, the other terrifying. After a years-long effort—with 200 other scientists—we made the first image of a black hole, its shadow of no return. Then I fell into debilitating pain. A deadly shadow...
1930, Madras, India. On a steamship to England, 19-year-old Indian physics prodigy, Subrahmanyan Chandrasekhar makes a paradigm-shattering discovery about the life cycle of stars: a discipline pioneered by Sir Arthur Eddington, the world’s most famous...
