MIT researchers have developed a novel algorithm that could help astronomers create the first image of a black hole. The algorithm, called CHIRP, would combine data collected from various radio telescopes around the globe.
The new project is under the auspices of the Event Horizon Telescope, an international collaboration which seeks to turn the planet into a large radio telescope dish. As Katie Bouman, who led the development of the new algorithm explains, radio wavelengths have a lot of advantages. However, because of their long wavelengths, they require large antenna dishes. So, for example, the largest single radio telescope dish has a diameter of 1,000 feet. However, the image it produced of the Moon would actually be blurrier than the image seen through an ordinary optical telescope (the one you might have in your backyard).
Taking an image of a black hole is similar to taking an image of a grapefruit on the Moon with a radio telescope. This would require a telescope with a 10,000 kilometer diameter, which is almost impossible considering the fact that the diameter of Earth is not even 13,000 kilometers.
In order to solve this issue, Event Horizon Telescope will coordinate measurements made by radio telescopes at different locations, and the large gaps in data will be filled by the new algorithm. The technique used is called interferometry. It combines the signals detected by pairs of telescopes so that the signals interfere with each other. CHIRP is precise, because if the measurements from three telescopes are multiplied, the delays caused by atmospheric noise cancel each other out. In other words, CHIRP requires data from three telescopes (and not two which is usually the case). It uses a model that’s more complicated than traditionally used algorithms, but has proven to be better than its predecessors at reproducing the original image from the measurements.
The Wonderful World of Algorithms 