How to Take a Picture of a Black Hole

A member of the team that captured the world's imagination this week by unveiling the first-ever image of a black hole explained the science and engineering behind the image to a packed audience on Caltech's campus Friday.

Katie Bouman, who will join Caltech's faculty as an assistant professor of computing and mathematical sciences in the Division of Engineering and Applied Science in June, is a member of the Event Horizon Telescope (EHT) team and worked on the computational imaging that helped to tease an image from the noisy data captured by EHT's telescopes.

"I really wanted to come here first after the announcement, as [Caltech is] my new home," she told the audience.

The image presented this week depicts the black hole at the center of the galaxy Messier 87, or M87, which is around 55 million light-years away. It is so far away that the shadow of the black hole— the dark signature of the event horizon, which is the point beyond which not even light can escape—is just 40 microarcseconds wide. Trying to see it is equivalent to trying to shoot a photo of an orange on the moon while standing on Earth, Bouman said.

Further complicating the task is the fact that the black hole is surrounded by a cloud of gas that is optically thick, meaning that the black hole shadow can only be imaged via narrow wavelengths of light capable of piercing the cloudy veil. Given the black hole's distance and the wavelength of light needed, taking an image of it with a single telescope would require one that is 13 million meters wide: basically, a telescope as big as the earth.

Since that is not possible, the EHT team instead created a virtual telescope of that size using eight radio observatories at six locations spread across the globe, linked together using the precise timing of atomic clocks. Gathering data with such precision from such disparate locations was an unprecedented feat, but also only the beginning of the challenge, Bouman said.

"At this point ... we can abstract away all of the astrophysics of the problem and really just think of it as a purely computational imaging problem," Bouman said. "We have these sparse noisy data and our challenge is to find the image that actually caused it."

Multiple classes of imaging algorithms were employed to try to ascertain exactly what had generated the data captured by the eight telescopes. Those algorithms were painstakingly tested by making sure they faithfully reproduced known images from data akin to the EHT observations of M87.

After much testing and processing of the M87 data, an image finally did take shape, clearly showing a bright ring surrounding a void of black. Beyond producing a first-of-its-kind image, the process also allowed the team to estimate the black hole's mass, which had been the subject of some debate. It had been estimated to be between 3 and 7 billion solar masses, but the image allowed the team to pin that down to around 6.5 billion solar masses.

This technique of imaging using a network of globally distributed telescopes is also being employed to capture an image of the black hole Sagittarius A*, which lies at the center of the Milky Way galaxy.

Bouman, who earned her doctorate from MIT in 2017 and is now a postdoc at the Harvard-Smithsonian Center for Astrophysics, was one of about 200 scientists and engineers from across the globe who worked on the EHT project. After the announcement, she quickly became one of the most visible members of the team when a photo of her excitement over seeing the first iteration the black hole image went viral. However, throughout her presentation, she emphasized the collaborative nature of the project.

"This was a huge team effort. I know right now there's a lot of stuff in the media going around that I single-handedly led this project. That is as far from the truth as possible. ... This is the effort of lots and lots of people for many years," she said.

Next, she said, the team will try to add additional telescopes to generate even better data to work from, and may attempt to generate a video of a black hole in order to see its evolution over time.