Tonima Tasnim Ananna is bringing the heaviest black holes out of hiding. She has drawn the most complete picture yet of black holes across the universe — where they are, how they grow and how they affect their environments. And she did it with the help of artificial intelligence.
As far as astronomers can tell, nearly every galaxy stows a black hole at its center, weighing millions or billions of times the mass of the sun. Though these supermassive black holes can heat surrounding material until it glows brighter than all the galaxy’s stars combined, the light can be concealed by gas and dust. High-energy X-rays cut through that dusty veil. So for her Ph.D., completed in 2019, Ananna gathered surveys from four X-ray telescopes, more datasets than any previous study had used. Her goal was to create a model of how black holes grow and change across cosmic history. “It was supposed to be a short paper,” Ananna says. But models that explained one or a few of the datasets didn’t work for the full sample. “It stumped us for some time.”
To break the gridlock, she developed a neural network, a type of artificial intelligence, to find a description of the black hole population that explained what all the observatories saw. “She just went off and taught herself machine learning,” says Meg Urry, an astrophysicist at Yale University and Ananna’s Ph.D. adviser. “She doesn’t say, ‘Oh, I can’t do this.’ She just figures out a way to learn it and do it.” One early result suggests that there are many more active black holes out there than previously realized.
Galaxies live and die by their black holes. “When a black hole puts out energy into the galaxy, it can cause stars to form,” Ananna says. “Or it could blow gas away,” shutting down star formation and stunting the galaxy’s growth (SN: 4/25/20, p. 9). So understanding black holes is key to understanding how cosmic structures — from galaxy clusters down to planets and perhaps even life — came to be. Ananna’s model is built on data describing black holes at different cosmic distances. Because looking far in space is like looking back in time, the model shows how black holes change over time. It could also help figure out how efficiently black holes eat, which may help explain how some got so big so fast.
When Ananna was a 5-year-old in Dhaka, Bangladesh, her mother told her about the Pathfinder spacecraft landing on Mars. Her mother was a homemaker, she says, but was curious about science and encouraged Ananna’s curiosity, too. “That’s when I realized there were other worlds,” she says. “That’s when I wanted to study astronomy.” There were not a lot of opportunities to study space in Bangladesh, so she came to the United States for undergrad, attending Bryn Mawr College in Pennsylvania. She chose an all-women’s school not known for a lot of drinking to reassure her parents that she was not “going abroad to party.” Although Ananna intended to keep her head down and study, the social opportunities surprised her. “The women at Bryn Mawr were fiercely feminist, articulate, opinionated and independent,” she says. “It really helped me grow a lot.” Internships at NASA and CERN, the European particle physics laboratory near Geneva, and a year at Cambridge University, boosted her confidence. (She did go to some parties — “no alcohol for me, though.”)
Now, Ananna is giving back. She cofounded Wi-STEM (pronounced “wisdom”), a mentorship network for girls and young women interested in science. She and four other Bangladeshi scientists who studied in the United States mentor a group of 20 female high school and college students in Bangladesh, helping them find paths to pursue science.