With the release of its first images yesterday, the Vera C. Rubin Observatory in Chile is one giant step closer to launching its Legacy Survey of Space and Time (LSST) - a decade-long astronomical investigation that will potentially transform humanity's understanding of the universe.
Several Northwestern University scientists have played integral roles in LSST since 2014, contributing scientific leadership and innovation that helped shape the endeavor.
"Northwestern's involvement has been critical in preparing for the immense discovery potential of the Rubin Observatory," said Northwestern astrophysicist Vicky Kalogera. "This is a transformative moment for astronomy, and we're proud to be at the forefront."
Jointly funded by the National Science Foundation (NSF) and the U.S. Department of Energy, the LSST will generate the widest, deepest images of the southern sky and create the first-ever comprehensive, multicolor movie of the night sky as it changes over time. The project will generate 20 terabytes of data per night, potentially holding answers to the universe's most elusive secrets, including the nature of dark matter and dark energy, the formation of galaxies, and the creation of elements by stars and their explosions.
Members of Northwestern's Center for Interdisciplinary Exploration and Research in Astrophysics (CIERA) and the NSF-Simons Institute for AI in the Sky (SkAI, pronounced "sky") will develop and refine AI tools capable of processing the unprecedented amounts of data as well as lead projects to understand the evolution of stars, search for the origins of heavy elements and uncover new cosmic explosions. Kalogera, who is the Daniel I. Linzer Distinguished Professor of Physics and Astronomy at Northwestern's Weinberg College of Arts and Sciences, directs CIERA and the SkAI Institute.
A universe of new data
While LSST promises transformational advances in astrophysics and cosmology, it also presents a massive challenge: processing an unprecedented volume of information. New AI tools will be crucial to meeting this demand. To help the effort, members of the SkAI Institute will develop and deliver innovations at the intersection of astronomy and AI, transforming the speed and reliability of simulations and optimizing the design of instruments and surveys.
"The SkAI Institute was established this past fall to develop new AI models in support of astronomical surveys," said Adam Miller, an assistant professor of physics and astronomy at Weinberg and director of the LSST-Discovery Alliance Data Science Fellowship Program. "For the next decade, LSST will be the flagship survey for U.S. astronomers. With its massive 3-billion-pixel camera, the Rubin Observatory will map the entire sky visible from the southern hemisphere every three days. This onslaught of data must be met with AI models. There is absolutely no way that any research team could actually look at the tens of billions of sources that Rubin will regularly monitor."
Northwestern researchers are teaming up across departments and disciplines to meet Rubin's various data challenges. Tjitske Starkenburg, a research assistant professor of physics and astronomy at Weinberg, and Emma Alexander, an assistant professor of computer science at the McCormick School of Engineering, will harness AI tools to study the stars and galaxies captured by LSST. Starkenburg and Alexander will build a new AI system to uncover ultra-faint wisps on the outskirts of nearby galaxies. Miller and Han Liu, the Orrington Hunt Professor of Computer Science at McCormick and professor of statistics at Weinberg, are leading a team using AI models - that have been trained on diverse data like stock market variations and weather patterns - to better predict how stars' brightness changes over time.
Miller also will collaborate with Aravindan Vijayaraghavan, an associate professor of computer science at McCormick, to develop new AI systems that will help schedule observations for the Rubin Observatory and other surveys.
"At any given moment, there are thousands of different locations where these telescopes could be observing," Miller said. "Determining the optimal location - and stringing together multiple consecutive optimal observations - is a complex problem that requires information about the weather, the atmosphere and past observations. New AI models will rapidly synthesize all this information to execute the best possible strategy."
In other work, Kalogera will collaborate with Aggelos Katsaggelos, deputy director of the SkAI Institute and the Joseph Cummings Professor of Electrical and Computer Engineering at McCormick, to develop new AI models to simulate the evolution of binary stars - condensing billions of years of evolution into mere hours. By studying these simulations, Kalogera, Katsaggelos and their team will explore how stars change over the course of their lives.
Explosive discoveries await
By consistently monitoring half the night sky, LSST will capture faint and fleeting phenomena, including supernovae, which can flare up quickly and suddenly vanish. With a constant eye toward space, astronomers expect to catch more of these transient objects.
"Right now, if you add up all the existing telescopes in the world, they catch about 40,000 new supernovae every year," Miller said. "The Rubin Observatory will discover 2,000 new explosions every night."
After pinpointing new supernovae, Charlie Kilpatrick, a research assistant professor at CIERA, will look at images of the sky taken before the explosions. By examining stars prior to their demise, Kilpatrick seeks to understand how stars lose mass during their death throes. For the first time, LSST will provide a detailed picture of how massive stars shed their outer layers immediately before exploding - a concept that theory has struggled to explain.
While Kilpatrick examines nearby explosions, Miller will co-lead a team with Allison Strom, an assistant professor of physics and astronomy at Weinberg, to study distant explosions that occur more than half-way across the universe. Because the early universe contained less iron than it does now, astronomers expect to find far more super luminous supernovae in these distant galaxies. The stars that give rise to these extraordinary explosions are formed with very little iron. Miller, Strom and their team will leverage these discoveries to understand how supernovae occur in high-mass galaxies with a low abundance of iron.
Wen-fai Fong, an associate professor of physics and astronomy at Weinberg, also will search for rare explosions in the distant universe. By studying these explosions, Fong and her team aim to understand where heavy elements are created and to uncover direct evidence for the origins of fast radio bursts. She also notes the possibility that the Rubin Observatory might uncover entirely new types of explosions that scientists have never seen before.
"The Rubin Observatory will instantly provide the deepest map of the southern sky in history," Fong said. "That map will help us find distant galaxies that host cosmic explosions. In the past, I've had to point a telescope specifically at a location of an interesting explosion, but now the Rubin Observatory will provide that information automatically. It will be a game changer particularly in finding distant galaxies that host such explosions."
