Using Big Data and Machine Learning to Tackle Fundamental Cosmological Questions

Junior Faculty Fellow Spotlight: Dr. Dillon Brout, Cosmologist. Astrophysicist. Data Enthusiast.

By Hariri Institute Staff

How do we understand dark matter, dark energy, or the history and fate of the Cosmos? These are questions that Hariri Institute Faculty Fellow Dillon Brout seeks to answer through his research. Brout, assistant professor of astronomy and physics, is a leader in the field of observational cosmology. He uses large datasets, state-of-the-art machine learning, statistical tools, and image analysis to answer questions surrounding the dynamics of the cosmos. Previously Brout was a NASA Einstein Fellow at the Center for Astrophysics | Harvard & Smithsonian.

Brout has played a significant role in cosmology projects such as the Pantheon+ and SH0ES , which involved collecting the last 30 years of supernova data and reanalyzing it. The paper which analyzes Pantheon+ was awarded the IOP Publishing Top Cited Paper award. The team combined this data with 1000+ Hubble space telescope orbits, revealing the current expansion rate of the universe. The significance of this finding is that it contradicts the Planck collaboration, from when the universe was 380,000 years old (as far back as researchers today can see). The contradiction is one of the most widely discussed topics in cosmology today and could indicate a gap in the current theory of the universe. It motivates a lot of the research done today to map the expansion of the universe and test Einstein’s theory on general relativity and the physics of the universe in the time period shortly after the big bang. 

Marc Chalufour for Arts x Sciences Magazine sat down with Brout to discuss how he became interested in this field, and how his great-granduncle in Belgium – “one of the most important physicists of all time,” – inspired him to pursue cosmology and led him to where he is today. Read an excerpt of this article below.

When Brout became interested in physics in high school, he learned a surprising fact from his parents. “You know, we have some distant relative who lives in Belgium and is apparently some important physicist,” they said. “I came to realize very quickly that he is one of the most important physicists of all time,” Brout says.

Turns out, his great-granduncle, Robert Brout, along with François Englert, published the first paper to theorize that a “God particle” gives everything in the universe mass. Their theory dominated particle physics for more than half a century. Then, in 2012, scientists identified such a particle, which by then was called the Higgs boson. In 2013, Englert and Peter Higgs were awarded the Nobel Prize in physics for their work in developing the theory; Robert Brout likely would have received the award with them, but he’d died a year earlier.

Dillon Brout has been researching his great-granduncle’s legacy and hopes to write a book. “He’s up there with Einstein because of his impact on so many different aspects of physics,” he says.

Brout only learned of his great-granduncle a few years before his death, but they exchanged frequent letters in that time. When Brout shared his interest in particle physics, his great-granduncle encouraged him to study cosmology instead. “In 2008, he was thinking about the birth of the universe,” Brout says. “That set me on the trajectory that brought me here today.”

In 1978, Robert Brout coauthored a paper about how the universe may have expanded in the initial moments of the big bang, a theory that is now called inflation. For Dillon Brout, working on precisely measuring the cosmos is a chance to build on the idea. “If we could connect the dots between what the universe is doing now and what caused that initial inflation—that would be a dream for me.”

The Hariri Institute asked Brout about his research focus, recent projects on collecting and reanalyzing supernova data, and future work as a Hariri Institute Junior Faculty Fellow:

Hariri Institute: Can you describe your research focus and its applications? 

Brout: My research group ties together cosmological distance and velocity measurements of the universe with the fundamental physical properties such as dark energy, dark matter, ordinary matter, and Einstein’s general relativity. Cosmology is a field that is very fortunate. We have incredibly fundamental unanswered questions but we also have some of the largest datasets in all of science coming to us in the next decade.

We have these huge problems, like the fact that the nature of dark energy and the accelerating universe and the nature of dark matter are not known yet they make up roughly 95% of the matter/energy of the universe. In addition we face the fact that our “end to end” test of the universe (which compares the universe at its earliest times to the universe as we see it today) does not pass; this is the so-called Hubble Constant Crisis. We are now looking farther and farther back into the history of the universe to see if dark energy could be evolving, or if in fact it is a “cosmological constant” as given to us by Einstein himself. If the latter, then we still have to try and explain why theoretical predictions for the size of the cosmological constant are off from the observed value by 120 orders of magnitudes. This is sometimes called “the largest discrepancy between theory and observations in all of science”!

Hariri Institute: What are the main goals or objectives of your research?

Brout is now leading a project to answer those questions: the Rubin Observatory Legacy Survey of Space and Time (LSST) Dark Energy Science Collaboration, located one mountaintop over in the Chilean Andes. The LSST will build upon the technology of the Dark Energy Survey (DES) that Brout previously led , but with upgraded hardware, a larger field of view, more sensitive electronics, and a 60 second alerts stream—“major improvements in every facet of building the telescope,” 

The DES identified 1,500 supernovae; the LSST could find 1 million. “This is revolutionary. We can ask different questions,” Brout says. “We can probe the universe in completely new ways. It’s a whole new game we’re going to be playing. First, however, we need to wrangle the firehose of data that is coming our way.”

Here is an article about my work on LSST https://www.space.com/rubin-observatory-legacy-survey-space-time-dark-energy-expansion

Hariri Institute: Has there been a recent development or finding that you find particularly exciting?

Brout: LSST is coming online next year. The LSST data volume will be 20TB per night (15PB in total). The current set of challenges are how to handle this immense data volume and produce alerts within 60 seconds and mature analysis products within a year of data taking.

Hariri Institute: What do you feel is most rewarding about your work – either as a professor or researcher?

Brout: On a daily basis working with extremely large datasets is fun, but the most rewarding part is the search for answers to the birth, makeup, and fate of the universe. 

Hariri Institute: How do you plan on using this fellowship opportunity?

Brout: This fellowship will allow me to host LSST workshops that include the AI community. As this data mining effort will rely on AI algorithms, statistical and computational improvements, and image processing, this fellowship provides a unique opportunity to fund gatherings of cosmologists and data scientists with the broader Hariri and CDS community. Our first workshop will bring in the LSST Deputy Technical Coordinator to visit and work with us this spring.

Learn more about Professor Brout’s work here.