Reaching for the Stars May 3, 2023 Profiles AstrophysicsCAPSStudents Share this page: Twitter Facebook LinkedIn Email Portrait of Alex Gagliano By NCSA News Staff Over our long and storied existence, every culture on Earth has observed and measured the great expanse of Earth’s cosmic ceiling. We watched the stars to know when to plant food, or where to go on long journeys. We’ve made up stories to entertain, horrify and enrapture the people around us as well as those to come. In that sense, Alex Gagliano is simply among the newest generation of stellar storytellers. He’s just fortunate enough to have some really cool tools and access to other smart storytellers with cool tools. “I’ve been very lucky to have found a happy home at the intersection of astrophysics and machine learning,” says Gagliano. “I try to devise machine learning tools to analyze supernovae more quickly.” Gagliano is an NSF graduate research fellow at the University of Illinois at Urbana-Champaign who will be finishing graduate school this spring. He also developed GHOST, which is a database of more than 16,000 exploding stars – known as supernovae – and the properties of their host galaxies. A website Gagliano created shows the properties of all galaxies that have hosted supernovae in the GHOST catalog. The properties of the galaxy and the properties of the supernova that took place are shown on the right, and helped quickly compare many events at once. Images are taken from the Pan-STARRS 3-pi survey. While Gagliano’s work is uniquely his, the experience he found with the Astrophysical Sciences Project Office (ASPO) at the NCSA is a testament to the benefits of being part of a scientific community. Why NCSA? Our stellar storytellers across the globe use a variety of methods to hear the message of the heavens, but what all these people have in common is a deep and rich community devoted to their art. Gagliano’s interest and abilities lean toward the highly technical. He trains machines to combine many pieces of evidence to discover supernovae more quickly. But, his community is just as important to him and was a vital part of his decision in graduate schools. “I wanted to find an academic home that sat right at that intersection between computation and astrophysics,” says Gagliano. “I was enthusiastic about NCSA for just how excited other people in the department were about being able to develop tools that were new in the realm of computer science. They were thinking about using those tools for really rigorous physical insight.” While Gagliano found the number of other departments that were truly interested in this work was less than he anticipated, he felt like he had found a home at NCSA. What’s more, NCSA and other organizations are pioneering this scientific intersection so well that the community behind computational astrophysics is rapidly growing. Counting the Explosions On the left: A composite image of the center of galaxy M100, one of the host galaxies in GHOST. A supernova exploded in the bright spot at inset in 2020 (named SN2020oi); Gagliano wrote a paper analyzing the explosion a year after he created the catalog.On the right: A new method Gagliano developed for GHOST to find the galaxy where a supernova took place. The method starts at the location of the supernova and follows the direction of increasing brightness, to finally arrive at the center of the right galaxy. In the most basic of terms, a star can go supernova because it has either gobbled up so many of its neighbors that it’s too big to go on or because it has burned through all of the fuel needed to maintain pressure and the core collapses and rebounds out. It then explodes in a spectacular light show that can be brighter than an entire galaxy. Obviously, the fact that these natural wonders even exist is reason enough to study them. On top of that, supernovae libraries like GHOST can help us understand the fundamental physics that governs the entire universe. Particularly, both nuclear and particle physics can benefit from the study of supernovae, and they’re critical tools for studying how the universe is expanding from dark energy. Gagliano points out that at the time of creation, GHOST was the largest catalog of its kind. Although it was knocked from the top spot within a year of debut, Gagliano certainly isn’t disappointed. If anything, this is just the natural progression that he wants to see. “The Rubin Observatory in its first year is going to discover half a million to 1 million supernovae,“ says Gagliano. “That will absolutely blow GHOST out of the water. But in preparation for those catalogs, we’ve got to start building our tools with something.” Connecting with Communities Gagliano has a deep appreciation for the community he found at NCSA, but he also tries hard to know and understand the global community he’s a part of. He spoke about his time in Goa, India, at an astronomy outreach festival called the Story of Space. Along with a music-major colleague in his undergraduate studies, Gagliano discovered that each element had a unique fingerprint when you shot light through it. By using this mapping of certain stellar elements, they were able to turn the data into sound. They then created an exhibit that allowed listeners to hear the sounds of the stars. One walkthrough allowed listeners to hear how the composition of a star changes over its lifespan as it produces more heavy elements. Another was a Midi keyboard that had each key make a different elemental sound, thereby allowing people to create their own star songs. “I have always loved traveling,” says Gagliano. “I think one of the benefits of astrophysics is that there are people doing it all around the world, seeing a slightly different night sky. It behooves you to connect with those different communities and better understand all the different ways that you could see and conceptualize the night sky.” A sky map from Gaia with the number of discovered explosions by 2022 overlaid. The majority of these supernovae (and the galaxies where they came from) are in the GHOST catalog. Of course, Gagliano’s journey is just beginning. After looking at the properties of supernovae that have already been discovered, Alex simulated what supernovae the next generation of observatories will find. He then used neural networks to predict the properties of these upcoming explosions as soon as they’re found, all part of his dissertation on “Building a Comprehensive Picture of Stellar Death.” After graduation, Gagliano will take his PhD to Cambridge where he has accepted a postdoctoral fellowship that’s jointly affiliated with Harvard, MIT, Northeastern and Tufts. As a member of the Institute for Artificial Intelligence and Fundamental Interactions, he’ll investigate how to combine physics constraints with new machine-learning models. Of course, he’ll never forget his time at U of I or the people he’s met. I’m just immensely grateful at how strong the community is at U of I. I feel like I’ve had the flexibility to do research that I’m incredibly excited about and be supported along the way. A lot of other departments don’t have that community aspect, but I’ve always felt rooted to the researchers here. Alex Gagliano, graduate research fellow at NCSA Gagliano isn’t leaving the U of I community. If anything, he’s simply exploring the vast global community of storytellers who have spent generations craning their necks up at the distant points of light in the darkness.