Alex Gagliano

I’m a Postdoctoral Research Fellow with the NSF-Funded Institute For AI and Fundamental Interactions (IAIFI), where I use machine learning to study the astrophysical systems that the Universe lets us watch evolve in real time (supernovae, kilonovae, gamma-ray bursts, etc). I'm particularly interested in scalable techniques to infer a supernova's progenitor properties using synoptic surveys such as the Vera C. Rubin Observatory, and translating noisy observations into robust and interpretable strategies for follow-up studies of peculiar transients. You can find me either at the Harvard Center for Astrophysics or the Lab for Nuclear Science at MIT.

I earned my Ph.D. in Astronomy at the University of Illinois at Urbana-Champaign, and spent some time as a Pre-Doctoral Fellow at the Flatiron Institute's Center for Computational Astrophysics in Spring 2022.

Complementing my research goals, I’m also passionate about building community and creating welcoming spaces within astrophysics. I founded astro[sound]bites, the audio counterpart to the popular blog astrobites, to highlight the work of early-career scientists. During my Ph.D., I led Astronomy Cultural Journal Clubs at UIUC, and I've spearheaded outreach projects in the US, Gabon, and India. In my free time, you can find me practicing cello, biking, running (I ran my first marathon in 2021!), and reading. You can download my CV here.

Research

Rapid Inference from Survey Streams

Rubin, Roman, and JWST will revolutionize our understanding of the time-domain sky, but only if we have the architecture in place to investigate the weird, unexpected phenomena that they will discover. I develop fast, scalable methods to find and prioritize events within massive survey streams, and am an active member of the ANTARES alert broker.

Data stream entering computer, where the streams are categorized and directed to dedicated follow-up observatories.

Transient-Host Galaxy Studies

The local environment where transients occur can provide tantalizing evidence for the nature of their progenitors. Within the Dark Energy Science Collaboration., I combine observations with state-of-the-art simulations to investigate these correlations for poorly-understood classes of events.

Galaxies pulled from a statistical sample, one small galaxy hosting one transient and one large hosting many.

Early-Time Supernova Signatures

Data obtained in the first few days of an explosion provide an in-depth view into the final moments of a star's life and the earliest moments of its death. As a member of the Young Supernova Experiment, I search for these young transients and model signatures of shock breakout, CSM interaction, and flash ionization.

Chemical Enrichment from High Redshift Supernovae

Multi-scale simulations are required to understand the atomic and molecular chemistry of the early Universe. At the Los Alamos National Laboratory, I built a chemical solver that runs inline with cosmological simulations and probed the abundances of formed CO, OH, and H2O.

An SN surrounding by atoms and molecules.

Publications

An updated list of my publications can be found here:
ADS Publications

Preprints & Conference Proceedings

  1. A Physics-Informed Variational Autoencoder for Rapid Galaxy Inference and Anomaly Detection, , Machine Learning and the Physical Sciences Workshop, NeurIPS 2023.

Recent and Upcoming Talks

  • Machine Learning and the Physical Sciences Workshop at NeurIPS (poster), December 2023.
  • Unveiling the dynamic universe: cosmic streams in the era of Rubin, December 2023.
  • University of Michigan Astronomy Department Colloquium, November 2023.
  • Astro AI Meeting Discussion, November 2023.
  • "Building Software to Study Cosmic Explosions", Astronomy on Tap Boston, October 2023.
  • Five Colleges Astronomy Department Colloquium, October 2023.
  • LSST Transient and Variable Sciences Colloquium, October 2023.
  • Rubin Project and Community Workshop, August 2023.
  • Transient and Variable Universe, June 2023.
  • Rubin Observatory LSST @ Europe4, October 2022.
  • Caltech Time-Domain Astronomy Center Seminar, September 2022.
  • UC Berkeley Astronomy Department, September 2022.
  • MIT Brown Bag Lunch, May 2022. Announcement.
  • Harvard CfA, April 2022.
  • Caltech Tea, Mar 2022. Recorded Talk.
  • Princeton Thunch Talk, Feb 2022. Slides.
  • DESC Time Domain Working Group, Feb 2022. Slides.
  • LSST DESC Meeting Research Byte, Feb 2022. Slides.
  • Exploring the Transient Universe with Roman, Feb 2022. Recorded Talk.
  • Tri-State Cosmology x Data Science, Jan 2022. Slides.

I maintain/contribute to multiple open source software projects.

Name Description
PHotometric Analysis Software Tools (PHAST) A package for real-time transient classification using transient photometry and host galaxy information.
Simulated Catalog of Optical Transients and Correlated Hosts (SCOTCH) A set of analysis tools for simulating transient-host galaxy correlations for common (SNe Ia/II/Ibc) and rare (KNe) classes of events, and a catalog of 5M simulated events.
SuperNova ANAlysis package (SNANA) A software pipeline for simulating transients and generating realistic host galaxy properties.
Galaxies Hosting Supernovae and other Transients (GHOST) A package for associating transients with their most likely host galaxies and estimating host photo-zs.
-->

Outreach & Academic Service

An in-person discussion of the Astronomy Cultural Journal Club.

Astro[sound]bites

I am a creator and host of the podcast astro[sound]bites, where I discuss recent astronomy research with other graduate students. We amplify the voices of early-career scientists through regular interviews and explore sound-based representations of astronomy data. To date, we've published over 72 episodes, received outreach grants from the ASP, AGU, and AAS, and have listeners in 56 countries. Find our website here.

Alex laughing while presenting at the 2017 Story of Space
Alex giving a talk at the Literary on machine learning in astronomy.

Sounds of the Spectrum

As part of the 2017 Story of Space, an exploration of the overlap between astronomy and art, I designed and implemented an interactive exhibit exploring the chemical composition of the universe through sound. The exhibit ran for a week in Goa, India and reached thousands of members of the public. Read more about it here and here.

Alex laughing while presenting at the 2017 Story of Space

Visualizations

    You are welcome to use these for your own talks (but please credit me)! Click a video to download it.

    1. Optical spectral sequence of a type Ia supernova. Movie inspired by The Supernova Cosmology Project and data interpolated from the 'SUGAR' SN Ia spectral templates at http://supernovae.in2p3.fr/sugar/#sec-3. 2. Mollweide sky map of supernovae classified (spectroscopically or photometrically) from 1800 to 2022. Right histogram shows breakdown by redshift.

Contact