Research

I work in the fields of galactic archaeology—the study of a galaxy’s ancient stellar populations—and near-field cosmology—the study of the high-redshift Universe through nearby galaxies. A galaxy’s long-lived stars preserve a fossil record of its chemical and dynamical (i.e., chemodynamical) evolution, encoding the composition of the galaxy at the time of their birth as well as the mergers and accretion events that assembled the galaxy over cosmic time. Interpreting this archaeological record provides valuable clues into the physics of galaxy formation and the nature of dark matter. It is an incredibly exciting time for galactic archaeology, with wide-field spectroscopic surveys like DESI collecting millions of stellar spectra throughout the Milky Way and neighboring galaxies, while JWST is extending our reach to resolved stars in galaxies that are fainter, more crowded, and more distant than ever before. Leveraging these facilities, I study systems ranging from the faintest ultra-faint dwarf galaxies to massive galaxies like Andromeda and Centaurus A, combining Bayesian statistics, machine learning, and galactic chemical evolution modeling to turn observations into constraints on galaxy formation.

Chemodynamics of Ultra-Faint Dwarf Galaxies

As the oldest, lowest mass, most chemically pristine, and most dark matter-dominated galaxies in the Local Group, dwarf galaxies represent unparalleled laboratories to investigate a wide variety of astrophysics, including cosmic reionization, the nature of dark matter, hierarchical galaxy assembly, and the formation of the first stars. They are the building blocks of more massive galaxies, their stellar halos record structure formation on the smallest scales, their low baryon fractions make them unique sites to study dark matter, and their shallow potential wells and simple star formation histories make them sensitive probes of stellar feedback. The faintest of these systems, ultra-faint dwarf galaxies (UFDs), provide a particularly valuable window into the low-mass extreme of galaxy formation in the early universe. However, the paucity of bright stars in UFDs has historically precluded robust statistical analyses of their chemical and dynamical properties.

Recent narrow-band Hubble Space Telescope imaging has dramatically increased the number of stars in UFDs with measured metallicities (Fu et al. 2022, 2023). As a co-investigator on these programs, I developed a Bayesian galactic chemical evolution framework to constrain the physical processes governing the rapid but inefficient star formation and chemical enrichment of these galaxies. Applying this framework to the ancient UFD Eridanus II, I placed one of the first observational constraints on the strength of supernova-driven galactic winds in UFDs, finding that strong outflows ejected the vast majority of the galaxy’s enriched gas (Sandford et al. 2024). With collaborators, I am extending this analysis to the full sample of UFDs, as well as to similar observations of faint M31 satellites and isolated dwarf galaxies (Fu et al. 2024a, 2024b), which will provide constraints on the star formation efficiencies, star formation timescales, and galactic outflows for a large homogeneous sample of low-mass galaxies and yield new insight into the impact of stellar feedback as a function of mass and environment.

Meanwhile, by combining new wide-field spectroscopy of the UFD Boötes I from the Southern Stellar Stream Spectroscopic Survey (S⁵) with nearly 15 years of archival observations, I assembled the largest and most spatially extended sample of stars for any Milky Way UFD, enabling the most comprehensive chemodynamical study of a single UFD to date (Sandford et al. 2026). This included the detection of a line-of-sight velocity gradient aligned with Boötes I’s orbit, a possible signature of ongoing tidal disruption, and dynamical modeling that weakly favors a centrally concentrated (“cuspy”) dark matter profile, in line with the predictions of Λ Cold Dark Matter (ΛCDM) cosmology—though follow-up analysis has since complicated the story (Nguyen et al. 2026). Its stellar chemistry, like that of Eridanus II, supports a picture in which UFDs experience rapid, inefficient star formation and strong galactic outflows.

Mapping the Outskirts of Galaxies with DESI

In ΛCDM cosmology, extended stellar features like stellar halos and tidal streams form around galaxies as a natural consequence of hierarchical structure formation. Spectroscopic measurements of stellar chemistry and kinematics in galaxy outskirts therefore provide a unique probe of dark matter and galaxy assembly and a powerful test of the ΛCDM paradigm. The diffuseness of these features has historically limited their spectroscopic study in galaxies outside of the Milky Way, but wide-field, highly multiplexed spectrographs like the Dark Energy Spectroscopic Instrument (DESI) are poised to revolutionize this landscape. With its 3.2° field of view and ~5000 robotic fibers, DESI can efficiently target sparse stellar populations spread across many degrees on the sky. Already, the DESI Milky Way Survey (MWS) has acquired spectra of millions of stars throughout the Milky Way halo during its initial survey (DESI Run 1a; 2021–2025). In its extended run (DESI Run 1b; 2025–2029), DESI has begun a dedicated survey of Milky Way stellar streams, nearby dwarf galaxies, and Andromeda. As a co-chair of the MWS Working Group, I am heavily involved in these ongoing efforts, in particular:

• The Outskirts of Milky Way Satellite Galaxies
  I lead the DESI Milky Way Dwarf Galaxy Program, which will collect thousands of stellar spectra in five Milky Way satellite galaxies, more than doubling the number of member stars with precise velocities and chemistry in each galaxy. With these data, we will search for low-surface-brightness stellar halos and tidal features, constrain the galaxies’ dark matter distributions, and quantify the effects of stellar feedback on their evolution.
• A Panoramic View of Andromeda’s Halo
  I am also a key contributor to the DESI Andromeda Region Kinematic (DARK) survey, which is mapping the rich substructure of Andromeda’s complex stellar halo. With these data, we will produce the first panoramic chemodynamical map of a Milky Way-like galaxy’s stellar halo, allowing us to kinematically characterize known streams, detect new streams, reconstruct the mass and shape of its dark matter halo to unprecedented radii, and test cosmological predictions for galaxy assembly.

New Frontiers in Galactic Archaeology with JWST

The sensitivity and angular resolution of JWST enable us to extend the reach of galactic archaeological studies to resolved stellar populations in galaxies that are too distant, faint, and/or crowded for even the most powerful ground-based telescopes. As a result, JWST will play a pivotal role in following up the many distant, low-mass, low-surface-brightness galaxies, star clusters, and stellar halos and streams that next-generation imaging surveys (e.g., LSST, Roman, Euclid) will discover throughout the Local Volume (1–10 Mpc). I am involved in three ongoing JWST programs:

• A Closer Look at the Formation and Evolution of M31’s Inner Disk (GO-4735; PI Sandford)
  This program will obtain JWST spectroscopy and imaging in M31’s crowded inner disk, which will yield the first spatial profile of detailed stellar chemical abundance patterns in an external Milky Way-mass galaxy. With these data, we hope to make the first radial characterization of the [α/Fe] bimodality in M31’s disk, measure stellar ages, radial migration, and disk heating, and model the disk’s chemical evolution.
• Exploring the Accretion History and Halo of Giant Elliptical Cen A (GO-10246; PI Sandford)
  This program will obtain deep JWST spectroscopy and imaging in the halo of Centaurus A, which will provide the first resolved-star chemistry and kinematics in any massive elliptical galaxy, yielding transformative new understanding of the chemical enrichment, mass assembly, and formation history of ellipticals.
• Alpha Elements and the Baryon Cycle in Isolated Dwarf Galaxies (GO-3788; PI Weisz)
  This program, led by UC Berkeley Ph.D. student Katie Sharpe, is using JWST spectroscopy to measure stellar chemistry in isolated dwarf galaxies to reveal the physics of low-mass galaxy evolution in the absence of environmental effects (e.g., ram pressure, tidal stripping) that complicate studies of Milky Way satellite galaxies.

Stellar Spectroscopy

The absorption features imprinted on a star’s spectrum encode its physical structure, chemical composition, and line-of-sight velocity, which in turn provide a fossil record of its host galaxy’s chemical and dynamical evolution. As such, spectroscopy of resolved stars is an incredibly powerful tool of galactic archaeology. Supernovae (of various kinds), stellar winds, neutron star mergers, and gas in/outflows each leave a unique chemical signature in the abundance patterns of stars observed today. Accordingly, studying the detailed chemical abundances of large numbers of resolved stars provides valuable insight into everything from the formation and evolution of galaxies to the detailed nuclear and quantum physics that underlie stellar evolution and the production of heavy elements.

However, even for nearby galaxies, extragalactic distances require that stellar spectra be observed at low resolution, where heavily blended absorption features make detailed chemical abundances difficult to measure. Using the Fisher information of stellar spectra, I demonstrated that full-spectrum fitting techniques (as opposed to line-by-line analyses) applied to low-resolution spectra could enable robust chemical abundance measurements for galaxies throughout the Local Group (Sandford et al. 2020). The accompanying open-source package, Chem-I-Calc, has since informed multiple observing programs and the science cases for next-generation spectrographs like Keck/FOBOS (Bundy et al. 2020) and VLT/HRMOS (Magrini et al. 2023). I then tested these forecasts empirically by degrading high-resolution spectra of metal-poor giants in the globular cluster M15 (Sandford et al. 2023). The spectroscopic pipeline I developed for this work has since supported other analyses, including studies of massive stars in the extremely metal-poor galaxy Leo A (Gull et al. 2022) and the ongoing analysis of JWST spectroscopy. More recently, my students and I have been developing a machine learning technique known as contrastive learning that leverages the wealth of ground-based spectroscopic data with theoretical stellar models to enable more robust measurements from JWST spectra.

Software

To facilitate collaboration and reproducibility, I make all of my code and analysis publicly available on my GitHub. You can learn more about the software I have developed here. Don’t hesitate to reach out if you have any questions, find any mistakes, or would like to contribute to this code.

Students

Over the last three years, I have mentored 10 amazing students through four undergraduate capstone projects, seven summer projects, a master’s thesis, and a PhD thesis, resulting in several student research awards. My former mentees have gone on to pursue grad school as well as positions in industry and education. If you are a student at UCSC interested in any of the above research topics, please reach out!

Current

• Katie Sharpe, Ph.D. student, UC Berkeley (2023–present)
  • Alpha Elements and the Baryon Cycle in Isolated Dwarf Galaxies
• Matteo Moretti, B.Sc. student, University of Toronto (2026–present)
  • Unraveling the Formation History of the Andromeda Galaxy with JWST Spectroscopy
  • AST 425: Research Topics in Astronomy and Astrophysics, University of Toronto
  • 2026 Summer Undergraduate Research Program, University of Toronto (SURP Student Spotlight)
  • Honorable Mention for SURP Poster Award and Students’ Pick SURP Poster Award
• Aishani Chaudhuri, B.Sc. student, University of Toronto Scarborough (2025–present)
  • Hunting for Binary Stars in DESI Multi-epoch Spectroscopy
  • 2025 Summer Undergraduate Research Program, University of Toronto
  • Recipient of a University of Toronto Excellence Award and the Prof. Mercedes T. Richards Award for Excellence in Summer Undergraduate Research (News Article)

Former

• Nava Wolfish, B.Sc., University of Toronto (2025–2026) → Toronto Blue Jays R&D Intern
  • Reduction and Analysis of JWST Resolved Star Spectroscopy
  • AST 425: Research Topics in Astronomy and Astrophysics, University of Toronto
  • 2025 Summer Undergraduate Research Program, University of Toronto (SURP Student Spotlight)
• Rosayla Coulthard, B.Sc., University of Toronto (2024–2026; co-supervised with Mairead Heiger)
  • Validating Predicted Abundances for the Spectra of DESI Data Releases
  • Carbon-Enhanced Metal-Poor Star Properties within Milky Way Halo Satellites
  • 2024 Research Opportunity Program, University of Toronto
  • AST 430: Supervised Study in Astronomy & Astrophysics, University of Toronto
• Juan Carlos Aranda Muñoz, B.Sc., National University of Colombia (Summer 2025)
  • Forecasting Dwarf Galaxy Dark Matter Profile Constraints for DESI
  • 2025 Mitacs Research Internship, University of Toronto
• Julia Kim, B.Sc., University of Toronto (2024–2025) → Ph.D., MIT
  • Applying Contrastive Learning to JWST Stellar Spectroscopy
  • Part III Essay, University of Cambridge
  • 2024 Summer Undergraduate Research Program, University of Toronto
  • Recipient of a University of Toronto Excellence Award
• Xuanqi Wei, B.Sc., University of Toronto (Summer 2024) → M.S.E., University of Pennsylvania
  • Searching for Stars in the Outskirts of Ursa Minor with Gaussian Mixture Models
  • 2024 Research Opportunity Program, University of Toronto
• Colin McCurdy, B.Sc., University of Toronto (2023–2024) → Physics Learning Services Specialist, University of Toronto
  • Spectroscopic Data Reduction and Analysis of Triangulum II Keck/LRIS Observations
  • AST 425: Research Topics in Astronomy and Astrophysics, University of Toronto
• Shuhan Zheng, B.Sc., University of Toronto (2023–2024) → M.Eng., University of Toronto
  • Literature Review on the Chemical Abundances of Stars in Local Group Dwarf Galaxies
  • AST 424: Introduction to Astrophysical Research, University of Toronto

Earlier Work

As an undergraduate, I studied how galactic feedback shapes the metallicity gradients of disk galaxies using a semi-analytic model with Dr. Yu Lu at the Carnegie Observatories, and searched for gamma-ray signatures of dark matter annihilation in dwarf galaxies using Fermi-LAT data with Dr. Eric Charles through the DOE Science Undergraduate Laboratory Internship at KIPAC.

Publications

A full list of my publications is available on my Publications page.