Research

Supernova progenitor stars

Localization of the progenitor star of supernova 2019yvr in NGC 4666
Localization of the progenitor star to supernova 2019yvr in NGC 4666. The progenitor was identified in Hubble Space Telescope imaging with the aid of follow-up imaging of 2019yvr from the Gemini-South adaptive optics system. From Kilpatrick et al. (2021).

I use pre-explosion imaging to identify supernova progenitor systems. The classic example is Sanduleak -69 202 for SN 1987A. I search HST, laser-guide-star adaptive optics, and JWST data for counterparts, including the unusually cool progenitor of the Type Ib SN 2019yvr (above). Only a few dozen progenitors are known. Expanding that sample involves a range of state-of-the-art analysis techniques, including novel space telescope pipelines, high-precision astrometry and distortion modeling, and developing new models for the spectral energy distributions of massive stars.

I build image-registration and comparison tools for pre-SN fields and maintain updated Keck/OSIRIS distortion solutions for archival data. I am on the science teams for new Keck and Gemini AO imagers, including SCALES. Circumstellar dust can hide or redden progenitors at optical wavelengths, and so I work with Spitzer, HST, ground-based IR, and JWST to probe dust emission and extinction around massive stars before explosion.

CIERA: Webb Telescope unveils doomed star hidden in dust
CNN: ‘Oddball supernova’ reveals star’s death throes before exploding
STScI: Astronomers find possible elusive star behind supernova
National Geographic: The mystery of the universe’s missing exploding stars

Gravitational wave astronomy

Hubble image of NGC 4993 before GW170817 and Swope discovery image of SSS17a
(Left) Hubble Space Telescope image from 4 months before the discovery of GW170817. (Right) The Swope discovery image of SSS17a, the optical counterpart to GW170817. SSS17a is marked with the red arrow.

I work on optical counterpart searches, multi-wavelength follow-up, and physical modeling of gravitational-wave sources. I contributed to identifying SSS17a, the optical counterpart to the binary neutron star merger GW170817, and I help plan coordinated follow-up to future LIGO/Virgo/KAGRA events.

Broadband data from radio to X-ray are needed to separate kilonova, afterglow, and host contributions and to test for r-process signatures in spectra and light-curve decline rates. That work requires rapid scheduling, good models, and wide wavelength coverage.

I am the PI of the Multi-Messenger Treasure TROVE collaboration, an effort with partners including the University of Arizona, UC San Diego, and Northwestern. TROVE (the Tool for Rapid Object Vetting and Examination) is building a public, open-source, API-enabled platform to vet transients in real time for gravitational-wave and other multi-messenger follow-up. See the TROVE website for the team, software, and publications.

Science: early-career researchers and Science’s Breakthrough of the Year

Fast radio bursts

Rapid optical imaging of the fast radio burst FRB 180916
Rapid optical imaging of the fast radio burst FRB 180916 from Kilpatrick et al. (2021).

Fast radio bursts are millisecond radio flashes. Some repeaters are localized to host galaxies, and a Galactic magnetar has produced similar bursts, but non-radio counterparts remain scarce. A prompt optical or X-ray flash in coincidence with the radio pulse would pin down the explosion site in space and time, constrain the radiating plasma, and test whether the same engine powers both the radio beam and a high-energy outflow. Even deep limits rule out bright synchrotron or thermal counterparts and tighten the allowed energetics and beaming geometry. I run optical and X-ray follow-up, including high-cadence imaging of the nearby periodic source FRB 180916 (above).

Supernova remnants

The Cassiopeia A supernova remnant in 12CO J=2-1
The Cassiopeia A supernova remnant in 12CO J=2-1 from Kilpatrick et al. (2014).

Galactic supernova remnants interact with ambient molecular gas. Those encounters regulate how explosions deposit energy and momentum into the cold interstellar medium, which drives galactic feedback. They also stir interstellar turbulence and multiphase structure in molecular clouds, and they highlight interfaces where blast waves and cloud shocks can operate as sites of cosmic-ray acceleration. Quantifying when and where remnants actually shock dense molecular gas is a basic constraint on all of these processes.

Using 12CO J=2-1 (231 GHz) data from the Heinrich Hertz Submillimeter Telescope, I searched for broad molecular lines, signatures of shocked gas, near Cas A (above) and in a systematic survey of 50 Galactic SNRs. Shocked CO is detected toward Cas A. That survey found broadened CO toward nineteen SNRs in total, including nine newly identified broad-line regions associated with SNR-molecular cloud interaction.