Several long, thin, “filament” structures have been detected shooting out from nearby pulsars, misaligned with the pulsar velocity. They’re made of electrons and positrons recently manufactured by the pulsar and accelerated into the ambient interstellar medium with Lorentz factors up to 108, at almost the speed of light.
I searched for filaments in new Chandra data of high velocity pulsars. We found several candidates, but no confirmed sources. See the interactive display of my results. However, our null results place strong empirical limits on the acceleration mechanism of filament particles (Dinsmore & Romani 2024).
Several filaments have two thin structures, where one is an X-ray trail and one a filament. It was unknown which was which. I measured the proper motions of these pulsars to determine which was misaligned with the velocity. We measured new or improved existing velocity estimates for all filament pulsars but one, and discovered a new filament (Dinsmore & Romani 2026).
I have also tested and confirmed the theoretical belief that these filaments are aligned with the Galactic magnetic field; once through starlight polarization measurements of the Guitar filament (Dinsmore & Romani 2025), and once through IXPE polarimetry of the Lighthouse filament (Dinsmore et al., in prep).
These measurements poured the foundation for a new theoretical model of filament propagation through the ISM. I built a simulation based on this model which matches filament morphology and polarization measurements, and predicts that filaments leak a large but hitherto invisible fraction of their particles into space (Dinsmore & Romani, in prep.). These particles could contribute to the positron excess detected at Earth by AMS, Fermi and PAMELA, with could also be dark matter annihilation.
Advanced Polarimetry with IXPE
The Imaging X-ray Polarimetry Explorer (IXPE) is the first ever telescope to image X-ray polarization. It opens a new frontier for observing many high energy phenomena, including pulsars and their nebulae, as well as supernova remnants, black hole jets, X-ray binaries, and other high energy phenomena.
But X-ray polarization measurements are polluted by detector systematics that add distorting patterns onto the imaged polarizations (left). I developed a method to predict and subtract this polarization to reveal the true polarization of the source (Dinsmore & Romani, 2024). My publicly available tools have aided several other analyses: e.g. (Wong, Romani, & Dinsmore, 2023) and (Di Lalla et al. 2025).
I also measured accurate point-spread functions (PSF) of the satellite, which enable us to assign precise probabilities for each photon as to which source it came from. With these probabilities, I made a polarization analysis method which is twice as sensitive than the standard analysis too (Dinsmore & Romani, 2025). With this technique, Andrew Sullivan and I made the first X-ray polarization measurement of a spider pulsar's intra-binary shock (Sullivan, Dinsmore, & Romani, 2026). I also made the first polarization measurement of a confirmed X-ray filament, and a pulsar wind trail with an observation of the Lighthouse pulsar (Dinsmore et al., in prep).
Rapid, low-noise imaging with proto-Lightspeed
proto-Lightspeed is a new low noise, high speed imager for the Magellan telescopes, developed primarily by Kevin Burdge's group at MIT. It delivers 0.3 electron read noise on millisecond-length exposures, which enables a host of new science from solar system small-body searches to pulsar science in the optical.
With proto-Lightspeed data, I measured the optical light curve of PSR B0540-69. proto-Lightspeed delivered signal-to-noise comparable to previous results, but in 1/100 of the exposure time (Layden et al., 2026).
Probing Asteroids with Gravity
In 2029, an asteroid named Apophis will fly closer to Earth than any yet-measured asteroid. Earth’s tidal tug will change its spin axis in a way that will reveal the composition deep within the asteroid where light cannot reach.
Our work determined the observational precision needed to extract the asteroid’s properties from the changes in its spin axis, and demonstrated a few methods to translate those observations into an asteroid map (Dinsmore & de Wit, 2022).
Dark Matter Physics in the Galactic Center
The Fermi Gamma-ray Space Telescope detected an excess of high energy emission from the center of our Galaxy, which might have been a signature of dark matter colliding with itself and annihilating into into light.
But our work joins a growing set of literature which suggests that the excess could instead come from many dim, undiscovered pulsars. We show that the pulsar physics we know is consistent with this pulsar explanation, and that improved data or data analysis techniques could reveal some—even many—of these hidden pulsars (Dinsmore & Slatyer, 2021).
