Daniel Gilman
Astrophysics and Cosmology
I am currently a Brinson Prize Fellow in observational cosmology at the University of Chicago. Prior to this, I was a Schmidt AI in Science Fellow, and a postdoctoral scholar working with Professor Jo Bovy at the University of Toronto. I finished a PhD in Physics at UCLA in May 2020. I grew up in Yorktown, Virginia.
I am an astrophysicist working on cosmic probes of fundamental physics. My research connects theoretical predictions to observations from many different ground and space based observatories. I am particularly interested in developing techniques to infer properties of dark matter substructure, and how we can connect these observations to the particle nature of dark matter and the early Universe.
Strong gravitational lensing as a probe of dark matter substructure
Simulated gravitational lens observations across Space UV, Keck adaptive optics, ELT, HST F814W, Roman F106, and Euclid VIS.
The figure above this text shows an example of a quadruple image strong lens system as it would appear when observed by a variety of existing and potential ground and space based observatories. Strong gravitational lensing allows us to measure the properties of dark matter substructure across most of the observable Universe, even on mass scales where halos are too small to host stars and emit light. This capability enables direct inferences on the abundance and internal structure of dark subhalos and field halos along the line of sight.
The figure below this text illustrates the application of a modeling pipeline I developed on a real lens system, GRAL1131-4419. The figure shows a reconstruction of this lens system for one possible realization of dark matter subhalos and field halos. We are now analyzing populations of quad lenses recently observed by JWST to test key predictions of the cold dark matter paradigm, and to search for hints of new physics in small scale structure.
Image zooms of lens system 1131. Figure adapted from Gilman et al. (2026).
Galactic dynamics and snails
The Gaia mission has measured the positions and velocities of millions of stars near the Galactic disk. These data reveal a coherent spiral pattern in phase space that suggests some kind of dynamic perturbation took place at some point in the last billion years. I developed an open source package “darksnails” to model these perturbations, including both the known luminous satellites in our galaxy and the many dark subhalos predicted by cold dark matter. An example configuration of these perturber orbits is shown below.
Orbits of Milky Way satellite galaxies and dark matter substructure perturbers in the Galactic disk. Figure adapted from Gilman et al. (2025).
Self-interacting dark matter
Cold dark matter (CDM) predicts that dark matter is collisionless, but relaxing this assumption has some very interesting ramifications for cosmology, and in particular, dark matter substructure. The first figure below shows a couple of simulated projected mass maps for a strong gravitational lens system, with CDM on the left and self-interacting dark matter (SIDM) on the right. SIDM alters the internal structure of halos, which can lead to stronger gravitational lensing perturbations when core collapse occurs in halos. You’ll also see an animation of a self-interaction cross section for an attractive Yukawa potential. Small tweaks to the coupling strength, for a fixed mass ratio between the DM particle and a light force carrier, lead to order of magnitude changes in the cross section across the halo mass scales relevant for small-scale structure.
Convergence maps for CDM, SIDM with cores, and SIDM with cores and core collapse.
Animation of dark matter cross section evolution. Click to pause/play. Animation inspired by Gilman et al. (2023).
The primordial matter power spectrum
In addition to the particle nature of dark matter, the abundance and internal structure of dark matter halos depends on the initial conditions for structure formation. This means that an inference of halo properties through strong lensing, or any other technique, can be recast in terms of the primordial power spectrum. The primordial power spectrum depends on the physics of inflation and the properties of the early Universe on small scales (k > 10 Mpc−1), which are not currently accessible with measurements of the CMB or large-scale structure.
Constraints on the primordial matter power spectrum. Figure adapted from Gilman et al. (2022).
Fuzzy dark matter
Several dark matter models make ‘smoking gun’ predictions that make them more easily distinguishable from vanilla cold dark matter. Self-interacting dark matter, as described above (scroll up), is one such model, provided the cross section is large enough to drive halos into core collapse. Fuzzy dark matter, or ‘ultra-light DM’, is another example. In this class of theory, the de Broglie wavelength of a DM particle is of order 1 kpc, and wave interference effects described by the Schrödinger equation manifest themselves on galactic scales.
Below this text is a figure from Laroche, Gilman et al. that shows how the projected mass density around the Einstein radius of a strong lens system changes as the mass of the dark matter particle increases from 10−22 eV to 10−18.5 eV.
Gravitational lensing convergence maps across a range of fuzzy dark matter particle masses. Figure adapted from Laroche, Gilman, et al. (2022).
Time delay cosmography
Strong gravitational lensing enables direct measurements of cosmological distances, which depend on the expansion rate of the Universe and the Hubble constant. The TDCOSMO collaboration (see recent milestone paper) combines precise measurements of image arrival times with state of the art strong lens modeling to measure the Hubble constant, H0. In Gilman, Birrer, Treu (2020) we explored how dark matter substructure can affect the arrival time delays between lensed images, and how this could affect measurements of H0. Below, you’ll see figures adapted from Gilman et al. (2020) that show mock analogs of real lens systems generated with substructure in the lens model, including the perturbation to the arrival time surface (second from right) and the projected mass in dark substructure (far right).
WFI2033-4723 (top) and PG1115+080 (bottom): mock observed images, lens reconstructions, normalized residuals, residual time delay maps, and residual convergence maps. Figure adapted from Gilman et al. (2020).
A Python package for simulating populations of dark matter halos and subhalos for gravitational lensing simulations in a variety of dark matter models.
github.com/dangilman/pyHalo →A Python package for forward modeling vertical phase-space spirals in the Galactic disk.
github.com/dangilman/darkspirals →Compute exact solutions for the phase shifts and the differential scattering cross sections corresponding to a Yukawa potential.
github.com/dangilman/pykawa →