Strong gravitational lensing as a cosmological probe: new ways of finding substructure and constraining the particle nature of dark matter

Studying the smallest self-bound dark matter structure in our Universe can yield important clues about the fundamental particle nature of dark matter, and galaxy-scale strong gravitational lensing provides a unique way to detect and characterize dark matter substructures at cosmological distances from the Milky Way. Research in this field can be broadly separated into works that aim to directly detect individual subhalos, carrying out a Bayesian likelihood analysis minimizing the residual between the smooth lens model and the perturbations produced by the subhalos, and works that aim to statistically constrain the substructure distribution, by looking at collective perturbations caused by an unresolved population of subhalos. We present recent advances in both of these approaches. Traditionally, pipelines that fall into the former approach can take weeks to characterize a lensing system, and often don’t find any compelling evidence for substructure. Setting our sights on the huge influx of strong lens images expected over the next few years, we present new work using convolutional neural networks to quickly analyze huge numbers of images and determine whether or not they are likely to contain substructure that can be detected through traditional methods, and therefore whether they warrant further analysis, with the key added benefit that no smooth lens modeling is required. With respect to the latter, we introduce the convergence power spectrum as a promising statistical observable that can be extracted from strong lens images and used to distinguish between different dark matter scenarios, showing how different properties of the dark matter get imprinted on different scales. Finally, we present PCATLens, a trans-dimensional Bayesian framework that refrains from requiring model subhalos to improve the goodness of fit above some detection threshold and instead uses all the information contained in photon count maps to constrain population characteristics such as the subhalo mass function and mass fraction, showing that fixed-dimensional inference can significantly mismodel the data.