Constraining the dark matter particle mass through galaxy-galaxy strong gravitational lensing

We study several possible ways of constraining the dark matter particle mass through galaxy-galaxy strong gravitational lensing, where the key underlying idea is that a strong lens image could be perturbed by low-mass haloes that happen to be close to light-rays emitted from a background source galaxy. Under the Approximate Bayesian Computation framework, we develop a forward-modelling procedure aimed at directly placing constraints on the dark particle's thermal relic mass using a set of observations. We use mock images to demonstrate its ability to extract subhalo information from the power spectrum of image residuals and unbiasedly recover the input mass of the dark matter particles. We re-derive the number density of detectable intervening line-of-sight haloes relative to lens subhaloes in galaxy-galaxy strong lens observations. Unlike previous methods determining the detectability through fitting deflections or idealistic images, we regard a perturber as detectable only if adding it to a smooth model generates a statistically significant improvement in the reconstructed image. Our new results show that line-of-sight haloes are still important but, in most cases, no longer completely dominate detections over those of subhaloes. Finally, we study the effects on subhalo inference from the complexity of the lens through simulating mock lensed images with a lens from a high-resolution hydrodynamic simulation. We find that the commonly used lens model, a power-law profile, could wrongly infer the subhalo information if it cannot capture the non-elliptical features of the lens galaxy. We also consider a decomposed model for the lens mass, which separately models the stellar and dark matter mass. We find the decomposed model can successfully capture the complexity of the simulated galaxy and correctly infer the subhalo information.