Quantifying the power spectrum of small-scale structure in semi-analytic galaxies

Abstract

In the cold dark matter (CDM) picture of structure formation, galaxy mass distributions are predicted to have a considerable amount of structure on small scales. Strong gravitational lensing has proven to be a useful tool for studying this small-scale structure. Much of the attention has been given to detecting individual dark matter subhalos through lens modeling, but recent work has suggested that the full population of subhalos could be probed using a power spectrum analysis. In this paper we quantify the power spectrum of small-scale structure in simulated galaxies, with the goal of understanding theoretical predictions and setting the stage for using measurements of the power spectrum to test dark matter models. We use a sample of simulated galaxies generated from the Galacticus semi-analytic model to determine the power spectrum distribution first in the CDM paradigm and then in a warm dark matter scenario. We find that a measurement of the slope and amplitude of the power spectrum on galaxy strong lensing scales (k ∼ 1 kpc −1 ) could be used to distinguish between CDM and alternate dark matter models, especially if the most massive subhalos can be directly detected via gravitational imaging.

The lensing properties of subhaloes in massive elliptical galaxies in sterile neutrino cosmologies

“…However, the scatter from different projections is larger than the difference between the medians and values from the different models overlap, as in previous works (Brennan et al 2019). This suggests that it is virtually impossible to distinguish between the three models by analysing only a few lensing images.…”

“…In general, the convergence power spectrum P( ) of subhaloes is characterized by a number of features: first of all, the normalization depends on the overall convergence in subhaloes, which is by construction the main difference between the three models that we consider. As already shown in Brennan et al (2019), the normalization of the subhalo power spectrum is dominated by the highest mass subhaloes and thus removing them increases the relative difference between different models, as happens when excluding the 'luminous' subhaloes from our calculation. As a test, we computed the power spectra using maps that contain all the subhaloes and we also find that in this case the median P( ) is essentially identical across the three models.…”

Quantifying the Line-of-Sight Halo Contribution to the Dark Matter Convergence Power Spectrum from Strong Gravitational Lenses

“…The idea of using the subhalo convergence power spectrum [29,34,53,54] was developed as a statistical detection method to obtain population-level constraints without having to individually resolve clumps. Unlike in direct detection efforts where by construction the sensitivity is maximal for the most massive clump close to the lensed images, in a power spectrum approach the higher number of lower-mass halos can actually make the sensitivity peak for the mass range of 10 7 − 10 8 M for a CDM population of subhalos, and still maintain some sensitivity at lower masses [53].…”

“…In this paper, we focus on the LOS contribution to the convergence power spectrum. This observable has been analyzed extensively in the context of the subhalo contribution [29,34,53,54] (it has interchangeably been referred to as the substructure power spectrum) and identified as a powerful statistical method to constrain dark matter from strong gravitational lens images. It is a particularly valuable observable because it ties mass scales (what dark matter theories provide) to length scales (deflection angles on the lens plane).…”

Mining for Dark Matter Substructure: Inferring Subhalo Population Properties from Strong Lenses with Machine Learning

“…This complicated marginalization integral makes the likelihood for populationlevel parameters effectively intractable. Methods based on summary statistics (Birrer et al 2017a) and studying the amplitude of spatial fluctuations on different scales through a power spectrum decomposition (Hezaveh et al 2016a;Cyr-Racine et al 2016;Diaz Rivero et al 2018;Chatterjee & Koopmans 2018;Díaz Rivero et al 2018;Cyr-Racine et al 2019;Brennan et al 2019) have been proposed as ways to reduce the dimensionality of the problem and enable substructure inference in a tractable way. This class of methods is well-suited to studying dark matter substructure since they can be sensitive to the population properties of low-mass subhalos in strongly lensed galaxies which are directly correlated with the underlying dark matter particle physics.…”