Signatures of Mass-to-Light Deviation Scenarios in Abell 370

A Free-form reconstruction using Grale.

Related Publication:

Since lens reconstructions recover the mass distribution in the clusters of galaxies, it also creates the opportunity to study the dynamics of the cluster itself. Clusters are relatively young in cosmic evolution. They are formed later than other cosmic objects. Sometimes two clusters merge together, triggered by their gravitational pull. These merging clusters are very interesting because they provide us the scope to study the nature of interactions between their merging member galaxies and the dark matter they host. Specifically, different theoretical models of dark matter give different predictions for the distribution of dark matter in merging clusters (Massey et al. 2021).

We studied one such merging cluster - Abell 370 - and used Grale to trace out its mass (Ghosh et al. 2021). In this project We used the strong lensing data from the Beyond Ultra-deep Frontier Fields and Legacy Observations (BUFFALO) collaboration. This data was collected by the Hubble Space Telescope in 2018. We am a part of the lens reconstruction group of this collaboration.

In the reconstructed mass model of Abell 370 (see Figure 2), We identified three mass features where there is no visible counterpart found: (i) a \( \sim 35\) kpc offset between the northern BCG and the nearest mass peak, (ii) a \( \sim 100\) kpc mass concentration of roughly critical density \( \sim 250\) kpc east of the main cluster, representing a possible dark matter substructure within the cluster and (iii) a probable filament-like structure passing N-S through the cluster. We tested the hypothesis on the presence of the filament-like structure by creating a synthetic cluster which mocks the situation of a cluster with external mass. When comparing lens inversion methods, it is sometimes instructive to compare specific regions of reconstructed mass maps generated by various methods.

We found that, while the offset in the northern BCG is present in some form in most publicly available reconstructions spanning the range of modeling techniques: parametric, hybrid, and free-form. The substructure and the possibility of the filament are recovered by only about half of the reconstructions. These comparisons show that many reconstructions with distinct modeling philosophies - parametric vs. free-form - can lead to converging results regarding specific mass features, while at the same time, models using the same algorithm can draw contrasting conclusions, just by using slightly different model priors and data constraints. We further concluded that because free-form models enjoy the freedom from fixed parameter spaces of parametric models, it enables them to recover mass features that are elusive for some parametric models.