MrMARTIAN: A Multi-resolution Mass Reconstruction Algorithm Combining Free-form and Analytic Components Astropy ( Astropy Collaboration et al. 2013 ; Astropy Collaboration et al. 2018 ; Astropy Collaboration et al. 2022 ) , Matplotlib ( Hunter 2007 ) , NumPy ( Harris et al. 2020 ) , PyTorch ( Paszke
Sangjun Cha, M. James Jee · 2025
Read MrMARTIAN: A Multi-resolution Mass Reconstruction Algorithm Combining Free-form and Analytic Components Astropy ( Astropy Collaboration et al. 2013 ; Astropy Collaboration et al. 2018 ; Astropy Collaboration et al. 2022 ) , Matplotlib ( Hunter 2007 ) , NumPy ( Harris et al. 2020 ) , PyTorch ( Paszke by Sangjun Cha, M. James Jee (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-13262.txt.
Introduction
Sangjun Cha and M. James Jee introduce a new algorithm for mapping the mass of galaxy clusters, in a 2025 paper. The method, called MrMARTIAN, reads how a cluster's gravity bends light from background galaxies, an effect known as strong lensing. It blends a free-form grid, which can take any shape, with fixed analytic nodes placed on known bright galaxies. Earlier free-form tools tended to smooth the inner mass profile too much.
A multi-resolution grid adds freedom only where lensing data are dense, which saves computing time. On simulated data, the method beat its predecessor at recovering both mass and magnification, and stayed stable when images were sparse. Applied to a real cluster using 412 multiple images, it reached the smallest scatter reported for that dataset.
Catalog record
| CREATOR | Sangjun Cha, M. James Jee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 44 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-13262
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
MrMARTIAN: A Multi-resolution Mass Reconstruction Algorithm Combining Free-form and Analytic Components Astropy ( Astropy Collaboration et al. 2013 ; Astropy Collaboration et al. 2018 ; Astropy Collaboration et al. 2022 ) , Matplotlib ( Hunter 2007 ) , NumPy ( Harris et al. 2020 ) , PyTorch ( Paszke et al. 2019 ) , SciPy ( Virtanen et al. 2020 ) Sangjun Cha, M. James Jee Abstract We present MrMARTIAN (Multi-resolution MAximum-entropy Reconstruction Technique Integrating Analytic Node), a new hybrid strong lensing (SL) modeling algorithm. By incorporating physically motivated analytic nodes into the free-form method MARS , MrMARTIAN enables stable and flexible mass reconstructions while mitigating oversmoothing in the inner mass profile. Its multi-resolution framework increases the degrees of freedom in regions with denser strong lensing constraints, thereby enhancing computational efficiency for a fixed number of free parameters. We evaluate the performance of MrMARTIAN using publicly available simulated SL data and find that it consistently outperforms MARS in recovering both mass and magnification. In particular, it delivers significantly more stable reconstructions when multiple images are sparsely distributed. Finally, we apply MrMARTIAN to the galaxy cluster MACS J0416.1–2403, incorporating two analytic nodes centered on the northeastern and southwestern BCGs. Our mass model, constrained by 412 multiple images, achieves an image-plane rms scatter of ∼ 0 . ′′ 11 \raise 0.81805pt\hbox{$\scriptstyle\sim$}0\farcs 11 , the smallest to date for this dataset. I Introduction As the most massive gravitationally bound structures in the universe, galaxy clusters have been the subject of extensive studies. Galaxy clusters enable the study of a wide range of topics, including the evolution of large-scale structures ( Eckert et al. 2015 ; Kuchner et al. 2020 ; HyeongHan et al. 2024 , e.g.,) , cosmology ( Acebron et al. 2017 ; Magaña et al. 2018 ; Caminha et al. 2022 , e.g.,) , and the properties of dark matter (DM) ( Markevitch et al. 2004 ; Robertson et al. 2019 ; Ragagnin et al. 2024 , e.g.,) . Among various techniques, strong gravitational lensing (SL) is one of the most powerful methods for studying galaxy clusters. Thanks to its unparalleled sensitivity to fine details in mass distributions, SL enables high-resolution and precise mass reconstructions around cluster cores and their inner profiles ( Zitrin et al. 2013 ; Kawamata et al. 2016 ; Sebesta et al. 2016 ; Bergamini et al. 2021 ; Cha & Jee 2023 , e.g.,) . Also, massive SL galaxy clusters serve as cosmic telescopes, magnifying faint background objects ( Welch et al. 2022 ; Hsiao et al. 2023 ; Furtak et al. 2024 , e.g.,) . In addition, SL can be used to infer the Hubble constant H_{0} by measuring time delays between multiple images of time-varying sources such as quasars or supernovae ( Kelly et al. 2015 ; Wong et al. 2020 ; Pascale et al. 2025 , e.g.,) . A number of SL modeling algorithms exist. Broadly, there are two types of SL modeling methods: parametric and free-form. Parametric methods utilize analytic profiles such as NFW ( Navarro et al. 1996 ) , dPIE ( Elíasdóttir et al. 2007 ) or PIEMD ( Kassiola & Kovner 1993 ) . Parametric methods reconstruct mass distributions by superposing these analytic profiles, with their positions constrained by the locations of cluster member galaxies. In contrast, free-form methods do not assume specific profiles for mass modeling. The mass is reconstructed on a grid without incorporating light distributions as priors. Both approaches have advantages and disadvantages. Parametric methods typically require fewer free parameters for modeling and can yield numerically stable results even with a small number of multiple images. However, the limited number of degrees of freedom restricts their ability to model complex and unexpected mass distributions. This limitation would be particularly concerning if the true mass distribution deviates from the light distribution. On the other hand, free-form methods offer excellent flexibility, but their high dimensionality often leads to degeneracies and unphysical features in the resulting mass models. Recently, Cha & Jee 2022 introduced a new grid-based free-form MAximum-entropy ReconStruction ( MARS ) algorithm, which employs maximum cross-entropy regularization. Thanks to this additional constraint, MARS can provide a quasi-unique solution and suppress spurious fluctuations ( Cha & Jee 2023 ; Cha et al. 2024 ; Cha et al. 2025 ) . However, despite these successes, MARS still has limitations. One such limitation is the smoothing of density peaks, which arises from both regularization and the finite grid size, especially in regions where SL constraints are sparse. Another limitation is the model’s dependence on priors in regions lacking SL constraints. Since free-form models are solely constrained by the multiple images, an asymetric or sparse distribution of SL constraints can introduce biases in the reconstructed mass distribution. ( Natarajan et al. 2024 ) . In this study, we propose a new hybrid Multi-resolution MAximum-entropy Reconstruction Technique Integrating Analytic Node ( MrMARTIAN ) to mitigate the aforementioned drawbacks of MARS . Although we choose a truncated NFW profile ( Baltz et al. 2009 ) as an analytic node in this study, MrMARTIAN can use a wide range of analytic profiles. By combining both free-form and parametric approaches, MrMARTIAN can leverage the advantages of both methods. In addition, we implement a multi-resolution framework in lens modeling. By assigning a higher resolution grid (i.e., more degrees of freedom) to regions with denser SL constraints, the total number of free parameters can be reduced without a significant loss of reconstruction quality. This enhancement enables MrMARTIAN to achieve improved computational efficiency. To evaluate the performance of the MrMARTIAN algorithm, we reconstruct lens models of simulated galaxy clusters. We use the publicly available synthetic clusters Ares and Hera from Meneghetti et al. 2017 , which are designed to mimic galaxy clusters observed by the Hubble Frontier Fields (HFF) program ( Coe et al. 2015 ; Lotz et al. 2017 ) . We also apply MrMARTIAN to the real galaxy cluster MACS J0416.1-2403 ( z=0.396 , hereafter MACSJ0416). As one of the targets of the HFF program, MACSJ0416 is well-known for its large number of multiple images ( Richard et al. 2021 ; Cha & Jee 2023 ; Bergamini et al. 2023 ) . Moreover, MACSJ0416 has also been observed by JWST as part of the Prime Extragalactic Areas for Reionization and Lensing Science ( Windhorst et al. 2023 , PEARLS;) and the CAnadian NIRISS Unbiased Cluster Survey ( Willott et al. 2022 , CANUCS;) programs. Based on the HFF and JWST data, lensing studies of MACSJ0416 have been published by both collaborations ( Diego et al. 2023 ; Diego et al. 2024 ; Rihtaršič et al. 2025 ) . Rihtaršič et al. 2025 report 415 multiple images across the MACSJ0416 field of view (FOV), which is the largest SL dataset to date. This extensive dataset makes MACSJ0416 an ideal testbed for evaluating whether MrMARTIAN is applicable in the current JWST era when hundreds of multiple images become available. This paper is organized as follows. In § II , we introduce the data used in the lensing analysis. We describe our new hybrid algorithm for mass reconstruction in § III . We present and discuss our results of MACSJ0416 in § IV and § V ,