An Image-Plane Approach to Gravitational Lens Modeling of Interferometric Data

Nan Zhang 1, Sreevani Jarugula 2, Justin S. Spilker 3, Simon Birrer 4, Jared Cathey 5, Scott C. Chapman 6,7,8, Veronica J. Dike 9, Anthony H. Gonzalez 5, Gilbert Holder 1, Kedar A. Phadke 9,10,11, Cassie Reuter 12, Joaquin D. Vieira 9,10,1, David Vizgan 9, Dazhi Zhou 8 · 2025

Read An Image-Plane Approach to Gravitational Lens Modeling of Interferometric Data by Nan Zhang 1, Sreevani Jarugula 2, Justin S. Spilker 3, Simon Birrer 4, Jared Cathey 5, Scott C. Chapman 6,7,8, Veronica J. Dike 9, Anthony H. Gonzalez 5, Gilbert Holder 1, Kedar A. Phadke 9,10,11, Cassie Reuter 12, Joaquin D. Vieira 9,10,1, David Vizgan 9, Dazhi Zhou 8 (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-2508-08393.txt.

Introduction

Nan Zhang and co-authors present a faster way to model strong gravitational lensing in radio telescope data, in a paper from 2025. Lensing acts as a cosmic telescope, magnifying distant galaxies. Standard modeling works in a raw data space that is costly to compute.

Their method works instead on the image plane, with care for the linked noise in the rough images such telescopes produce. Tests on simulated data with many noise draws gave accurate models. Applied to two real lensed sources, results matched earlier models built the costly way, for both smooth and pixel based reconstructions. The method is released in an open source software package for lens modeling.

Catalog record

CREATORNan Zhang 1, Sreevani Jarugula 2, Justin S. Spilker 3, Simon Birrer 4, Jared Cathey 5, Scott C. Chapman 6,7,8, Veronica J. Dike 9, Anthony H. Gonzalez 5, Gilbert Holder 1, Kedar A. Phadke 9,10,11, Cassie Reuter 12, Joaquin D. Vieira 9,10,1, David Vizgan 9, Dazhi Zhou 8
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-08393

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

An Image-Plane Approach to Gravitational Lens Modeling of Interferometric Data Nan Zhang 1, Sreevani Jarugula 2, Justin S. Spilker 3, Simon Birrer 4, Jared Cathey 5, Scott C. Chapman 6,7,8, Veronica J. Dike 9, Anthony H. Gonzalez 5, Gilbert Holder 1, Kedar A. Phadke 9,10,11, Cassie Reuter 12, Joaquin D. Vieira 9,10,1, David Vizgan 9, Dazhi Zhou 8 Abstract Strong gravitational lensing acts as a cosmic telescope, enabling the study of the high-redshift universe. Astronomical interferometers, such as the Atacama Large Millimeter/submillimeter Array (ALMA), have provided high-resolution images of strongly lensed sources at millimeter and submillimeter wavelengths. To model the mass and light distributions of lensing and source galaxies from strongly lensed images, strong lens modeling for interferometric observations is conventionally performed in the visibility space, which is computationally expensive. In this paper, we implement an image-plane lens modeling methodology for interferometric dirty images by accounting for noise correlations. We show that the image-plane likelihood function produces accurate model values when tested on simulated ALMA observations with an ensemble of noise realizations. We also apply our technique to ALMA observations of two sources selected from the South Pole Telescope survey, comparing our results with previous visibility-based models. Our model results are consistent with previous models for both parametric and pixelated source-plane reconstructions. We implement this methodology for interferometric lens modeling in the open-source software package lenstronomy . 1 Introduction Strong gravitational lensing is a powerful tool for studying cosmic history and the expansion of the universe. This phenomenon occurs when massive foreground galaxies or galaxy clusters distort the light from distant background objects. Lensing magnifies the flux of background galaxies, facilitating the study of high-redshift galaxies ( Stark et al. 2008 ; Jones et al. 2010 ; Treu 2010 ; Leethochawalit et al. 2016 ; Spingola et al. 2020 ) . Through observations and modeling of static and variable lensed objects, such as quasars, fundamental cosmological parameters, such as the Hubble constant, can be constrained ( Birrer et al. 2020 ; Millon et al. 2020 ) , and the distribution of dark matter can be mapped through gravitational imaging ( Vegetti et al. 2010 ; Vegetti et al. 2012 ; Hezaveh et al. 2016 ; Inoue et al. 2016 ; Birrer et al. 2017 ; Minor et al. 2021 ; Sengül et al. 2022 ; Şengül & Dvorkin 2022 ; Ballard et al. 2024 ) . The flux-ratio statistics of strong lensing also currently provide the most stringent constraints on primordial black holes and the free-streaming properties of dark matter ( Gilman et al. 2020 ; Hsueh et al. 2020 ; Dike et al. 2023 ; Keeley et al. 2024 ) . Submillimeter observations of high-redshift galaxies are important for studying the evolutionary history of early-type galaxies. Several gravitational lenses have been discovered at millimeter or submillimeter wavelengths with the Herschel -ATLAS survey ( Eales et al. 2010 ) , Atacama Cosmology Telescope ( Marsden et al. 2014 ) , Planck ( Planck Collaboration XXVIII 2013 ) , and the South Pole Telescope (SPT) ( Vieira et al. 2013 ) and have been followed up with Atacama Large Millimeter/submillimeter Array (ALMA) observations, which provide higher sensitivity and resolution ( Hezaveh et al. 2013 ; Vlahakis et al. 2015 ; Spilker et al. 2016 ; Berman et al. 2022 ) . Lens modeling techniques are needed to extract information from observed images of strongly lensed galaxies. The physical properties of both the lens and source galaxies are inferred statistically through lens modeling. This is achieved by assuming that the lensed images can be traced back to the flux from the source galaxies, while the light-path deflection angles are determined by the density distribution of the lens galaxy. To estimate lens models that best fit strongly lensed images and provide reasonable statistical descriptions, good noise models of data images are needed. However, while interferometers such as ALMA provide high resolution, the noise in interferometric images is highly correlated because signals are sampled in Fourier space, known as the uv -plane. The image-space noise correlation makes it more challenging to analyze strong lensing systems observed by interferometers. Several lens modeling packages have been developed to address this challenge. Examples include Spilker et al. 2016 , Hezaveh et al. 2016 , Powell et al. 2021 , and Nightingale et al. 2021 , which focus on lens modeling in visibility space. Recently, Powell et al. 2022 proposed a “fast \chi^{2} ” technique that enables the computation of \chi^{2} using only dirty images and a sub-grid of the uv -plane, providing an image-plane approach to lens modeling of interferometric images. In this paper, we re-derive the mathematical description of the fast \chi^{2} method and develop an independent implementation within the widely-used lenstronomy analysis package. The resulting likelihood function enables image-plane lens modeling of interferometric images. The lens modeling computation requires only three images of an interferometric observation as input: naturally weighted data images (also known as dirty images), the corresponding dirty beam response maps, which we refer to as the Point Spread Function (PSF), and the primary beam image of the data. We validate this image-plane lens modeling approach using simulated ALMA images, ensuring that the image-plane likelihood function yields the expected statistical properties. We also perform fittings on real ALMA images to verify that the results of image-plane-based fittings align with previous uv -fitting results obtained from the same data. We integrate our technique into the open-source Lenstronomy package ( Birrer & Amara 2018 ; Birrer et al. 2021 ) . Lenstronomy includes a large library of parametric lens and source profiles for parametric fitting. It also provides auxiliary tools for computing mappings between source planes and image planes, enabling pixelated source reconstructions. With the integrated capability to fit interferometric images, the multiband fitting feature of Lenstronomy also facilitates the joint analysis of strong lensing images from both interferometers like ALMA and pixel-based imaging telescopes, including the Hubble Space Telescope and the James Webb Space Telescope. This paper is organized as follows. In Section 2 , we introduce the image-plane natural-weighting likelihood function and present the method for pixelated source reconstruction using the natural-weighting covariance matrix. In Section 3 , we evaluate the statistical properties of the image-plane likelihood function by fitting a set of simulated ALMA data with an ensemble of random noise realizations. We also compare image-plane fitting with uv -plane fitting to verify the validity of our method. In Section 4 , we test our method with two well-studied ALMA sources. Parametric fitting is performed on SPT0346-52 and compared with previous analyses based on the uv -plane fitting package visilens ( Hezaveh et al. 2013 ; Spilker et al. 2016 ; Dong et al. 2019 ) . We test pixelated source reconstruction on SPT0311-58, which was previously analyzed using the uv -plane pixelated source fitting code RIPPLES ( Hezaveh et al. 2016 ; Spilker et al. 2022 ) . Finally, we summarize our work in Section 5 . 2 Methodology of Lens modeling Figure 1: Flowchart of the lens modeling pipeline for interferometric images. The pipeline consists of two components: natural-weighting parametric fitting (left-hand panel) and pixelated source reconstruction (right-hand panel). Our method takes as input a set of three data files

Read the complete text