🔭 astrophysics

Lensing in the Blue III: Weak Lensing Shape Catalogs of 30 Merging Galaxy Clusters

This paper presents the weak gravitational lensing shape catalogs for 30 merging galaxy clusters derived from near diffraction-limited SuperBIT balloon-borne observations, utilizing the metacalibration algorithm and extensive diagnostics to achieve an unbiased reconstruction with a measured multiplicative shear bias of $(1.1 \pm 7.8)\%$ .

Original authors: Sayan Saha, Jacqueline E. McCleary, Spencer W. Everett, Maya Amit, Georgios N. Vassilakis, Emaad Paracha, Leo W. H. Fung, Steven J. Benton, William C. Jones, Gavin Leroy, Eric M. Huff, Richard Massey

Published 2026-03-20

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Sayan Saha, Jacqueline E. McCleary, Spencer W. Everett, Maya Amit, Georgios N. Vassilakis, Emaad Paracha, Leo W. H. Fung, Steven J. Benton, William C. Jones, Gavin Leroy, Eric M. Huff, Richard Massey, Thuy Vy T. Luu, Ajay S. Gill, Mohamed M. Shaaban, Philippe Voyer, Anthony M. Brown, Giulia Cerini, Paul Clark, Matthew Craigie, Christopher J. Damaren, Tim Eifler, David Harvey, Eric Habjan, John W. Hartley, Bradley Holder, Mathilde Jauzac, David Lagattuta, Jason S. -Y. Leung, Lun Li, Johanna M. Nagy, C. Barth Netterfield, Susan F. Redmond, Jason D. Rhodes, Andrew Robertson, L. Javier Romualdez, Jurgen Schmoll, Ellen Sirks, Sut Ieng Tam, André Z. Vitorelli, Alfredo Zenteno

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Balloon-Borne Eye in the Sky

Imagine you are trying to take a photo of a distant, faint firefly in a forest. If you take the photo from the ground, the air is thick with humidity, dust, and heat waves that make the firefly look blurry and wobbly. If you go to space, the air is gone, but the camera is incredibly expensive and heavy.

SuperBIT is the "Goldilocks" solution. It is a high-tech telescope launched on a giant scientific balloon that floats 20 miles up (in the stratosphere). At that height, it is above 98% of the Earth's atmosphere. It's like taking your camera to the top of a mountain where the air is thin and crisp, but without the cost of a rocket ship.

The goal of this paper is to show that SuperBIT can take incredibly sharp photos of galaxy clusters (huge groups of galaxies held together by gravity) and use them to map out Dark Matter.

The Mystery: Invisible Glue

Galaxy clusters are the biggest structures in the universe. They are held together by Dark Matter, an invisible substance that we can't see but know is there because it has gravity.

The Analogy: Imagine looking at a clear glass window. You can't see the glass itself, but if you look at the trees behind it, the glass bends the light, making the trees look slightly distorted.
The Science: This is called Weak Gravitational Lensing. The gravity of the Dark Matter in a cluster acts like that glass window, bending the light from galaxies far behind it. By measuring exactly how those background galaxies are stretched and distorted, scientists can map where the Dark Matter is hiding.

The Challenge: The "Jello" Problem

Measuring this distortion is incredibly hard.

The Signal is Tiny: The distortion is only about 1% (like stretching a rubber band just a tiny bit).
The Noise is Huge: Galaxies aren't perfect circles; they are naturally lumpy and irregular. It's like trying to measure the wind's effect on a leaf by looking at a pile of leaves that are already crumpled and twisted.
The Camera Blur: Even a perfect telescope blurs images slightly (this is called the Point Spread Function or PSF). If you don't know exactly how your camera blurs the image, you can't tell if a galaxy is naturally lumpy or just blurry.

The Solution: A Three-Step Recipe

The authors of this paper built a sophisticated "recipe" to clean up the data and measure the distortion accurately.

1. Cleaning the Lens (Data Pre-processing)

Before measuring anything, they had to clean the raw photos.

The Analogy: Imagine taking a photo with a dirty lens. You have to wipe off the dust (hot pixels), remove the glare from the sun (satellite glints), and stack 36 different photos of the same spot on top of each other to make the image brighter and clearer.
The Result: They created a "deep coadded image" where faint galaxies that were invisible in a single photo are now visible.

2. Mapping the Blur (PSF Modeling)

They needed to know exactly how the telescope blurs the light.

The Analogy: Think of the telescope's blur like a specific type of "fuzz" that changes depending on where you look in the photo. In the center, the fuzz is tight; in the corners, it might be stretched.
The Method: They used bright stars in the photo as "test dummies." Since stars are just points of light, any shape they have is purely due to the telescope's blur. They used a computer program (PSFEx) to map this blur across the entire image, creating a "blur map" for every single galaxy.

3. The Magic Trick (Metacalibration)

This is the most clever part. How do you measure a tiny stretch when the galaxies are already messy?

The Analogy: Imagine you have a potato chip that is naturally crinkled. You want to know how much it stretches when you pull it. You can't just pull it once; you might break it.
The Method: Instead of pulling the real chip, the computer creates a "virtual twin" of the galaxy. It artificially stretches this virtual twin slightly, measures how much the shape changed, and then shrinks it back. By doing this mathematically for thousands of galaxies, they can calculate exactly how sensitive their "ruler" is. This allows them to subtract the camera's blur and the galaxy's natural crinkles to find the tiny stretch caused by Dark Matter.

The Results: A New Map of the Invisible

The paper presents a catalog of shapes for 30 galaxy clusters.

The Stats: They measured the shapes of over 400,000 galaxies.
The Accuracy: They tested their method using fake images (simulations) that looked exactly like the real data. They found their method was accurate to within 1.1%. This is like measuring the width of a human hair from a mile away and being off by less than a millimeter.
The "Leakage" Check: They worried that the telescope's own blur might "leak" into their measurements, making it look like there is Dark Matter where there isn't. They proved that this leakage is negligible in the center of the images, though it gets a bit messy at the very edges (which they plan to fix in future work).

Why Does This Matter?

This isn't just about taking pretty pictures.

Testing Dark Matter: By comparing where the Dark Matter is (from the lensing) vs. where the hot gas is (from X-rays), they can test if Dark Matter particles bump into each other. If they do, it changes our understanding of the universe's fundamental laws.
Blue Light is Key: Most telescopes look at red light. SuperBIT looks at blue and ultraviolet light. Because the telescope is so high up, the blue light isn't scattered by the atmosphere. This allows them to see fainter, younger galaxies that other telescopes miss.
A Legacy: This data is now a public resource. Other scientists can use these "shape catalogs" to study everything from the dust between galaxies to the life cycles of stars.

Summary

In short, the SuperBIT team flew a balloon-borne telescope high above the atmosphere, took sharp blue-light photos of 30 massive galaxy clusters, and used a clever computer algorithm to "undo" the camera's blur. They successfully mapped the invisible Dark Matter holding these clusters together, proving that a balloon can do science as good as a satellite.

1. Problem Statement

Weak gravitational lensing is a premier tool for mapping dark matter distributions and constraining cosmological parameters. However, precise measurement of galaxy shapes (ellipticities) is hindered by observational systematics, primarily the Point Spread Function (PSF), pixelization, and noise.

The Challenge: Intrinsic galaxy ellipticities are an order of magnitude larger than the lensing shear signal (which is typically <1%). Distinguishing the shear signal from intrinsic noise and instrumental artifacts requires rigorous calibration.
The Context: This paper is part of the "Lensing in the Blue" series, focusing on data from the Super-pressure Balloon-borne Imaging Telescope (SuperBIT). Unlike ground-based telescopes (limited by atmospheric seeing) or space telescopes (high cost), SuperBIT operates in the stratosphere (~33 km), above 98% of the atmosphere. This offers near-diffraction-limited imaging in blue/UV wavelengths, which is crucial for studying high-redshift galaxies and minimizing shape noise, but introduces unique PSF characteristics (high-frequency structure due to lack of atmospheric smoothing) and stability challenges.
Goal: To produce a validated, unbiased weak-lensing shape catalog for 30 merging galaxy clusters observed during the 2023 SuperBIT flight, quantifying systematic biases and demonstrating the pipeline's readiness for cosmological analysis.

2. Methodology

The paper details a comprehensive end-to-end pipeline from raw data to calibrated shape catalogs.

A. Data Acquisition and Pre-processing

Instrument: SuperBIT uses a 0.5m Dall-Kirkham telescope with a Sony IMX-455 CMOS sensor (0.144"/pixel).
Observations: 45-night flight in 2023 targeting 30 merging clusters and 6 COSMOS calibration fields in three filters (F400W, F480W, F600W).
Pre-processing:
- Dark/Flat Correction: Creation of master darks using "haircut" thresholds to handle hot/cold pixels without biasing the mean.
- Astrometry: Solved using astrometry.net with specific index files (5202/5200) optimized for the wide field of view.
- Stacking: Clean exposures are stacked using SWarp with weight maps to handle hot pixels and dithering.
- Source Detection: Performed using SExtractor on the F480W (b-band) coadd. Spurious detections near bright stars (BCGs) and diffraction spikes are masked manually using DS9.

B. PSF Modeling

Star Selection: Stars are selected based on a size-magnitude diagram, excluding saturated stars and low-SNR objects.
Modeling: PSFEx is used to model the PSF spatially across the CCD.
Validation: The model is validated using a reserved 20% test set of stars. Residuals in size and ellipticity are computed, and $\rho$ -statistics (two-point correlation functions of PSF residuals) are calculated to ensure they do not contaminate the shear signal.

C. Shape Measurement: Metacalibration

Algorithm: The paper employs the Metacalibration algorithm (Huff & Mandelbaum 2017; Sheldon & Huff 2017) implemented via the ngmix package.
Mechanism: Instead of relying solely on simulations, Metacalibration measures the shear response ( $R$ ) of the shape estimator directly from the data by applying artificial shears ( $\pm \gamma$ ) to the deconvolved images.
Calibration Strategy:
- Gridded Response: A novel approach where the shear response is binned in a 2D grid of Signal-to-Noise Ratio (SNR) and galaxy-to-PSF size ratio ( $T/T_{PSF}$ ). This reduces shear bias compared to a global mean response.
- Selection Response: Accounts for selection biases introduced by the shear-dependent detection efficiency.
- Additive Bias: Corrected using terms derived from the PSF shear response ( $c_{PSF}$ ) and galaxy shear response ( $c_{\gamma}$ ).

D. Validation Framework

Unit Tests: A sequence of simulations (from idealized Gaussian PSFs to realistic SuperBIT PSFs and COSMOS galaxy populations) isolates sources of bias.
Fiducial Simulations: Realistic simulations injecting galaxies into blank exposures with real SuperBIT PSFs, noise, and sky backgrounds. These include source blending and are processed through the full pipeline.
Bias Metric: The multiplicative shear bias is quantified using the recovery factor $\zeta = \langle \hat{\gamma} \rangle / \gamma_{true}$ .

3. Key Contributions

First Large-Scale Blue-Wavelength Catalog: Delivers the first weak-lensing shape catalog for 30 merging clusters from a balloon-borne mission, leveraging the unique blue/UV capabilities of SuperBIT.
Gridded Calibration Technique: Introduces and validates a gridded shear-response calibration scheme that significantly reduces multiplicative bias in the presence of realistic observational conditions.
Comprehensive Systematics Control:
- Demonstrates that PSF modeling residuals (via $\rho$ -statistics) are well-controlled.
- Investigates PSF ellipticity leakage, finding that apparent correlations in real data are largely geometric artifacts of the lensing signal rather than true instrumental leakage in the central field.
Pipeline Robustness: Validates the ngmix Gaussian-mixture (em5) model for SuperBIT's diffraction-limited PSF, showing that despite smoothing high-frequency features, the residuals remain within acceptable limits for weak lensing.

4. Results

Source Density: Achieved a mean selected source density of $n_{sel} = 14.16 \text{ arcmin}^{-2}$ and an effective weighted density of $n_{eff} = 11.04 \text{ arcmin}^{-2}$ across all targets.
Shear Bias:
- Using a conventional global mean response calibration, the fiducial simulations showed a multiplicative bias of $\sim 6.3\%$ ( $\hat{\zeta} = 0.935$ ).
- Applying the gridded shear-response calibration reduced the bias to $1.1 \pm 7.8\%$ ( $\hat{\zeta} = 1.011$ ), consistent with an unbiased measurement.
PSF Modeling: The $\rho$ -statistics of PSF residuals fall within the "safe zone" (defined as 5% of the expected shear signal), confirming that PSF interpolation errors do not significantly contaminate the lensing signal.
Leakage Analysis: A correlation between galaxy shear ( $e_1$ ) and PSF ellipticity ( $e_{PSF,1}$ ) was detected ( $3.45\sigma$ ) in real data. However, fiducial simulations without lensing showed no significant leakage, suggesting the correlation in real data arises from the alignment of the NFW lensing shear pattern with the CCD-coordinate PSF pattern, rather than direct instrumental leakage in the central field.

5. Significance

Cosmological Utility: This catalog provides the foundational dataset for Paper IV (Convergence Maps), which will constrain the self-interaction cross-section of dark matter ( $\sigma/m$ ) by analyzing offsets between dark matter, stars, and intracluster gas in merging clusters.
Technological Validation: The work proves that balloon-borne telescopes can achieve space-quality weak lensing precision at a fraction of the cost, specifically in the blue/UV regime where ground-based observations are limited by atmospheric scattering.
Methodological Advancement: The successful application of gridded metacalibration to real, complex observational data sets a new standard for handling shear bias in high-resolution surveys, particularly those with non-Gaussian PSFs.
Legacy Value: The three-band photometric data (including near-UV) will support diverse astrophysical studies beyond cluster cosmology, including photometric redshift estimation, circumgalactic dust studies, and searches for OB stars and white dwarfs.

In summary, this paper establishes the SuperBIT mission as a viable and powerful platform for precision weak lensing, delivering a rigorously calibrated shape catalog that minimizes systematic errors through advanced calibration techniques and extensive simulation-based validation.