🔭 astrophysics

UNIONS-3500 Weak Lensing: I. A Galaxy Shape Catalogue in the Northern Sky

This paper presents the first weak gravitational lensing galaxy shape catalogue from the UNIONS survey, covering 3,500 square degrees of the Northern Sky with 62 million galaxies and detailing the comprehensive image processing, validation, and calibration procedures that establish a robust foundation for future cosmic shear cosmological analyses.

Original authors: F. Hervas-Peters, S. Guerrini, M. Kilbinger, L. Baumont, A. Guinot, C. Daley, C. Bonini, A. Wittje, C. Murray, L. W. K. Goh, A. Paradis, A. Tersenov, M. J. Hudson, L. Van Waerbeke, H. Hildebrandt, S.

Published 2026-05-14

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: F. Hervas-Peters, S. Guerrini, M. Kilbinger, L. Baumont, A. Guinot, C. Daley, C. Bonini, A. Wittje, C. Murray, L. W. K. Goh, A. Paradis, A. Tersenov, M. J. Hudson, L. Van Waerbeke, H. Hildebrandt, S. Fabbro, J. -C. Cuillandre, A. W. McConnachie

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

Imagine the universe as a giant, invisible web made of dark matter. We can't see this web directly, but we can see how it bends the light coming from distant galaxies, much like looking through a funhouse mirror or a wavy glass window. This bending effect is called weak gravitational lensing.

This paper, titled "UNIONS-3500 Weak Lensing," is the first major step in a five-part series by a team of astronomers. Their goal was to create a massive, high-quality "map" of galaxy shapes across the northern sky to study this invisible web.

Here is a breakdown of what they did, using simple analogies:

1. The Big Picture: Taking a Snapshot of the Northern Sky

Think of the universe as a giant canvas. For a long time, the "Northern Hemisphere" of this canvas was less explored than the South. The team used a powerful camera called MegaCam on the Canada-France-Hawai'i Telescope to take a massive picture of about 3,500 square degrees of the sky (roughly 10% of the entire sky).

They didn't just take a photo; they created a catalogue. Imagine a library card catalog, but instead of book titles, it lists 62 million galaxies. For every galaxy, they recorded its shape, size, and how much its light was distorted by the dark matter web.

2. The Challenge: The "Wavy Window" Problem

When you look through a dirty or wavy window, the view gets distorted. In astronomy, the "window" is the Earth's atmosphere and the telescope's own lenses. This distortion is called the Point Spread Function (PSF).

If you try to measure the shape of a galaxy while looking through a wavy window, you might think the galaxy is stretched when it's actually round. This is the biggest enemy of this research.

The Paper's Solution: The team spent a huge amount of effort building a mathematical model of this "wavy window." They used thousands of stars (which are perfect points of light) to measure exactly how the window distorts things. Then, they used a clever technique called Metacalibration to mathematically "undo" the distortion for every single galaxy. It's like taking a photo of a reflection in a rippling pond and using software to calculate exactly how the water moved to show you what the object looked like before the ripples.

3. The Process: Filtering the Noise

The team had to be very picky about which galaxies to include in their final list.

The Masking: Just like you wouldn't want to measure a galaxy if a bright star or a satellite streak was blocking it, they created digital "masks" to cover up bad spots, cosmic rays, and bright stars.
The Selection: They filtered out galaxies that were too faint, too small, or too blurry to measure accurately. They ended up with a "clean" list of galaxies that they could trust.

4. The Results: A Reliable Map

After all the cleaning and correcting, they produced a final catalogue with some impressive stats:

62 Million Galaxies: A massive dataset.
High Precision: The "noise" (random errors) in their measurements is very low, meaning the shapes are measured with high confidence.
Validation: They ran many tests to make sure their "wavy window" correction actually worked. They checked if the galaxies' shapes were still correlated with the telescope's position or bright stars (which would mean the correction failed). The tests showed that the remaining errors were small enough to be ignored for their main goal: studying the universe's structure.

5. What's Next?

This paper is just Part I of a five-part series. Think of it as laying the foundation and building the walls of a house.

Part I (This Paper): Building the catalogue and proving the measurements are accurate.
Part II: Checking for "ghost signals" (errors that look like real physics but aren't).
Parts III & IV: Using this clean data to actually calculate how much dark matter exists and how the universe is expanding.
Part V: Detailed simulations to double-check their math.

In Summary

The authors have successfully built a massive, high-precision "shape library" of galaxies in the northern sky. They have proven that they can correct for the distortions caused by the telescope and atmosphere well enough to use this data for serious cosmology. They haven't yet published the final answers about the universe's secrets (that comes in the next papers), but they have provided the reliable, high-quality data needed to find those answers.

Technical Summary: UNIONS-3500 Weak Lensing: I. A Galaxy Shape Catalogue in the Northern Sky

Problem and Context
Weak gravitational lensing (cosmic shear) is a primary probe for mapping the distribution of dark matter and constraining cosmological parameters such as $\Omega_m$ , $\sigma_8$ , and the dark energy equation of state $w$ . However, the precision of these measurements is limited by systematic errors, particularly those arising from the Point Spread Function (PSF) and image processing artifacts. While previous Stage-III surveys (e.g., KiDS, DES, HSC) have made significant strides, the Northern Hemisphere has remained relatively underserved by deep, wide-field photometric surveys. The Ultraviolet Near Infrared Optical Northern Survey (UNIONS) aims to fill this gap, but requires a robust, calibrated galaxy shape catalogue to enable cosmological analysis. This paper addresses the creation and validation of the first major weak lensing catalogue for UNIONS, covering approximately 3,500 square degrees of the Northern sky.

Methodology
The paper details the processing pipeline from calibrated MegaCam images (r-band) to the final shape catalogue, utilizing the ShapePipe framework.

Data and Pre-processing: The survey utilizes r-band data from the Canada-France-Hawai'i Telescope (CFHT) with a median seeing of 0.7 arcsec. Images are co-added into 0.5° × 0.5° tiles. Astrometric calibration is matched to Gaia EDR3, and photometric calibration is anchored to Pan-STARRS.
Source Detection and Masking: SExtractor is used for conservative and exhaustive source detection. Masking strategies include pre-processing chip defect flags, ShapePipe-generated masks for bright stars and artifacts (using the Guide Star Catalogue), and post-processing masks generated by maximask (a neural-network-based tool) to remove cosmic rays and fringing. A strict selection criterion of FLAGS = 0 is applied to remove potentially blended or deblended objects, though this is noted to introduce a selection bias that is later calibrated.
Shape Measurement and Calibration: The pipeline employs Metacalibration with a single-Gaussian model (ngmix) to measure galaxy ellipticities. This method estimates the shear response ( $R$ $R$ ) and selection response ( $R_{sel}$ $R_{se l}$ ) by artificially shearing the images.
- Multiplicative Bias ( $m$ ): Residual multiplicative biases are quantified using extensive image simulations (Paper V). These simulations revealed biases related to WCS offset propagation ( $\Delta m_{offset} \approx -0.02$ ) and SExtractor flag selection ( $\Delta m_{flag-sel} \approx -0.02$ ). Additionally, a blend bias ( $\Delta m_{blend} \approx -0.017$ ) was identified. The total multiplicative bias is calibrated to $m = -0.0572 \pm 0.0141$ (with a conservatively tripled error).
- Additive Bias ( $c$ ): Global additive biases are measured and subtracted.
PSF Modeling and Leakage Mitigation: A data-driven PSF model is constructed using stars in the magnitude range $18 \le r \le 22$ $18 \leq r \leq 22$ . The paper implements a two-step object-wise leakage correction:
1. Binning galaxies by Signal-to-Noise Ratio (SNR) and resolution ( $R$ ) to estimate a leakage coefficient $\alpha$ via linear regression.
2. Fitting a global trend function to $\alpha$ and subtracting the predicted leakage from observed ellipticities.
  Residual leakage is further quantified using $\rho$ -statistics (star-star correlations) and $\tau$ -statistics (star-galaxy correlations) to constrain the parameters $\alpha, \beta, \eta$ of the PSF systematic error model.

Key Contributions and Results

Catalogue Statistics: The final catalogue contains 62 million galaxies with an effective source density of $n_{eff} = 4.96$ arcmin $^{-2}$ over an effective area of 2,894 deg $^2$ (derived from a 3,648 deg $^2$ footprint). The shape noise (ellipticity dispersion) is $\sigma_\epsilon = 0.27$ .
Systematic Control:
- Additive Bias: The mean additive bias is measured as $c_1 = -1.6 \times 10^{-4}$ and $c_2 = 2.3 \times 10^{-4}$ , with uncertainties significantly smaller than Euclid requirements.
- PSF Leakage: The object-wise leakage correction significantly reduces the PSF systematic contribution to the two-point correlation function $\xi_+(\theta)$ . While significant leakage remains at scales beyond 100 arcminutes, the paper demonstrates that the residual contamination is sufficiently controlled for cosmological analysis when combined with scale cuts.
- Null Tests: The catalogue passes various null tests, including checks for correlations between ellipticity and image position, and galaxy-galaxy lensing around tile centers and Gaia stars, all consistent with zero.
Validation: The paper validates the pipeline through B-mode tests (detailed in a companion paper), image simulations, and cross-correlations with other survey data.

Significance and Claims
The authors position this work as the foundational step for the UNIONS collaboration's first major cosmological analysis. The paper claims that the UNIONS-3500 catalogue is not only a large, competitive dataset but, crucially, a reliable resource for cosmology. By demonstrating that systematic effects—specifically PSF leakage and multiplicative biases—are sufficiently characterized and mitigated, the authors assert that the catalogue is ready for the subsequent papers in the series, which will present cosmological constraints in real space (Paper III) and harmonic space (Paper IV).

The significance of this work lies in its provision of a high-quality weak lensing dataset for the Northern Hemisphere, offering substantial overlap with spectroscopic surveys like SDSS and DESI, and serving as a precursor to next-generation surveys like Euclid and LSST. The paper emphasizes that the robustness of the catalogue is established through rigorous validation, including dedicated image simulations and extensive systematic error modeling, rather than simply the volume of data.