🔭 astrophysics

A direct detection method of galaxy intrinsic ellipticity-gravitational shear correlation in non-linear regimes using self-calibration

This paper extends the self-calibration method to non-linear regimes by modifying its scaling relation to account for complex galaxy bias and intrinsic alignment models, demonstrating that it can effectively detect and suppress galaxy intrinsic ellipticity-gravitational shear contamination in Rubin LSST Year 1 survey data with high accuracy and robustness.

Original authors: Avijit Bera, Leonel Medina Varela, Vinu Sooriyaarachchi, Mustapha Ishak, Carter Williams, The LSST Dark Energy Science Collaboration

Published 2026-01-23

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Avijit Bera, Leonel Medina Varela, Vinu Sooriyaarachchi, Mustapha Ishak, Carter Williams, The LSST Dark Energy Science Collaboration

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: Cleaning Up the Cosmic Mirror

Imagine the universe as a giant, slightly warped mirror. When light from distant galaxies travels through space, the gravity of invisible dark matter bends that light, distorting the shapes of the galaxies we see. Astronomers call this "Cosmic Shear." By studying these tiny distortions, they can map out the dark matter and understand the universe's expansion.

However, there is a problem. The galaxies themselves are not perfect circles; they are slightly squashed or stretched (elliptical). Sometimes, the local gravity in a galaxy's neighborhood stretches it in a specific direction before the light even leaves. This is called Intrinsic Alignment (IA).

Think of it like this: You are trying to measure how much a funhouse mirror distorts a person's reflection (Cosmic Shear). But the person is already wearing a funny, stretched costume (Intrinsic Alignment) before they even step in front of the mirror. If you don't account for the costume, you'll think the mirror is more distorted than it actually is. This "costume" is the IG correlation (Intrinsic Ellipticity-Gravitational Shear), and it is the biggest source of noise in these measurements.

The Old Tool vs. The New Tool

For years, scientists have used a technique called Self-Calibration (SC) to remove this "costume" noise.

The Old SC Method: This worked well, but only for "linear" scales—think of it as looking at the universe from very far away, where things are smooth and simple. It assumed that the way galaxies cluster is a simple, straight-line relationship.
The Problem: The universe is messy up close. On smaller scales (non-linear regimes), galaxies cluster in complex, clumpy ways. The old method broke down here, like trying to use a ruler to measure a crumpled piece of paper.

What this paper does: The authors have upgraded the Self-Calibration tool to work in these "messy," non-linear zones. They created a new mathematical "scaling relation" (a conversion formula) that accounts for the complex, clumpy nature of galaxy clustering and the specific ways galaxies align.

How the New Method Works: The "Shadow" Trick

The authors propose a clever way to separate the "costume" (Intrinsic Alignment) from the "mirror distortion" (Cosmic Shear) using a concept similar to casting shadows.

The Setup: They look at pairs of galaxies. Some act as "lenses" (foreground) and some as "sources" (background).
The Shadow (gI): By looking at how the number of galaxies in a specific area correlates with their shapes, they can isolate the "Intrinsic Alignment" signal. It's like noticing that in a crowded room, people tend to stand in a certain formation that matches the shape of the room itself.
The Conversion (The Scaling Relation): Once they have measured this "Intrinsic" signal, they use their new, upgraded formula to calculate exactly how much of that signal is leaking into the "Cosmic Shear" measurement.
The Cleanup: They subtract this calculated noise from the total signal, leaving behind a much cleaner picture of the actual dark matter distribution.

The Results: A Cleaner View

The team tested this new method using data simulations for the upcoming LSST (Rubin Observatory) survey, which will scan the sky with incredible detail.

Accuracy: They found that for galaxies that are far apart in their redshift bins (different "layers" of the universe), the new method is accurate within 10%. For galaxies in the same layer, it's accurate within 20%.
Noise Reduction: Because the method is so accurate, it can suppress the "costume" noise by a factor of 10 for distant pairs and 5 for close pairs. This is a massive improvement, effectively silencing the background chatter so the main signal can be heard clearly.
Robustness: Even if the scientists aren't 100% sure about the exact parameters of how galaxies align or cluster (which is often the case), the method remains stable. It's like a sturdy bridge that doesn't wobble even if the wind picks up a little.

Why This Matters

The paper claims that by extending this method into the "non-linear" (messy, small-scale) regime, astronomers can now use a much larger portion of the data collected by future telescopes like LSST. Previously, they had to throw away the "messy" small-scale data because the old tools couldn't handle it. Now, they can keep that data, leading to much tighter constraints on the nature of Dark Energy and Dark Matter.

In short, the authors have built a better filter. It allows us to look through the cosmic mirror and see the true shape of the universe, even when the view is cluttered with the complex, clumpy reality of how galaxies actually behave.

Technical Summary: A Direct Detection Method of Galaxy Intrinsic Ellipticity-Gravitational Shear Correlation in Non-Linear Regimes Using Self-Calibration

Problem Statement
Intrinsic alignment (IA) of galaxies represents a primary systematic contaminant in Cosmic Shear (GG) measurements, which are critical for probing the dark sector of the universe. Specifically, the correlation between a galaxy's intrinsic ellipticity and the gravitational shear of a background source (IG correlation) acts as a dominant nuisance factor, particularly when cross-correlating redshift bins. Ignoring this contamination can bias the determination of cosmological parameters, such as the dark energy equation of state and the matter fluctuation amplitude, by significant margins (up to 50% and 30%, respectively, in some forecasts).

While the Self-Calibration (SC) method has been established as an effective technique to extract and remove IG contamination by utilizing correlations within the same lensing survey, its applicability has historically been restricted to linear scales and linear galaxy bias models. As upcoming Stage-IV surveys like the Rubin Legacy Survey of Space and Time (LSST) aim for high-precision measurements on smaller, non-linear scales, the existing linear approximations in SC become insufficient. The challenge lies in extending the SC method to the non-linear regime where galaxy bias and IA models exhibit complex behaviors, without introducing prohibitive systematic errors.

Methodology
The authors propose an extension of the Self-Calibration method designed to operate within the non-linear regime, specifically tailored for the proposed Year 1 (Y1) survey of the LSST. The methodology involves three core components:

Non-Linear Modeling: The study moves beyond the linear bias assumption. It incorporates a non-linear galaxy bias model expanded up to the second order in both the matter density field and the tidal field using Standard Perturbation Theory (SPT). For the IA physics, the authors utilize the Tidal Alignment and Tidal Torquing (TATT) model, which encapsulates both linear alignment (TA) and non-linear tidal torquing (TT) mechanisms.
Modified Scaling Relation: The core of the SC method relies on a scaling relation to compute the IG correlation ( $C_{IG}$ ) from the observable galaxy number density-intrinsic ellipticity correlation ( $C_{gI}$ ). In the linear regime, the ratio of the galaxy-intrinsic ellipticity power spectrum ( $P_{gI}$ ) to the matter-intrinsic ellipticity power spectrum ( $P_{mI}$ ) is simply the linear galaxy bias. In the non-linear regime, this ratio is no longer constant. The authors derive a new scaling relation where the ratio $R_i(\ell) = P_{gI}/P_{mI}$ is calculated explicitly using the TATT model and non-linear bias parameters. This ratio is computed up to the 1-loop order term using a modified branch of the FAST-PT code.
Signal Separation: The method first separates the galaxy number density-gravitational shear ($gG$) and galaxy number density-intrinsic ellipticity ($gI$) signals from the observed galaxy-galaxy lensing correlation ( $C_{g\gamma}$ ) within the same redshift bin. This is achieved using a suppression factor $Q_i(\ell)$ , which accounts for the lens-source selection geometry and photometric redshift (photo-z) uncertainties.

Key Contributions

Extension to Non-Linear Regimes: The paper provides a formal framework to apply the SC method beyond linear scales, accommodating non-linear galaxy bias and the TATT IA model.
Derivation of $R_i(\ell)$ : The authors derive the specific scaling factor $R_i(\ell)$ required for non-linear regimes, demonstrating that it is a combination of galaxy bias, IA parameters, and the matter power spectrum, rather than a simple linear bias.
Code Implementation: A branch of the FAST-PT code is made available to compute the $gI$ power spectra up to the 1-loop order, facilitating the calculation of the necessary power spectra for the new scaling relation.
LSST Y1 Application: The method is tested specifically against the redshift bin configurations proposed for the LSST Y1 survey, using the TATT model and standard non-linear bias parameters.

Results
The performance of the new SC method was evaluated across various redshift bin combinations (auto-correlating, adjacent, and non-adjacent):

Accuracy of Scaling Relation: The new scaling relation is found to be accurate within 10% for cross-correlating redshift bins (non-adjacent pairs) and within 20% for auto-correlating bins. For adjacent bins, the accuracy drops slightly but remains within the 10% threshold.
Suppression of Contamination: Based on these accuracies, the method can suppress IG contamination by a factor of 10 for cross-correlating bins and a factor of 5 for auto-correlating bins.
Robustness to Uncertainties: The study tested the method's robustness against moderate uncertainties in IA parameters (20% for amplitudes $A_1, A_2$ ; 50% for slopes $\eta_1, \eta_2$ ) and galaxy bias parameters (10% for $b_1$ ). Even with these uncertainties, the scaling relation remains accurate within 20% for all bin combinations, allowing for a suppression factor of 5.
Error Budget: The residual statistical errors ( $\Delta f^{(a)}_{ij}$ ) and systematic errors ( $\Delta f^{(b)}_{ij}$ ) resulting from the SC method are shown to be smaller than or comparable to the minimum fractional statistical error ( $e^{min}_{ij}$ ) inherent in the GG measurement due to cosmic variance and shape noise. This indicates that the SC method introduces negligible loss of cosmological information.
Photo-z and Other Systematics: The paper qualitatively assesses catastrophic photo-z errors, photo-z bias, and magnification bias. It concludes that for LSST-level photo-z accuracy (outlier rates $<0.1\%$ ), these effects do not invalidate the SC technique, though they require careful handling.

Significance
The paper claims that this work enables the LSST and similar future surveys to fully realize their potential by mitigating IA contamination on non-linear scales, where the statistical power of the survey is highest. By extending the SC method to the non-linear regime, the authors provide a pathway to extract a "pure" cosmic shear signal without relying solely on marginalization over IA parameters, which can degrade cosmological constraints. The authors emphasize that the residual errors after self-calibration are negligible compared to the fundamental measurement limits, ensuring that the method does not compromise the precision of dark energy and dark matter constraints. Furthermore, the availability of the FAST-PT implementation for $gI$ correlations supports broader 3x2 point analyses in the non-linear regime. The study remains modest, noting that while the method is robust, the accuracy is slightly lower for bins involving the lowest redshift (Bin 1) and for auto-correlations, reflecting the challenges of lensing signal strength and redshift evolution in those specific regimes.

The Big Picture: Cleaning Up the Cosmic Mirror

The Old Tool vs. The New Tool

How the New Method Works: The "Shadow" Trick

The Results: A Cleaner View

Why This Matters

More like this