🔭 astrophysics

Improving Precision in Kinematic Weak Lensing with MIRoRS: Model-Independent Restoration of Reflection Symmetries

This paper introduces MIRoRS, a novel model-independent technique that combines kinematic and photometric data to restore reflection symmetries in lensed galaxies, significantly reducing shear measurement uncertainty to 0.028 through validation on mock data and application to Illustris TNG simulations with an outlier rejection method.

Original authors: Christopher Hopp, David Wittman

Published 2026-02-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Christopher Hopp, David Wittman

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: Measuring the Invisible

Imagine you are looking at a distant galaxy through a telescope. Between you and that galaxy, there is a massive cluster of invisible dark matter. This dark matter acts like a giant, wobbly lens made of glass. As the light from the distant galaxy travels through this "lens," it gets stretched and twisted.

In astronomy, this stretching is called shear. It's a crucial clue because it tells us exactly how much mass (including the invisible dark matter) is sitting in that spot.

The Problem: Usually, measuring this stretch is incredibly hard. Galaxies come in all different shapes and sizes. Some are round, some are oval, some are lopsided. It's like trying to guess how much a piece of rubber was stretched just by looking at a crumpled piece of paper. You don't know what the paper looked like before it was crumpled, so you can't be sure how much it was stretched. This uncertainty is called "shape noise."

The New Solution: MIRoRS

The authors of this paper developed a new tool called MIRoRS (Model-Independent Restoration of Reflection Symmetries). Instead of guessing the galaxy's original shape, they use a clever trick based on symmetry.

Think of a spinning merry-go-round (a galaxy). If you look at it from the side, the people on the left are moving toward you, and the people on the right are moving away. This creates a perfect, symmetrical pattern of motion.

The Twist: When the invisible dark matter lens stretches the galaxy, it breaks this perfect symmetry. The left side might look slightly different than the right side, and the "spin axis" might look tilted.
The Fix: The MIRoRS method acts like a digital "undo" button. It tries different amounts of stretching (shear) and rotation until it finds the exact setting that makes the galaxy's motion pattern look perfectly symmetrical again.
The Clue: The amount of "undo" required to make it symmetrical tells the scientists exactly how much the galaxy was stretched in the first place.

The Secret Sauce: Two Eyes Are Better Than One

Previous methods tried to fix the symmetry using only the motion of the stars (kinematics). The authors realized they could get a much better result by combining motion with light (photometry).

The Motion (Kinematics): Tells us how the galaxy is spinning.
The Light (Photometry): Tells us the shape of the galaxy's glow.

The Analogy: Imagine you are trying to straighten a crooked picture frame on a wall.

If you only look at the motion of the dust particles floating in the air around the frame, you might guess it's crooked.
But if you also look at the shape of the frame itself, you can see exactly how it's tilted.
By combining both clues, you can straighten the frame perfectly.

The paper shows that using both the "motion" and the "light" together reduces the error in their measurements significantly.

The "Garbage In, Garbage Out" Filter

Even with a great tool, sometimes the data is messy. Some galaxies have weird features like spiral arms, bars, or warps that break the symmetry in ways that aren't caused by the dark matter lens. If you try to fix these, your measurement will be wrong.

To solve this, the authors used a statistical test called Moran's I.

The Analogy: Imagine you are looking at a map of rainfall. If the rain is random, it's just scattered dots. But if you see a giant, coherent blob of rain in one spot, that's a pattern.
The authors looked at the "mistakes" (residuals) left over after they tried to fix the galaxy's symmetry. If the mistakes looked like random static noise, the galaxy was a good candidate. If the mistakes formed a weird, organized pattern (like a spiral or a bar), they knew the galaxy was too messy to trust.
They used this test to filter out the "bad" galaxies. By throwing out the messy ones, they made their final measurement even more precise.

The Results: A Major Leap Forward

The team tested their method on a massive computer simulation of the universe (Illustris TNG) containing 358 galaxies.

Old Method: Previous methods had an uncertainty (error margin) of about 0.08.
New Method (MIRoRS): By combining motion and light, they got the uncertainty down to 0.039.
New Method + Filter: By using the "garbage filter" (Moran's I) to remove the messy galaxies, they got the uncertainty down to 0.028.

Why This Matters

This is a huge deal. It means astronomers can now measure the mass of the universe's invisible scaffolding with much greater precision, one galaxy at a time. Instead of needing to average out thousands of galaxies to get a rough idea, they can now get a very sharp, clear picture of the dark matter around individual galaxies.

In short: They built a smarter way to "un-crumple" the universe's photos, using both the spin and the shape of galaxies to see the invisible mass holding them together.

1. Problem Statement

Weak gravitational lensing is a primary tool for mapping the mass distribution of the universe, but traditional methods suffer from "shape noise"—the intrinsic, random ellipticity and orientation of unlensed galaxies. To overcome this, statistical ensembles of thousands to millions of galaxies are typically required.

Kinematic Lensing (KL) offers a solution by utilizing the velocity fields of individual galaxies. An unlensed disk galaxy possesses specific symmetries: symmetry across its major axis and antisymmetry across its minor axis. Weak lensing shear ( $\gamma$ ) breaks these symmetries. While previous KL methods (e.g., Donet & Wittman 2023) have improved precision, they rely on parametric models of the velocity field. These models often fail when real galaxies deviate from idealized circularity or specific velocity profiles, leading to residual uncertainties (e.g., $\sigma_{\gamma_\times} \approx 0.08$ ) that are larger than the signal itself. Furthermore, fitting the cross-component of shear ( $\gamma_\times$ ) is particularly sensitive to assumptions about the intrinsic axis ratio of the galaxy.

2. Methodology: MIRoRS

The authors introduce MIRoRS (Model-Independent Restoration of Reflection Symmetries), a technique that fits $\gamma_\times$ without assuming a specific parametric form for the galaxy's velocity profile.

Core Concept:
The method operates on the principle that if the correct shear parameter is applied to reverse the lensing effect, the galaxy's velocity field will restore its intrinsic reflection symmetries.

Algorithmic Steps:

Data Transformation: Starting with observed velocity and photometric data in the detector frame, the algorithm applies a series of transformations to map the data back to the source plane:
- Translation: Moves the centroid to the origin.
- Inverse Shear: Applies $A^{-1}$ to remove the shear distortion.
- Rotation: Aligns the galaxy's kinematic axes with the coordinate axes using a position angle parameter ( $\phi$ ).
Symmetry Restoration: Two independent symmetry operations are performed on the transformed field:
- Major Axis Reflection: Reflects the field across the major axis.
- Minor Axis Inversion: Reflects across the minor axis and inverts the sign (to account for antisymmetry).
Model Construction: The two transformed fields are averaged to create a single non-parametric model. This model is then transformed back to the detector frame.
Likelihood Optimization: The model is compared to the original data to compute a residual and a log-likelihood. The optimization (using a dual annealing algorithm) varies parameters ( $\gamma_\times$ , position angle $\phi$ , centroid, central velocity) to maximize the likelihood.
Photometric Constraint: To break degeneracies and improve precision, the method incorporates photometric data. It calculates the difference between the observed photometric position angle and the kinematic position angle predicted by the shear transformation (derived from DiGiorgio et al. 2021). This term is added to the likelihood function.
Outlier Rejection: A Global Moran's I test is applied to the residuals. Spatial autocorrelation in residuals indicates unmodeled physical structures (e.g., warps, bars) that violate the symmetry assumptions. Halos with significant spatial correlation are filtered out.

3. Key Contributions

Model-Independence: Unlike previous KL approaches, MIRoRS does not assume a specific velocity profile (e.g., Tully-Fisher or arctangent profiles), making it robust against deviations in galaxy morphology.
Hybrid Data Integration: It uniquely combines kinematic (velocity field) and photometric (isophotal shape) data within a single likelihood framework, leveraging the different ways shear affects kinematic vs. photometric axes.
Statistical Outlier Rejection: The introduction of the Moran's I test as a standard filter for kinematic lensing allows for the objective identification and removal of galaxies with complex internal structures that bias shear measurements.
Validation on Realistic Simulations: The method was rigorously tested on idealized mock data and applied to a large dataset of 358 halos from the Illustris TNG simulations, allowing for direct comparison with previous parametric studies.

4. Results

The paper presents results from three distinct analyses:

A. Idealized Mock Data:

Performance: The combined kinematic+photometric model outperformed both kinematic-only and traditional parametric models across various shear values, inclinations, and noise levels.
Inclination Sensitivity: The combined model showed superior performance at mid-inclinations ( $40^\circ$ – $60^\circ$ ).
Position Angle Sensitivity: The kinematic-only model failed completely when the galaxy was rotated $45^\circ$ relative to the coordinate axes (as $\gamma_\times$ mimics $\gamma_+$ in that frame). The combined model remained robust because the photometric term constrained the fit.
Systematics: The authors identified a small bias when $\gamma_+$ (parallel shear) is non-zero, particularly at high inclinations, where it can cause a bimodal distribution in the fit. However, the uncertainty remained lower than parametric methods.

B. Illustris TNG Simulation Analysis:

Dataset: 358 halos at $z=0$ with a known true shear of $\gamma_\times = 0$ .
Uncertainty Reduction:
- Parametric Model (DW23): Overall uncertainty $\sigma \approx 0.078$ .
- MIRoRS (Combined): Overall uncertainty reduced to $\sigma = 0.039$ .
- Core Uncertainty: The narrow component of the distribution (representing well-behaved galaxies) improved to $\sigma_{core} = 0.016$ .
Bias: A slight negative bias ( $\approx -0.007$ ) was observed, attributed to the specific clockwise rotation of the simulated galaxies (mimicking warped disks), which is expected to average out in a real, randomly oriented ensemble.

C. Outlier Rejection (Moran's I):

Applying the Moran's I test ( $p \leq 0.001$ ) filtered out $\sim44\%$ of the sample (472 out of 1074 observations).
Final Precision: After filtering, the overall uncertainty dropped further to $\sigma = 0.028$ , with a core uncertainty of $\sigma_{core} = 0.014$ .
Validation: Visual inspection of rejected halos confirmed they contained substructure (bars, warps) that violated symmetry assumptions, while accepted halos showed clean residuals.

5. Significance

This work represents a substantial leap forward in single-source weak lensing. By moving away from parametric assumptions and integrating photometric constraints, MIRoRS reduces the uncertainty in measuring shear ( $\gamma_\times$ ) by a factor of roughly 2 to 3 compared to previous state-of-the-art methods on the same dataset.

The ability to achieve $\sigma_{\gamma_\times} \approx 0.028$ (and potentially lower with outlier removal) on a single line of sight suggests that kinematic lensing can eventually provide precise mass measurements for individual halos without requiring massive statistical ensembles. This opens new avenues for studying the mass distribution of specific galaxy clusters and the nature of dark matter on small scales. The methodology also provides a robust framework for handling "shape noise" in future high-resolution spectroscopic surveys (e.g., with the ELT or SKA).