🔭 astrophysics

Galaxy Power Spectrum at Two-Loop Order: Implications for Weak Lensing Surveys and New Physics

This paper presents a two-loop galaxy power spectrum calculation within effective field theory that achieves per-mille agreement with N-body simulations up to $k=0.85~h$ Mpc $^{-1}$ , significantly improving cosmological parameter constraints and offering a new probe for ultra-light axion dark matter in upcoming weak lensing surveys.

Original authors: Mikhail M. Ivanov

Published 2026-07-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Mikhail M. Ivanov

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, cosmic ocean. In the beginning, this ocean was smooth, but over billions of years, gravity has caused it to ripple, swirl, and form massive waves and whirlpools. These "waves" are clumps of matter, and the "stars" and "galaxies" are like buoys floating on top of them.

Astronomers want to understand how these waves formed to learn about the fundamental rules of the universe (like dark matter and dark energy). To do this, they use a mathematical tool called Cosmological Perturbation Theory. Think of this tool as a way to predict the shape of the waves.

However, there's a catch: the deeper you look into the ocean (the smaller the scales), the more chaotic the waves get. Simple math works for the big, gentle swells, but it breaks down when the water gets turbulent.

Here is what this paper does, broken down into simple concepts:

1. The Problem: The "One-Loop" Map is Too Simple

For a long time, scientists used a "one-loop" map to predict how galaxies are distributed.

The Analogy: Imagine trying to predict the weather. A "one-loop" model is like saying, "If it's raining today, it will probably rain tomorrow." It's a decent guess for the big picture, but it fails when you try to predict a specific thunderstorm or a sudden gust of wind.
The Limit: This simple model works well for large, smooth areas of the universe, but it starts to give wrong answers when scientists look at smaller, more crowded regions. It's like trying to navigate a busy city intersection using a map of a quiet country road.

2. The Solution: The "Two-Loop" Upgrade

The author of this paper, Mikhail Ivanov, has built a much more detailed map called the Two-Loop Galaxy Power Spectrum.

The Analogy: If the one-loop model is a country road map, the two-loop model is a satellite navigation system that accounts for traffic jams, construction zones, and sudden detours. It calculates the "turbulence" of the cosmic ocean much more accurately.
The Math: To build this, the author had to invent new mathematical "operators" (rules for how galaxies cluster) up to the fifth order. Think of this as adding five new layers of complexity to the recipe to account for every possible way the galaxies can interact.

3. The Challenge: Too Many Ingredients?

When you add more complexity to a recipe, you often need more ingredients. In this mathematical model, the author found that to make the "two-loop" map work, they needed 21 new "bias" parameters (adjustable knobs to tune the model).

The Analogy: It's like upgrading from a basic toaster to a high-end kitchen appliance. The new machine has 21 different dials and settings. At first, having so many dials seems scary because you don't know which way to turn them.
The Fix: The author proved that many of these dials are actually redundant (they do the same thing). After cleaning up the list, they found that 17 dials are the essential ones needed to make the model work perfectly.

4. The Test: Does the New Map Work?

The author tested this new, complex model against a massive computer simulation of the universe (called the "PT Challenge").

The Result: The new "two-loop" model matched the simulation data with extreme precision (within one part per thousand) up to very small scales.
The Comparison:
- The Linear Model (the simplest one) broke down and gave wrong answers very quickly.
- The One-Loop Model (the previous standard) worked okay for a while but started to drift off course.
- The Two-Loop Model stayed on track much longer, reaching scales where the others had already failed.

5. Why This Matters: Measuring the Universe's "Weight"

The main goal of these models is to measure $\sigma_8$ (sigma-8).

The Analogy: Imagine trying to weigh a cloud. If you use a simple scale (the linear model), your measurement is shaky and has a wide range of error. If you use a high-precision industrial scale (the two-loop model), you get a much tighter, more accurate weight.
The Gain: The author found that using this new two-loop model reduces the error in measuring the universe's "clumpiness" by three times compared to the simple linear model, and by 40% compared to the previous one-loop model.

6. Looking for "New Physics"

Because this new model can see deeper into the "turbulent" parts of the universe, it acts like a more powerful microscope.

The Analogy: If you are looking for a specific type of fish in the ocean, a simple net might miss it. But a high-tech sonar (the two-loop model) can detect the subtle ripples caused by that fish.
The Discovery: The author suggests this model could help detect Ultra-Light Axions (a hypothetical type of dark matter). These particles are so light they would change the shape of the cosmic waves in a very specific way. The new model is sensitive enough to spot these changes in a range of scales that previous models couldn't reach.

Summary

In short, this paper presents a super-charged mathematical tool for understanding how galaxies cluster. By adding more layers of detail (going to "two-loop" order), the author created a model that is far more accurate than previous versions. This allows scientists to look at smaller, more chaotic parts of the universe without losing their way, leading to much more precise measurements of the universe's properties and opening the door to discovering new types of dark matter.

Technical Summary: Galaxy Power Spectrum at Two-Loop Order

Problem Statement
Large-scale structure (LSS) surveys (e.g., DESI, Euclid, LSST, Roman) require precise analytic models of galaxy clustering to extract cosmological parameters and test new physics. While Effective Field Theory (EFT) of LSS has successfully modeled galaxy power spectra and bispectra at the one-loop order, extending these computations to the two-loop order is necessary to access smaller, quasi-linear scales ( $k \gtrsim 0.3 \, h\text{Mpc}^{-1}$ ) where non-linearities become significant. Previous one-loop models become biased at these scales, limiting the constraining power on parameters like the mass fluctuation amplitude $\sigma_8$ . Furthermore, existing two-loop calculations for the matter power spectrum had not been extended to the galaxy power spectrum with a consistent bias expansion and renormalization scheme compatible with standard analysis tools like CLASS-PT.

Methodology
The authors compute the real-space galaxy power spectrum ( $P_{gg}$ ) and galaxy-matter cross-spectrum ( $P_{gm}$ ) at the two-loop order within the framework of cosmological perturbation theory (EFT).

Bias Expansion and Operator Derivation:
- The galaxy overdensity $\delta_g$ is expanded up to the fifth order in the underlying density and velocity fields.
- The authors employ a hybrid approach: Eulerian "local evolution" (LE) operators for terms derived from non-linear fluid equations and Lagrangian "non-local evolution" (NLE) operators to capture time-non-local effects and displacement fields.
- They identify 29 distinct bias operators at the fifth order. Through systematic analysis, they determine that many of these operators produce redundant contributions to the two-loop power spectrum.
- The basis includes operators built from density potentials ( $\delta$ ), Galileon operators ( $G_2, G_3$ ), and higher-order Lagrangian displacement tensors.
Renormalization Scheme:
- The authors adopt the Bogoliubov-Parasiuk-Hepp-Zimmermann (BPHZ) renormalization scheme. This involves subtracting ultraviolet (UV) divergences directly in momentum space integrands.
- This scheme ensures that the two-loop bias parameters are consistent with the one-loop parameters used in standard codes (e.g., CLASS-PT), avoiding the need for complex renormalization group evolution between different cutoff scales.
- They derive new renormalization conditions for the third-order bias parameters ( $b_3, \gamma_2, \gamma_3, \gamma_{21}$ ) and establish that the two-loop power spectrum requires only 17 independent bias parameters per galaxy sample after removing redundancies.
Counterterms and Stochasticity:
- The authors derive higher-derivative counterterms up to $O(k^4)$ and stochastic contributions up to $O(k^4)$ in the stochastic expansion.
- They demonstrate that mixed deterministic-stochastic operators do not generate physical contributions to the power spectrum to all orders.
- The final model includes 17 bias parameters, 8 higher-derivative parameters, and 3 stochastic parameters for $P_{gg}$ , with additional parameters for $P_{gm}$ .
IR Resummation and Numerical Implementation:
- Infrared (IR) resummation is implemented using Time-Sliced Perturbation Theory (TSPT) to correctly model the damping of Baryon Acoustic Oscillations (BAO).
- The two-loop integrals (involving $P_{15}, P_{24}, P_{33,I}, P_{33,II}$ ) are computed numerically using the Cuba library with the VEGAS adaptive Monte Carlo algorithm.
- The results are validated against the PT Challenge N-body simulation at $z=0.61$ .

Key Results

Operator Counting: The two-loop galaxy power spectrum is governed by 17 non-redundant bias parameters, derived from a basis of 29 fifth-order operators.
Agreement with Simulations: The two-loop model achieves per-mille level agreement with PT Challenge simulation data for galaxy-galaxy and galaxy-matter power spectra up to $k_{\text{max}} = 0.85 \, h\text{Mpc}^{-1}$ at $z=0.61$ .
Cosmological Constraints:
- Linear Theory: Becomes biased at $k_{\text{max}} \approx 0.05 \, h\text{Mpc}^{-1}$ , yielding a $\sigma_8$ error of $\approx 2\%$ .
- One-Loop EFT: Becomes biased at $k_{\text{max}} \approx 0.29 \, h\text{Mpc}^{-1}$ , yielding a $\sigma_8$ error of $\approx 1\%$ .
- Two-Loop EFT: Remains unbiased up to $k_{\text{max}} = 0.85 \, h\text{Mpc}^{-1}$ . This results in a $\sigma_8$ measurement with an error bar of $\approx 0.63\%$ .
- The two-loop model provides a measurement of $\sigma_8$ that is three times more precise than the linear theory model and $\approx 40\%$ more precise than the one-loop model, even when varying all 35 relevant EFT parameters with conservative priors.

Significance and Claims
The paper claims that the two-loop EFT analysis offers significant gains for upcoming "2 $\times$ 2 pt" (galaxy clustering + galaxy-galaxy lensing) analyses from surveys like Euclid, LSST, and Roman. By extending the valid scale range to $k \sim 0.85 \, h\text{Mpc}^{-1}$ , the two-loop model breaks degeneracies that limit current analyses, leading to tighter constraints on cosmological parameters.

Furthermore, the authors highlight the model's utility for probing new physics. Specifically, the extended scale reach allows for the testing of scenarios that modify the matter power spectrum shape in the range $k \in (0.4 - 0.8) \, h\text{Mpc}^{-1}$ . As a concrete example, they note that this precision could probe ultra-light axion dark matter sub-components with masses $m_a \sim 10^{-24} \, \text{eV}$ , a regime currently only weakly constrained by weak lensing data. The authors emphasize that while their analysis uses a simplified Fisher matrix approach and specific simulation data, the results suggest that realistic analyses of projected survey data will benefit from these improved theoretical controls.