The size and shape dependence of the SDSS galaxy bispectrum

This paper presents a measurement of the SDSS galaxy bispectrum across various triangle sizes and shapes, finding it well-described by a power law that is accurately modeled by $\Lambda$CDM N-body simulations with a bias of 1.2, while revealing that red galaxies exhibit significantly higher bispectrum amplitudes than blue galaxies due to their stronger clustering and non-linear evolutionary interactions.

The Gist

Imagine the universe as a giant, cosmic ocean. For a long time, scientists have been trying to understand the waves in this ocean. They've looked at the height of the waves (how many galaxies are in a specific spot) using a tool called the Power Spectrum. This is like measuring the average size of the waves to understand the ocean's energy.

But this new paper by Anindita Nandi and her team is asking a different, more complex question: How do the waves interact with each other?

Instead of just looking at single waves, they are looking at groups of three waves that form a triangle. In the world of galaxies, this is called the Bispectrum. Think of it like this: If you drop three pebbles in a pond, the ripples they create don't just exist in isolation; they crash into each other, creating complex patterns. The shape of the triangle formed by these three ripples tells us a lot about the "weather" of the universe.

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

1. The "Galaxy Map" and the "Triangle Game"

The researchers used data from the Sloan Digital Sky Survey (SDSS), which is like a massive, high-resolution map of a chunk of our cosmic neighborhood. They zoomed in on a cubic box of space containing about 16,000 galaxies.

They didn't just look at random galaxies; they looked at every possible group of three galaxies that could form a triangle. They analyzed 137 million triangles!

• The Size: How big is the triangle? (Are the galaxies close together or far apart?)
• The Shape: What does the triangle look like?
  • Equilateral: All sides equal (like a perfect slice of pizza).
  • Squeezed: Two sides are long, one is tiny (like a very thin, sharp needle).
  • Stretched: Two sides are equal, one is half the length (like a long, flat kite).

2. The "Recipe" for the Universe

The team discovered that the strength of these galaxy triangles follows a very specific mathematical recipe, which they call a Power Law.

Think of the bispectrum as the "loudness" of the galaxy clustering. They found that:

• The Shape Matters Most: The "loudness" changes drastically depending on the shape of the triangle.
  • Equilateral Triangles (The Calmest): These have the lowest loudness. It's like a gentle hum.
  • Linear Triangles (The Loudest): When the triangle gets stretched out until the galaxies are almost in a straight line, the "loudness" explodes. It's like a shout.
• The "Bias" Factor: They compared their real galaxy data to computer simulations (mock data). They found that the real galaxies act like they are "biased" tracers. Imagine if you were trying to map a forest, but you only counted the big, old oak trees and ignored the saplings. The oak trees (red galaxies) are more clustered together than the saplings (blue galaxies). Their simulation showed that if you assume galaxies are slightly more clustered than the underlying dark matter (a "bias" of 1.2), the computer model matches the real universe perfectly.

3. The Red vs. Blue Galaxy Showdown

The researchers split the galaxies into two teams: Red and Blue.

• Red Galaxies: These are the "old timers." They are like the retired residents of a quiet, crowded suburb. They are older, have stopped forming new stars, and tend to hang out in dense clusters (like galaxy clusters).
  • Result: They have a much louder bispectrum. Their triangles are "noisier" because they have been interacting, merging, and wrestling with gravity for billions of years. They are highly biased.
• Blue Galaxies: These are the "youngsters." They are like energetic teenagers in a sprawling, open park. They are still forming stars and are more evenly spread out.
  • Result: They have a quieter bispectrum. Their distribution is closer to the underlying "skeleton" of the universe (dark matter) and hasn't been as distorted by complex interactions.

4. Why Does This Matter?

You might ask, "Why do we care about the shape of galaxy triangles?"

• It's a Fingerprint of Gravity: The universe started as a smooth, boring soup (Gaussian). But gravity is a bully; it pulls things together, creating clumps, filaments, and sheets. This process makes the universe "non-Gaussian" (messy and complex). The bispectrum is the best tool we have to measure exactly how messy it has become.
• The "Linear" Clue: The fact that the signal is strongest when galaxies are in a straight line (linear triangles) is a huge clue. It suggests that galaxies are forming along giant cosmic "highways" or filaments, like beads on a string.
• Testing Our Models: By showing that a simple computer model (with just a little bit of "bias") can perfectly recreate the complex triangle patterns of the real universe, the authors prove that our understanding of how gravity builds the universe is solid.

The Takeaway

This paper is like a detective story where the detectives (the scientists) look at the footprints (galaxies) left in the cosmic mud. Instead of just counting the footprints, they look at the triangles formed by three footprints at a time.

They found that:

1. Shape is everything: The universe looks very different depending on how you arrange your three galaxies.
2. Old galaxies are louder: The older, red galaxies are more tightly packed and show stronger signs of gravitational "brawling."
3. Our models work: The computer simulations we use to predict the universe are accurate enough to match the real, messy, beautiful data we see in the sky.

In short, they mapped the "personality" of the universe's structure, proving that gravity has turned a smooth cosmic soup into a complex, web-like structure, and they did it by playing a giant game of "connect the dots" with 137 million triangles.

Technical Summary

Here is a detailed technical summary of the paper "The size and shape dependence of the SDSS galaxy bispectrum."

1. Problem Statement and Motivation

The large-scale structure (LSS) of the Universe originates from linear, Gaussian primordial density perturbations. However, gravitational instability drives these perturbations into a non-linear regime, rendering the LSS non-Gaussian. While the two-point correlation function (or power spectrum) fully characterizes Gaussian fields, it is insufficient for describing the non-Gaussian features of the late-time Universe.

The bispectrum (the Fourier transform of the three-point correlation function) is the lowest-order statistic sensitive to this non-Gaussianity. It depends on both the size and shape of the triangle formed by the wavevectors $(k_1, k_2, k_3)$. While previous studies have used the bispectrum primarily to constrain cosmological parameters (like $\Omega_m$) or bias parameters, this paper adopts a different perspective: it aims to quantify the bispectrum as a descriptive statistic to understand how non-Gaussianity varies with triangle configuration (size and shape) in the local Universe, without focusing on cosmological parameter estimation.

2. Methodology

Data Selection

• Observational Data: The authors utilized the SDSS Main Galaxy Sample (DR17). To ensure uniformity and avoid complex survey geometry issues, they extracted a volume-limited sub-sample within a cubic volume of $[296.75 \text{ Mpc}]^3$.
  • Selection Criteria: Galaxies with absolute magnitude $-22.31 \le M_r \le -21.57$ and redshift $0.05 \le z \le 0.15$.
  • Sample Size: 16,324 galaxies with a median redshift of $z=0.102$ and a mean number density of $0.63 \times 10^{-3} \text{ Mpc}^{-3}$.
  • Classification: Galaxies were separated into Red and Blue classes using Otsu's thresholding technique on the $(u-r)$ color vs. stellar mass plane.
• Simulations: To estimate errors and validate models, the authors generated 50 mock galaxy samples using $\Lambda$CDM N-body simulations (Particle-Mesh code).
  • Biasing: Mock galaxies were created by placing particles in regions where the smoothed dark matter density exceeded a threshold ($\nu_{th}$). This Eulerian bias prescription produced samples with linear bias parameters $b_1 = 1.0, 1.2,$ and $1.4$.
  • Redshift Space: Simulated particles were mapped to redshift space to match the observational data, accounting for peculiar velocities.

Analysis Technique

• Bispectrum Estimation: The authors employed a fast binned bispectrum estimator (Shaw et al. 2021).
  • Parameterization: Triangle configurations were parameterized by the largest side $k_1$ (size) and two dimensionless shape parameters:
    • $\mu = \cos(\theta_{12})$: The cosine of the angle between the two largest sides.
    • $t = k_2/k_1$: The ratio of the second largest side to the largest.
  • Binning: The analysis covered $k_1 \in (0.034 - 0.434) \text{ Mpc}^{-1}$ and the $(\mu, t)$ plane (restricted to $0.5 \le \mu, t \le 1$and$2\mu t \ge 1$). The plane was divided into$5 \times 5$ bins.
  • Total Triangles: The analysis considered approximately $1.37 \times 10^8$ triangles.
• Noise Correction: Shot noise was corrected using the standard prescription for discrete galaxy sampling.
• Error Estimation: Statistical uncertainties were derived from the 50 mock realizations (specifically those with $b_1=1.2$).

3. Key Contributions

1. High-Resolution Shape Dependence: Unlike previous works that focused on specific configurations (e.g., equilateral or squeezed), this study provides a comprehensive map of the bispectrum across the entire allowed $(\mu, t)$ shape space for a local volume-limited sample.
2. Power Law Modeling: The authors demonstrate that for any fixed triangle shape, the bispectrum amplitude follows a robust power-law dependence on the size scale $k_1$.
3. Red vs. Blue Galaxy Analysis: The study explicitly compares the bispectrum of red (old, clustered) and blue (young, less clustered) galaxies, quantifying how galaxy formation history and environment affect non-Gaussian clustering.
4. Validation of Bias Models: The work validates a simple Eulerian bias prescription ($b_1 \approx 1.2$) for the SDSS Main Galaxy Sample, contrasting it with the more complex biasing required for high-redshift, high-bias samples like BOSS.

4. Key Results

General Bispectrum Behavior

• Power Law Fit: The measured bispectrum $B(k_1, \mu, t)$ is well-modeled by a power law:
  $$B(k_1, \mu, t) = A \left( \frac{k_1}{1 \text{ Mpc}^{-1}} \right)^n$$
  where $A$ is the amplitude and $n$ is the index (always negative).
• Shape Dependence of Amplitude ($A$):
  • $A$ is minimum for equilateral triangles ($\mu \approx 0.5, t \approx 1$).
  • $A$ increases as the triangle deforms toward linear configurations (where the two largest sides are nearly aligned, $\mu \to 1$).
  • The maximum amplitude is found for linear triangles with $\mu = 0.95, t = 0.75$.
  • The amplitude for linear triangles is nearly an order of magnitude larger than for equilateral triangles.
• Shape Dependence of Index ($n$):
  • The index $n$ is negative for all shapes.
  • The magnitude $|n|$ is minimum ($\approx 3.12$) for S-isosceles triangles ($\mu=0.65, t=0.75$).
  • The magnitude $|n|$ is maximum ($\approx 3.8$) for acute triangles ($\mu=0.65, t=0.85$).
  • Generally, $|n|$ decreases as the triangle shape transitions from equilateral to linear.

Comparison with Simulations

• The bispectrum from the mock samples with linear bias $b_1 = 1.2$ shows excellent agreement with the SDSS data across all triangle shapes.
• Mocks with $b_1=1.0$ under-predict the amplitude, while $b_1=1.4$ over-predicts it.
• The simple Eulerian bias model is sufficient for the SDSS Main Galaxy Sample, unlike the BOSS CMASS sample which requires non-local bias prescriptions.

Red vs. Blue Galaxies

• Amplitude: Red galaxies exhibit a significantly higher bispectrum amplitude ($A$) than blue galaxies for all triangle configurations.
• Interpretation: Red galaxies are older and reside in denser environments (high-mass halos). Their higher bispectrum indicates stronger non-linear evolutionary interactions and clustering. Blue galaxies are less biased ($b_1 \approx 1$) and trace the underlying dark matter field more linearly.
• Slope: The variation of the slope $n$ with shape differs between red and blue populations, though no single generic trend dominates.

Non-Gaussianity Metrics

• The dimensionless skewness $\tilde{\mu}_3 = \Delta_3 / (\Delta_2)^{1.5}$ (where $\Delta_3$ is the dimensionless bispectrum) is found to be $>1$ for all configurations, confirming the highly non-Gaussian nature of the LSS at these scales.
• Skewness is highest for linear triangles (values 6–50) compared to equilateral triangles (values 2–5), indicating that linear configurations are the most sensitive probes of non-Gaussianity.

5. Significance

This paper provides a rigorous, high-resolution characterization of the galaxy bispectrum in the local Universe. By decoupling the analysis from cosmological parameter estimation, it isolates the intrinsic statistical properties of galaxy clustering. The findings confirm that:

1. Non-Gaussianity is shape-dependent: The degree of non-Gaussianity is maximized in linear triangle configurations, reflecting the filamentary and sheet-like structure of the cosmic web.
2. Galaxy type matters: The bispectrum is a powerful tool to distinguish between different galaxy populations, revealing that red galaxies are significantly more biased and non-linearly clustered than blue galaxies.
3. Modeling validity: Simple bias models are effective for low-redshift, low-bias samples, providing a baseline for future studies using larger, more complex datasets (like DESI or Euclid) where higher-order bias terms may become necessary.

The work establishes a baseline for interpreting higher-order statistics in upcoming large-scale structure surveys and highlights the importance of shape-dependent analysis in understanding the non-linear evolution of the Universe.