🔭 astrophysics

Cosmological constraints from the DESI DR1 Bispectrum Full-Shape and DR2 BAO

This paper presents robust cosmological constraints from combining DESI DR1 full-shape bispectrum data with DR2 BAO measurements, demonstrating that the inclusion of the bispectrum tightens parameter uncertainties, shifts amplitude parameters, and consistently weakens evidence for evolving dark energy and massive neutrinos compared to power-spectrum-only analyses.

Original authors: D. Forero-Sánchez, S. Novell-Masot, H. Gil-Marín, L. Verde, J. Aguilar, S. Ahlen, D. Bianchi, A. Brodzeller, D. Brooks, F. J. Castander, S. Cole, A. de la Macorra, J. Della Costa, Biprateep Dey, P. Do

Published 2026-06-25

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: D. Forero-Sánchez, S. Novell-Masot, H. Gil-Marín, L. Verde, J. Aguilar, S. Ahlen, D. Bianchi, A. Brodzeller, D. Brooks, F. J. Castander, S. Cole, A. de la Macorra, J. Della Costa, Biprateep Dey, P. Doel, S. Ferraro, A. Font-Ribera, J. E. Forero-Romero, Satya Gontcho A Gontcho, G. Gutierrez, C. Hahn, H. K. Herrera-Alcantar, K. Honscheid, D. Huterer, M. Ishak, D. Kirkby, A. Kremin, O. Lahav, C. Lamman, M. Landriau, L. Le Guillou, M. E. Levi, M. Manera, A. Meisner, R. Miquel, J. Moustakas, S. Nadathur, J. A. Newman, G. Niz, N. Palanque-Delabrouille, W. J. Percival, F. Prada, I. Pérez-Ràfols, G. Rossi, L. Samushia, E. Sanchez, D. Schlegel, M. Schubnell, J. Silber, G. Tarlé, B. A. Weaver

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, expanding loaf of raisin bread. As the dough rises, the raisins (galaxies) move apart. For decades, cosmologists have been trying to measure exactly how fast this bread is rising and how the raisins are clumping together to understand the "secret ingredients" holding the recipe together: Dark Energy, Neutrinos, and Gravity.

This paper is like a master chef (the DESI collaboration) tasting a new, more complex batch of this cosmic bread. They aren't just looking at the distance between raisins; they are also listening to the sound of the dough rising and feeling the texture of the clumps.

Here is the breakdown of their findings in everyday terms:

1. The New Ingredients: "The Bispectrum"

In the past, scientists mostly looked at the Power Spectrum. Think of this as listening to a single note played by the universe. It tells you the average distance between galaxies. It's useful, but it's like trying to understand a symphony by only listening to the bass drum.

This paper introduces the Bispectrum. If the Power Spectrum is a single note, the Bispectrum is listening to how three notes interact at once. It's like listening to the harmony between three instruments.

The Analogy: Imagine trying to guess the shape of a room by walking in a straight line (Power Spectrum). Now, imagine walking in a triangle, measuring the angles and distances between three points (Bispectrum). The triangle gives you much more information about the room's shape and how the air (dark matter) is moving.

2. The Data: A Bigger, Better Map

The team used data from the Dark Energy Spectroscopic Instrument (DESI).

DR1 (Data Release 1): A map of about 6 million galaxies.
DR2 (Data Release 2): A newer, bigger map with over 14 million galaxies.
The Challenge: Because the new map overlaps with the old one, the data points are "correlated" (like taking two photos of the same scene from slightly different angles). The team had to be very careful to account for this overlap so they didn't double-count the evidence. They used computer simulations (mocks) to figure out exactly how these two datasets talk to each other.

3. The "ShapeFit" Trick: Avoiding the "Prior Volume" Trap

When scientists look for new physics (like changing Dark Energy), they often get confused by their own assumptions. This is called the "prior volume effect."

The Analogy: Imagine trying to find a lost coin in a dark room. If you assume the coin is definitely under the bed, you might ignore the fact that it's actually on the table. In complex math, if you assume too much about where the answer might be, you can accidentally push the answer away from the truth.
The Solution: The team used a method called ShapeFit. Instead of trying to fit the whole messy puzzle at once, they compressed the data into a few key "shape" parameters. This acts like a filter that removes the bias of their own assumptions, letting the data speak for itself more clearly.

4. What They Found: The "Taste Test" Results

A. The "Clumpiness" of the Universe ( $\sigma_8$ and $S_8$ )

The Finding: When they added the "Bispectrum" (the triangle measurement), the universe looked clumpier than before.
The Shift: The measurement of how much matter clumps together ( $\sigma_8$ ) went up by about 1.1 standard deviations. It's like realizing the raisin bread is actually denser and more tightly packed than the single-note measurements suggested.
Why it matters: This helps solve a long-standing puzzle where different telescopes disagreed on how "clumpy" the universe is. The new, more detailed measurement pushes the answer closer to what the Cosmic Microwave Background (the "baby picture" of the universe) predicts.

B. Is Dark Energy Changing? ( $w_0, w_a$ )

The Question: Is Dark Energy (the force pushing the universe apart) a constant "cosmological constant," or is it a dynamic fluid that changes over time?
The Finding: When using just the new DESI data, the "Bispectrum" made the results look more like a constant. It pushed the data back toward the idea that Dark Energy is unchanging (the standard $\Lambda$ CDM model).
The Twist: When they combined this with data from the Cosmic Microwave Background (CMB), the tension returned. The combined data suggested a 2.8-sigma preference for Dark Energy changing over time. However, the paper notes that the "Bispectrum" consistently tries to pull the answer back toward the "constant" model, weakening the evidence for change compared to older methods.

C. The Ghost Particles (Neutrinos)

The Finding: Neutrinos are tiny, ghostly particles that barely interact with anything. Their mass is hard to measure.
The Result: Without the Bispectrum, the data suggested the total mass of neutrinos was consistent with zero. With the Bispectrum, the data shifted to suggest a positive mass (around 0.26 eV).
The Analogy: It's like finally hearing the faint hum of a refrigerator in a quiet room. The "single note" measurement was too quiet to hear it, but the "harmony" of the Bispectrum made the hum audible, confirming that these ghost particles do have a tiny bit of weight, consistent with what particle physics experiments expect.

D. Modified Gravity

The Question: Does gravity work exactly as Einstein predicted, or is it slightly different on cosmic scales?
The Finding: The data is consistent with Einstein's General Relativity. The "Bispectrum" helped tighten the constraints, making the test for "new gravity" much more precise, but it didn't find any evidence that Einstein was wrong.

Summary

This paper is a "quality control" upgrade for our map of the universe. By listening to the "three-note harmony" (Bispectrum) instead of just the "single note" (Power Spectrum), and by using a clever compression trick (ShapeFit) to avoid bias, the team found:

The universe is clumpier than we thought.
The evidence for changing Dark Energy is weaker when using this new, more robust method.
We can now hear the mass of neutrinos more clearly.
Einstein's gravity still holds up under this stricter test.

The authors conclude that while the universe is still a bit of a mystery, this new way of listening to it gives us a much clearer, more reliable picture of the cosmic recipe.

Technical Summary: Cosmological Constraints from the DESI DR1 Bispectrum Full-Shape and DR2 BAO

Problem Statement
The Dark Energy Spectroscopic Instrument (DESI) provides unprecedented precision in mapping large-scale structure. While Baryon Acoustic Oscillation (BAO) measurements from DESI Data Release 2 (DR2) offer robust constraints on cosmic expansion, the Full-Shape (FS) analysis of the galaxy power spectrum contains additional information regarding structure growth, neutrino masses, and modified gravity. However, previous FS analyses of models beyond the standard $\Lambda$ CDM cosmology have suffered from "prior volume effects," where the interpretation of constraints becomes unreliable due to the sensitivity of the likelihood to prior choices in high-dimensional parameter spaces. Furthermore, standard power-spectrum-only analyses often leave the growth rate ( $f$ ) and amplitude ( $\sigma_8$ ) parameters degenerate. There is a need to combine DESI DR1 FS data (including higher-order statistics like the bispectrum) with DR2 BAO data while properly accounting for their cross-covariance to produce robust, prior-independent constraints on extended cosmological models.

Methodology
The authors present a joint analysis combining DESI DR1 Full-Shape measurements (specifically including the Luminous Red Galaxy (LRG) bispectrum) with DESI DR2 BAO data. The methodology relies on three key technical components:

ShapeFit Compression: Instead of "full modeling" (fitting perturbation theory directly to the broadband power spectrum), the authors employ the ShapeFit compression technique. This approach parameterizes the clustering signal via scale dilation parameters ( $q_{\parallel}, q_{\perp}$ ), the logarithmic growth rate ( $f$ ), the amplitude ( $\sigma_8$ ), and shape parameters ( $m, n$ ) capturing slope variations. This decouples data compression from cosmological inference, mitigating prior volume effects that typically plague beyond- $\Lambda$ CDM analyses.
Bispectrum Inclusion: The analysis incorporates the LRG bispectrum monopole alongside the power spectrum monopole, quadrupole, and hexadecapole. The inclusion of the bispectrum is critical for breaking the degeneracy between the growth rate ( $f$ ) and the amplitude ( $\sigma_8$ ), which cannot be disentangled using the power spectrum alone in the compressed framework.
Cross-Covariance Estimation: To combine DR1 (FS) and DR2 (BAO) data, the authors utilize the DESI EZmock suite to estimate the full cross-covariance matrix between the two datasets in the compressed parameter space. They demonstrate that while cross-correlations are negligible for power-spectrum-only analyses, they are non-negligible when including higher-order statistics like the bispectrum. The final covariance matrix combines data-derived diagonal blocks with mock-derived off-diagonal cross-blocks.

Key Contributions

First Joint DR1/DR2 Analysis with Bispectrum: This work provides the first robust constraints on extended cosmological models using the combination of DESI DR1 FS (with bispectrum) and DR2 BAO, properly accounting for their cross-covariance.
Mitigation of Prior Volume Effects: By utilizing ShapeFit compression, the analysis avoids the prior sensitivity issues that have historically hindered DESI-only constraints on dynamical dark energy and other extended models.
Decoupling of Growth and Amplitude: The inclusion of the bispectrum allows for the independent measurement of $f$ and $\sigma_8$ within the compressed ShapeFit framework, a capability previously limited to full modeling approaches.

Results
The paper presents constraints across several cosmological models using DESI-only data (supplemented by minimal priors on baryon density $\omega_b$ and spectral index $n_s$ ) and combinations with Cosmic Microwave Background (CMB) and Supernovae (DES-Dovekie) data.

$\Lambda$ CDM Constraints: The inclusion of the bispectrum shifts the inferred values of $\sigma_8$ up by $1.1\sigma$ and $S_8$ by $1.2\sigma$ relative to power-spectrum-only analyses. Uncertainties on these amplitude parameters are reduced by 26% and 28%, respectively. Background parameters ( $h, \Omega_m, n_s$ ) show marginal shifts, dominated by the precision of the DR2 BAO data.
Dynamical Dark Energy ( $w_0 w_a$ CDM): In DESI-only analyses, the bispectrum shifts the dark energy equation-of-state parameters ( $w_0, w_a$ ) closer to the $\Lambda$ CDM point ( $w_0=-1, w_a=0$ ), reducing the preference for evolving dark energy. When CMB data is added, the combination shows a $2.8\sigma$ deviation from $\Lambda$ CDM; however, the bispectrum consistently weakens the evidence for time-varying dark energy compared to power-spectrum-only analyses.
Neutrino Masses: The bispectrum enhances sensitivity to massive neutrinos. In DESI-only analysis, the power-spectrum-only posterior for the sum of neutrino masses ( $\sum m_\nu$ ) is consistent with zero, whereas adding the bispectrum shifts the mean to $0.26 \pm 0.17$ eV, with a 95% upper limit of 0.57 eV, aligning the peak with oscillation lower bounds.
Modified Gravity: The bispectrum breaks the growth-amplitude degeneracy in the ShapeFit framework, allowing for constraints on the modified gravity parameter $\mu_0$ . The result $\mu_0 = 0.12 \pm 0.49$ is consistent with General Relativity. While the constraint is $\sim10\%$ weaker than full modeling (due to the bispectrum being limited to LRGs while full modeling uses all tracers), it represents a significant improvement over power-spectrum-only ShapeFit constraints.
Spatial Curvature: The bispectrum shifts the mean of the spatial curvature parameter $\Omega_k$ by $0.6\sigma$ toward positive values in DESI-only analyses but does not significantly alter the uncertainty.

Significance and Claims
The paper claims that accounting for cross-dataset covariances and avoiding prior volume effects yields robust constraints on extended cosmological models using DESI data alone. The authors emphasize that the bispectrum consistently raises amplitude parameters ( $\sigma_8, S_8$ ) and tightens their uncertainties. Crucially, the inclusion of the bispectrum mitigates the apparent preference for evolving dark energy found in some power-spectrum-only analyses, shifting posteriors toward $\Lambda$ CDM. The work demonstrates that the ShapeFit methodology, when augmented with bispectrum information, provides a powerful and computationally efficient alternative to full modeling for testing deviations from standard cosmology, offering a handle on modified gravity and neutrino masses that was previously inaccessible with compressed statistics alone.