Validation of the DESI DR2 measurements of baryon acoustic oscillations from galaxies and quasars

U. Andrade, E. Paillas, J. Mena-Fernández, Qinxun Li, Ashley J. Ross, S. Nadathur, M. Rashkovetskyi, A. Pérez-Fernández, Hee‐Jong Seo, Nicole M. Sanders, O. Alves, Xinyi Chen, N. Deiosso, Arnaud de Mattia, Martin White, M. Karim, S. P. Ahlen, E. Armengaud, Alejandro Avilés, P. Bansal, Jayashree Behera, D. Bianchi, S. Brieden, A. Brodzeller, D. Brooks, E. Burtin, R. Calderón, Rebecca Canning, A. Carnero Rosell, Lidia Casas, F. J. Castander, M. Charles, E. Chaussidon, J. Chaves-Montero, T. Claybaugh, Shaun Cole, Andrew P. Cooper, Andrei Cuceu, Kyle Dawson, Axel de la Macorra, J. Della Costa, Arjun Dey, Biprateep Dey, Z. Ding, P. Doel, Daniel J. Eisenstein, Willem Elbers, E. Fernández-García, Simone Ferraro, Andreu Font-Ribera, J. E. Forero-Romero, C. García-Quintero, Lehman H. Garrison, E. Gaztañaga, Héctor Gil-Marín, Satya Gontcho A Gontcho, Alma X. González‐Morales, C. Gordon, G. Gutiérrez, J. Guy, ChangHoon Hahn, Song He, H. K. Herrera-Alcantar, K. Honscheid, Cullan Howlett, Dragan Huterer, Mustapha Ishak, S. Juneau, R. Kehoe, D. Kirkby, Theodore Kisner, Anthony Kremin, O. Lahav, C. Lamman, Martin Landriau, L. Le Guillou, Alexie Leauthaud, M. E. Levi, C. Magneville, Marc Manera, Paul Martini, William L. Matthewson, Aaron Meisner, R. Miquel, John Moustakas, A. Muñoz-Gutiérrez, D. Muñoz-Santos, Adam D. Myers, L. Napolitano, Jeffrey A. Newman, H. E. Noriega, N. Palanque‐Delabrouille, Jiaming Pan, Will J. Percival, Ignasi Pérez-Ràfols, C. Poppett, Francisco Prada, Anand Raichoor, C. Ramírez-Pérez, C. RavouxPhysical Review. D/Physical Review. D 2025 · DOI

TL;DR

This paper validates the Baryon Acoustic Oscillation (BAO) measurements from the Dark Energy Spectroscopic Instrument's (DESI) second data release (DR2). By rigorously testing the analysis pipeline with blinded data and mock catalogs, the authors confirm that DESI DR2 provides significantly more precise BAO constraints than previous surveys, paving the way for improved understanding of dark energy and the Universe's expansion history.

What Problem Does This Paper Address?

Understanding the expansion history of the Universe and the nature of dark energy is a central goal in cosmology. Baryon Acoustic Oscillations (BAOs) provide a powerful "standard ruler" imprinted in the large-scale distribution of matter, allowing us to measure cosmic distances and probe expansion. However, extracting precise BAO measurements from complex galaxy survey data is challenging due to various systematic effects, including galaxy bias, observational systematics, and non-linear structure growth. This paper addresses the critical need to validate the BAO measurements from DESI's Data Release 2 (DR2), ensuring their accuracy and reliability before they are used for cosmological inference. This validation is crucial because DR2 represents a significant leap in data volume and precision compared to previous releases and surveys.

Methods

The authors employed a comprehensive validation strategy, building upon the established DESI DR1 BAO analysis pipeline. Key methods include:

  • Catalog-level Blinding: The DR2 BAO measurements were kept "blinded" throughout the validation process to prevent confirmation bias. This involved modifying redshifts and weights to mimic a different cosmology.
  • Robustness Tests: A suite of tests were performed on the blinded data, including:
  • Spatial and Imaging Splits: Analyzing data from different survey regions (Galactic Caps, imaging sources) to check for biases related to observational conditions.
  • Intrinsic Property Splits: Dividing galaxy samples by stellar mass and magnitude to assess dependencies on galaxy properties and potential biases in galaxy bias modeling.
  • Reconstruction Techniques: Applying advanced methods to recover the linear BAO signal from non-linear effects, significantly improving precision.
  • Systematic Error Treatment: Carefully accounting for various sources of error, including theoretical systematics, HOD (Halo Occupation Distribution) systematics, and fiducial cosmology uncertainties, with tracer-dependent refinements for DR2.
  • Mock Catalogs: Extensive use of AbacusSummit 2nd-generation mocks (Abacus-2 DR2 mocks) to simulate the DESI survey and validate the analysis pipeline, compare results across different tracers and redshift bins, and assess statistical uncertainties.
  • BAO Modeling: Employing a detailed model for the galaxy power spectrum that includes smooth and oscillatory components, accounting for geometric distortions (Alcock-Paczynski effect) and non-linear damping.
  • Parameter Inference: Using Markov Chain Monte Carlo (MCMC) methods to derive BAO scaling parameters and assess their uncertainties.

Datasets

  • DESI Data Release 2 (DR2) LSS Catalogs: This is the primary dataset for the BAO measurements. It comprises spectra from nearly three years of DESI observations, covering a vast area of the sky and including millions of galaxies and quasars.
  • AbacusSummit 2nd-generation mocks (Abacus-2 DR2 mocks): These synthetic datasets were crucial for validating the analysis pipeline, testing systematic effects, and estimating uncertainties. They were generated to replicate the selection and clustering properties of the DR2 galaxy and quasar samples.

Key Insights and Ideas

  • Reconstruction is Key: The paper strongly emphasizes how density-field reconstruction significantly enhances the precision of BAO measurements, reducing uncertainties by up to 42-47% compared to unreconstructed catalogs. This is a critical technique for maximizing the scientific return from large-scale structure surveys.
  • DESI DR2 is a Major Leap: The DR2 BAO measurements offer a factor of 2.15 improvement in precision compared to DR1, and a factor of 2.56 improvement over previous SDSS BAO measurements. This highlights the power of DESI's increased survey volume and improved analysis.
  • Robustness Across the Board: The BAO constraints are shown to be remarkably stable across a wide range of analysis choices, including different data vectors, modeling approaches, and systematic treatments. This builds confidence in the reliability of the results.
  • Tracer-Specific Systematics Matter: For DR2, the analysis moved towards tracer-dependent treatments of systematic errors (e.g., HOD systematics), leading to a more refined and accurate error budget.
  • Gaussianity is a Good Approximation: The posterior distributions of the BAO scaling parameters are well-approximated by a Gaussian, simplifying their use in subsequent cosmological analyses.

Key Results

  • Improved BAO Precision: The combined BAO precision across six redshift bins has improved from ~0.52% in DR1 to ~0.24% in DR2.
  • Consistency with DR1: The DR2 BAO measurements are consistent with DR1, while achieving significantly reduced statistical uncertainties.
  • Stability Across Tests: BAO constraints remain stable across various robustness tests, including different data splits (spatial, magnitude, mass), modeling choices, and systematic treatments.
  • Successful Validation: A comprehensive "unblinding checklist" of tests was passed, confirming the readiness of the DR2 BAO analysis for cosmological interpretation.

Limitations and Future Work

  • BGS 1D vs. 2D Fit Decision: While 2D BAO fits were explored for BGS, the authors conservatively opted for a 1D fit due to concerns about non-Gaussianity in the anisotropic parameter and a higher-than-expected correlation between isotropic and anisotropic parameters. This decision was made to ensure robustness, but future work might explore 2D fits with improved modeling or larger datasets.
  • BGS Mock Discrepancy: A slight discrepancy was noted in the clustering amplitude of BGS mocks compared to data, which is being investigated. Upcoming analyses will incorporate improved mocks to address this.
  • Future Refinements: The authors suggest that future work will focus on refining systematics control and exploring alternative modeling approaches to further enhance the precision and reliability of DESI's BAO measurements.

Summary updated

Abstract

The Dark Energy Spectroscopic Instrument (DESI) Data Release 2 (DR2) galaxy and quasar clustering data represents a significant expansion of data from Data Release 1 (DR1), providing improved statistical precision in baryon acoustic oscillation (BAO) constraints across multiple tracers, including bright galaxies, luminous red galaxies, emission line galaxies, and quasars. In this paper, we validate the BAO analysis of DR2. We present the results of robustness tests on the blinded DR2 data and, after unblinding, consistency checks on the unblinded DR2 data. All results are compared with those obtained from a suite of mock catalogs that replicate the selection and clustering properties of the DR2 sample. We confirm the consistency of DR2 BAO measurements with DR1 while achieving a reduction in statistical uncertainties due to the increased survey volume and completeness. The combined BAO precision, including both statistical and systematic errors, improves from <a:math xmlns:a="http://www.w3.org/1998/Math/MathML" display="inline"> <a:mo>∼</a:mo> <a:mn>0.52</a:mn> <a:mo>%</a:mo> </a:math> in DR1 to 0.30% in DR2—a factor of 1.7 gain. We assess the impact of analysis choices, including different data vectors (correlation function vs power spectrum), modeling approaches and systematics treatments, and an assumption of the Gaussian likelihood, finding that our BAO constraints are stable across these variations and assumptions with a few minor refinements to the baseline setup of the DR1 BAO analysis. We summarize a series of pre-unblinding tests that confirmed the readiness of our analysis pipeline, the final systematic errors, and the DR2 BAO analysis baseline. The successful completion of these tests led to the unblinding of the DR2 BAO measurements, ultimately leading to the DESI DR2 cosmological analysis, with their implications for the expansion history of the Universe and the nature of dark energy presented in the DESI key paper (companion paper).
BibTeX
@article{andrade2025validation,
  author = {U. Andrade and E. Paillas and J. Mena-Fernández and Qinxun Li and Ashley J. Ross and S. Nadathur and M. Rashkovetskyi and A. Pérez-Fernández and Hee‐Jong Seo and Nicole M. Sanders and O. Alves and Xinyi Chen and N. Deiosso and Arnaud de Mattia and Martin White and M. Karim and S. P. Ahlen and E. Armengaud and Alejandro Avilés and P. Bansal and Jayashree Behera and D. Bianchi and S. Brieden and A. Brodzeller and D. Brooks and E. Burtin and R. Calderón and Rebecca Canning and A. Carnero Rosell and Lidia Casas and F. J. Castander and M. Charles and E. Chaussidon and J. Chaves-Montero and T. Claybaugh and Shaun Cole and Andrew P. Cooper and Andrei Cuceu and Kyle Dawson and Axel de la Macorra and J. Della Costa and Arjun Dey and Biprateep Dey and Z. Ding and P. Doel and Daniel J. Eisenstein and Willem Elbers and E. Fernández-García and Simone Ferraro and Andreu Font-Ribera and J. E. Forero-Romero and C. García-Quintero and Lehman H. Garrison and E. Gaztañaga and Héctor Gil-Marín and Satya Gontcho A Gontcho and Alma X. González‐Morales and C. Gordon and G. Gutiérrez and J. Guy and ChangHoon Hahn and Song He and H. K. Herrera-Alcantar and K. Honscheid and Cullan Howlett and Dragan Huterer and Mustapha Ishak and S. Juneau and R. Kehoe and D. Kirkby and Theodore Kisner and Anthony Kremin and O. Lahav and C. Lamman and Martin Landriau and L. Le Guillou and Alexie Leauthaud and M. E. Levi and C. Magneville and Marc Manera and Paul Martini and William L. Matthewson and Aaron Meisner and R. Miquel and John Moustakas and A. Muñoz-Gutiérrez and D. Muñoz-Santos and Adam D. Myers and L. Napolitano and Jeffrey A. Newman and H. E. Noriega and N. Palanque‐Delabrouille and Jiaming Pan and Will J. Percival and Ignasi Pérez-Ràfols and C. Poppett and Francisco Prada and Anand Raichoor and C. Ramírez-Pérez and C. Ravoux},
  title = {{Validation of the DESI DR2 measurements of baryon acoustic oscillations from galaxies and quasars}},
  journal = {Physical review. D/Physical review. D.},
  year = {2025},
  doi = {10.1103/kdys-w8vl},
}
Figure