Baryon Acoustic Oscillations

Baryon acoustic oscillations (BAO) are the imprint of sound waves in the early universe on the later distribution of matter. They produce a characteristic separation scale in the clustering of galaxies and other tracers of cosmic structure. In cosmology, this scale serves as a standard ruler: its observed angular extent and redshift extent constrain distances and the expansion rate at different epochs. The term describes both the original acoustic phenomenon and its surviving statistical signature, rather than sound waves still propagating through today’s galaxy distribution. (arxiv.org)

Physical origin

In the hot early universe following the Big Bang, ordinary matter existed as an ionized plasma. Photons scattered from free electrons, while electromagnetic interactions coupled electrons to atomic nuclei. Radiation and baryons therefore behaved approximately as a tightly coupled fluid. Baryons supplied inertia, while radiation supplied much of the restoring pressure. Density perturbations in this fluid underwent acoustic oscillations under the combined effects of pressure and gravity. Dark matter, which did not participate in the photon–baryon coupling, evolved differently. (arxiv.org)

An idealized, localized initial overdensity illustrates the process: the photon–baryon disturbance propagates outward as a sound wave, while a dark-matter overdensity remains near the original location. During recombination, neutral atoms form and photon scattering becomes inefficient. The subsequent cessation of effective photon drag on baryons ends their acoustic propagation. Later gravitational evolution preserves a weakened excess of matter at approximately the distance reached by the wave. Overlapping disturbances throughout space create a statistical preferred separation, not a collection of distinct observable shells. (arxiv.org)

The sound horizon

The relevant ruler is the comoving sound horizon at the baryon drag epoch, usually written $r_d$ . “Comoving” means that the overall expansion of the universe has been factored out. Its definition is

r_d=\int_{z_d}^{\infty}\frac{c_s(z)}{H(z)}\,dz,

where $z_d$ is the redshift of the drag epoch, $c_s(z)$ is the photon–baryon sound speed, and $H(z)$ is the expansion rate. In the tightly coupled approximation,

c_s=\frac{c}{\sqrt{3(1+R)}}, \qquad R=\frac{3\rho_b}{4\rho_\gamma},

with $c$ the speed of light and $\rho_b,\rho_\gamma$ the baryon and photon energy densities in consistent units. Increasing the baryon loading reduces the sound speed. (arxiv.org)

For cosmological parameters close to those favored by observations within the Lambda–cold dark matter model, $r_d$ is about 147 megaparsecs. This is a comoving length, not the physical size of the sound horizon when the waves stopped. It is also not a universal constant: its value depends on the early expansion history and the matter and radiation content. Precision calculations require an accurate treatment of recombination and baryon drag. (arxiv.org)

Statistical signature

BAO are measured from large samples rather than individual objects. Two closely related statistics describe the feature:

The two-point correlation function, $\xi(r)$ , measures the excess probability of finding pairs at separation $r$ . BAO produce a broad acoustic peak.
The matter power spectrum, $P(k)$ , measures the strength of density fluctuations at different spatial wavenumbers. BAO appear as a sequence of small oscillations superimposed on a smoother spectrum.

These descriptions are related through a Fourier transform. A preferred separation in real space corresponds to oscillatory structure in Fourier space. The fitted acoustic scale is more informative than simply reading the maximum of the broad correlation-function peak, whose position and shape are affected by other contributions to clustering. (arxiv.org)

The acoustic peaks in the cosmic microwave background arise from the same early photon–baryon dynamics. However, CMB anisotropies mainly record conditions at photon last scattering, whereas the late-time BAO ruler is associated with the baryon drag epoch. The two sound horizons are closely related but not identical. (arxiv.org)

Measuring distance and expansion

For a narrow redshift interval, transverse and radial BAO scales approximately satisfy

\Delta\theta\simeq\frac{r_d}{D_M(z)}, \qquad \Delta z\simeq\frac{H(z)r_d}{c},

where $D_M(z)$ is the transverse comoving distance and $\Delta\theta$ is measured in radians. An anisotropic analysis therefore constrains

\frac{D_M(z)}{r_d}, \qquad \frac{D_H(z)}{r_d}, \qquad D_H(z)=\frac{c}{H(z)}.

The angular-diameter distance is $D_A=D_M/(1+z)$ . Separating directions across and along the line of sight distinguishes accumulated distance from the expansion rate at the observed epoch. (arxiv.org)

When data do not support this separation, an approximately isotropic measurement commonly constrains

\frac{D_V(z)}{r_d}, \qquad D_V(z)=\left[zD_M^2(z)D_H(z)\right]^{1/3}.

This combines radial and transverse information. BAO thus directly measure distance ratios relative to the ruler; converting them into absolute distances requires a calibration of $r_d$ . (arxiv.org)

Measurements over several redshifts constrain cosmic geometry and the expansion history, including the effects of dark energy. They complement supernova distances, CMB observations, and gravitational lensing. Constraints on the Hubble constant, spatial curvature, or the dark-energy equation of state depend on the adopted cosmological model and additional data. BAO do not independently identify the physical substance responsible for accelerated expansion. (arxiv.org)

Observation and historical development

The first convincing late-time BAO detections were reported in 2005 by two independent survey analyses. The Sloan Digital Sky Survey detected an acoustic peak in the correlation function of 46,748 luminous red galaxies. The final Two-degree Field Galaxy Redshift Survey analysis found evidence for baryon oscillations in the galaxy power spectrum. These results connected early-universe acoustic physics with the distribution of galaxies billions of years later. (arxiv.org)

Subsequent surveys increased the volume sampled and measured the ruler over wider redshift ranges. Besides galaxies and quasars, observations use the Lyman-alpha forest: absorption features in distant quasar spectra that trace intervening intergalactic hydrogen. In 2025, the Dark Energy Spectroscopic Instrument collaboration reported its second-data-release galaxy and quasar BAO analysis, based on more than 14 million objects, together with companion Lyman-alpha forest measurements. (arxiv.org)

Reconstruction and observational limitations

Gravitational evolution moves matter away from its initial positions, broadening the acoustic feature and slightly shifting its scale. Density-field reconstruction estimates large-scale displacements from the observed density distribution and approximately reverses them. This sharpens the BAO feature and improves distance precision; it does not recover every detail of the primordial density field. (arxiv.org)

Important observational limitations include finite survey volume, sparse sampling, redshift errors, survey boundaries, and nonuniform target selection. Peculiar velocities distort clustering along the line of sight, while galaxies are biased tracers of the underlying matter distribution. Analyses account for these effects through modeling, survey weights, simulated catalogs, and tests of alternative fitting choices. The acoustic scale is comparatively robust, but precision measurements still require a systematic-error budget. (arxiv.org)

Model dependence and interpretation

The principal distinction is between measuring the acoustic feature and interpreting it cosmologically. Changing early-universe physics can change $r_d$ , so an accurately measured distance ratio need not imply the same absolute distance in every model. BAO-based inferences about the Hubble constant must retain this calibration dependence. (arxiv.org)

Similarly, evidence for time-dependent dark energy comes from comparing cosmological models against combined datasets, not from directly observing dark energy in the acoustic feature. The DESI second-data-release analysis reported that preferences for evolving dark energy depended on the combination of BAO, CMB, and supernova data used. Such results are statistical and model-dependent, and must be distinguished from the underlying measurements of the distance–redshift relation. (arxiv.org)