Cosmic Microwave Background

The cosmic microwave background (CMB) is the pervasive, nearly isotropic bath of thermal relic electromagnetic radiation that fills the entire observable universe, originating from the epoch of recombination and photon decoupling approximately 378,000 years after the Big Bang (at cosmological redshift z_∗ = 1089.80 ± 0.21). Before this epoch, the early universe existed as a hot, opaque, ionized photon–baryon plasma of protons (H⁺), helium nuclei (⁴He²⁺), free electrons (e⁻), and photons (γ) coupled tightly via Thomson scattering. As cosmic expansion cooled the plasma to T_rec ≈ 3,000 K (corresponding to a thermal energy k_BT ≈ 0.26 eV), protons captured free electrons to form electrically neutral hydrogen atoms (H), causing the photon mean free path to exceed the Hubble horizon and allowing photons to stream freely across space from the surface of last scattering.

Over the subsequent 13.8 billion years of metric cosmic expansion, the wavelengths of these decoupled photons have been stretched by a factor of 1 + z_∗ ≈ 1,090, redshifting their blackbody spectrum from the visible/near-infrared regime into the microwave band. Today, the CMB exhibits the most precise thermal Planck blackbody spectrum measured in nature, with a monopole temperature of T₀ = 2.72548 ± 0.00057 K, a spectral radiance peak at frequency ν_peak = 160.23 GHz (wavelength λ_peak ≈ 1.871 mm), a photon number density of n_γ ≈ 410.7 photons cm⁻³, and a radiation energy density of ρ_γc² ≈ 0.260 eV cm⁻³. Superimposed on this uniform background are tiny primary temperature anisotropies at the level of ΔT / T₀ ≈ 10⁻⁵ (~18–100 μK), whose acoustic angular power spectrum—mapped at high precision by COBE, WMAP, and the Planck space observatory—provides the primary observational foundation for the standard ΛCDM (Lambda Cold Dark Matter) cosmological model.

Classify the Cosmos: Explore deep-sky structures, galaxies, and cosmological landmarks across the observable universe in Galaxy.

From Alpher, Herman, and Gamow to Penzias and Wilson (1965)

The existence of a cosmic thermal relic background was first predicted quantitatively in 1948 by Ralph Alpher and Robert Herman (working with George Gamow) as a direct consequence of hot Big Bang nucleosynthesis (αβγ theory). Calculating the expansion cooling of the radiation-dominated early universe from the nuclear freeze-out epoch (T ~ 10⁹ K at t ~ 180 s) to the present matter density, Alpher and Herman estimated a present-day background radiation temperature of T₀ ≈ 5 K. Their prediction was largely overlooked for fifteen years until the early 1960s, when Yakov Zel'dovich and Igor Novikov in the Soviet Union and Robert Dicke, Jim Peebles, Peter Roll, and David Wilkinson at Princeton University independently recalculated the thermal relic radiation expected from an early hot universe and began constructing a Dicke radiometer to detect it.

Simultaneously, just 50 km away at Bell Telephone Laboratories in Crawford Hill, Holmdel, New Jersey, radio astronomers Arno Penzias and Robert Wilson were calibrating the ultra-sensitive 20-foot (6-meter) horn-reflector antenna originally built for Project Echo satellite communications. Operating at a wavelength of λ = 7.35 cm (4.08 GHz) with a liquid-helium-cooled maser amplifier, Penzias and Wilson detected an persistent, unpolarized, isotropic background antenna noise temperature of 3.5 ± 1.0 K that showed no diurnal, seasonal, or galactic-plane variation and remained unchanged after removing roosting pigeons and cleaning dielectric bird droppings from the horn throat. Upon learning of Dicke and Peebles' theoretical work at Princeton, the two groups published companion papers in the Astrophysical Journal in July 1965, decisively refuting the steady-state cosmological model (which lacked a hot dense early phase capable of thermalizing a cosmic blackbody) and earning Penzias and Wilson the 1978 Nobel Prize in Physics.

The COBE FIRAS Blackbody Spectrum and Photon Statistics

In November 1989, NASA launched the Cosmic Background Explorer (COBE) satellite, carrying the Far Infrared Absolute Spectrophotometer (FIRAS) led by John C. Mather and the Differential Microwave Radiometer (DMR) led by George F. Smoot (co-recipients of the 2006 Nobel Prize in Physics). Comparing the cosmic sky from 60 GHz to 630 GHz against an onboard precision isothermal blackbody calibrator, FIRAS demonstrated that the CMB spectral radiance I_ν obeys Max Planck's quantum Bose–Einstein blackbody radiation law (the quantum boson analogue of the Boltzmann distribution with zero chemical potential μ = 0):

I_ν(ν, T₀) = (2*h*ν³ / c²) × 1 / [exp(*h*ν / [k_BT₀]) − 1]

Deviations from a pure Planck blackbody across the peak of the spectrum are constrained to less than 50 parts per million (specifically, dimensionless Compton y-distortion |*y*| < 1.5 × 10⁻⁵ and chemical potential μ-distortion |μ| < 9 × 10⁻⁵ at 95% confidence). Integrating the Planck photon distribution over all frequencies yields the cosmic photon number density n_γ = [16π ζ(3) / (hc)³] (k_BT₀)³ ≈ 410.7 photons cm⁻³ and specific entropy density s_γ = (4π² / 45) k_B (k_BT₀ / ℏc)³. Comparing n_γ to the cosmic baryon density measured from acoustic peaks and primordial deuterium abundances yields the fundamental baryon-to-photon ratio η_B = n_b / n_γ = (6.12 ± 0.04) × 10⁻¹⁰—showing that there are roughly 1.6 billion CMB photons for every proton or neutron in the universe.

Saha Ionization Equilibrium and the Surface of Last Scattering

Prior to t ≈ 378,000 years, photons scattered continuously off free electrons with a Thomson scattering cross section σ_T = (8π/3)(e² / 4πε₀m_ec²)² = 6.652 × 10⁻²⁹ m². Although the ground-state binding energy of neutral hydrogen (H + γ ↔ p⁺ + e⁻) is B₁ = 13.6 eV (corresponding to T = B₁ / k_B ≈ 158,000 K), recombination was delayed until the plasma cooled to T_∗ ≈ 3,000 K (k_BT_∗ ≈ 0.26 eV). This delay occurs because the tiny baryon-to-photon ratio (η_B ≈ 6.1 × 10⁻¹⁰) means that even at T = 3,000 K, the high-energy Wien exponential tail exp(−*B*₁ / k_BT) of the Planck photon distribution still contains more than one ionizing photon with *h*ν > 13.6 eV per hydrogen atom, as quantified by the Saha ionization equation:

(1 − X_e) / X_e² = [2 ζ(3) / π²] η_B (2π*k*_BT / [m_ec²])^3/2 exp(B₁ / [k_BT])

where X_e = n_e / n_H is the free electron ionization fraction. Moreover, direct recombination to the n = 1 hydrogen ground state emits a 13.6 eV Lyman-continuum photon that ionizes a neighboring H atom; net recombination therefore proceeded via capture into excited states (n ≥ 2) followed by either the slow two-photon decay of the metastable 2*s* state (rate Λ_2s→1s = 8.22 s⁻¹) or cosmological redshifting of Lyman-α (2*p* → 1*s*) photons out of resonance, as modeled by Peebles (1968) and Zel'dovich, Kurt, and Sunyaev (1968). As X_e plummeted from ~1 to ~10⁻⁴ between z ≈ 1,300 and z ≈ 1,000, the optical depth τ(z) = ∫ n_e σ_T c dt dropped through unity at z_∗ = 1089.80, defining the surface of last scattering with a finite redshift thickness Δ*z* ≈ 80 (corresponding to ~40,000 years).

Multipole Expansion and Baryon Acoustic Oscillations (BAO)

To analyze the spatial distribution of temperature fluctuations Δ*T*(θ, φ) = T(θ, φ) − T₀ across the celestial sphere, cosmologists expand the temperature contrast field in spherical harmonics Y_ℓm(θ, φ):

Δ*T*(θ, φ) / T₀ = ∑_ℓ=1^∞ ∑_m=−ℓ^ℓ a_ℓm Y_ℓm(θ, φ), with C_ℓ = [1 / (2ℓ + 1)] ∑_m=−ℓ^ℓ |*a*_ℓm|²

where the multipole moment ℓ corresponds to an angular scale on the sky of θ ≈ 180° / ℓ, and D_ℓ = [ℓ(ℓ + 1) / 2π] C_ℓ T₀² (in μK²) is the angular power spectrum.

1. The Dipole (ℓ = 1, θ = 180°): The largest anisotropy is a kinematic Doppler dipole of amplitude ΔT_dipole = 3.362 mK (ΔT/T₀ = 1.23 × 10⁻³), caused by the peculiar velocity of the Solar System moving at 369.82 ± 0.11 km s⁻¹ relative to the CMB rest frame toward the constellation Crater/Leo.
2. The Sachs–Wolfe Plateau (2 ≤ ℓ ≲ 30, θ > 6°): Angular scales larger than the particle horizon at recombination (~1°–2°) could not have been modified by causal fluid physics before decoupling. Fluctuations on these scales preserve the nearly scale-invariant primordial curvature perturbations (spectral index n_s = 0.9665 ± 0.0038) generated by quantum fluctuations during cosmic inflation. As derived by Rainer Sachs and Arthur Wolfe (1967), photons climbing out of primordial gravitational potential wells Φ lose gravitational redshift energy, resulting in a net temperature fluctuation Δ*T* / T₀ = (1/3) Φ / c² (meaning cold spots on large angular scales correspond to overdense gravitational wells).
3. Baryon Acoustic Peaks (100 ≲ ℓ ≲ 1,000): On sub-horizon scales, the coupled photon–baryon fluid underwent baryon acoustic oscillations (BAO): dark matter gravitational wells pulled the baryon–photon fluid inward (compression), while massive radiation pressure (P_γ = [1/3]ρ_γc²) pushed it outward (rarefaction), propagating acoustic sound waves at speed c_s = c / √[3(1 + 3ρ_b / 4ρ_γ)] ≈ c / √3. At decoupling (t_∗ = 378 kyr), these sound waves froze in place at the comoving sound horizon r_s(z_∗) = 144.43 ± 0.26 Mpc.

• The first acoustic peak at ℓ₁ ≈ 220 (angular scale θ ≈ 0.8°–1.0°, amplitude ~5,700 μK²) represents modes that completed exactly one maximal compression into dark matter potential wells by recombination. Because r_s acts as a known physical standard ruler at z_∗ = 1089.8, its observed angular size directly measures the spatial geometry of the universe, proving that space is Euclidean and flat (Ω_k = 0.0007 ± 0.0019).
• The second acoustic peak at ℓ₂ ≈ 540 represents modes at maximal rarefaction (expansion), and the third peak at ℓ₃ ≈ 810 represents a second compression. Adding baryonic matter increases the inertial mass of the fluid without adding radiation pressure, enhancing odd (compression) peaks relative to even (rarefaction) peaks and pinning down the physical baryon density Ω_bh² = 0.02242.

1. Silk (Diffusion) Damping Tail (ℓ > 1,000, θ < 0.15°): During the finite duration of recombination, photons random-walked across a mean diffusion length λ_D ≈ 10 Mpc (comoving), erasing temperature fluctuations on small scales via photon viscosity and thermal conduction (Silk damping, 1968).

Beyond primary temperature fluctuations set at z_∗ ≈ 1,090, CMB photons acquire linear polarization and secondary anisotropies as they traverse the evolving large-scale structure of the universe over the next 13.8 billion years:

• E-Mode and B-Mode Polarization: Anisotropic quadrupole temperature gradients around free electrons during decoupling (and later during cosmic reionization at z_re ≈ 7.7 by the first stars) generate curl-free E-mode linear polarization via Thomson scattering (first detected by the DASI interferometer in 2002). Primordial gravitational waves (tensor perturbations) from cosmic inflation are predicted to generate a divergence-free, curl-like B-mode polarization pattern at large angular scales, parameterized by the tensor-to-scalar ratio r (currently constrained by BICEP/Keck and Planck to r < 0.036).
• Sunyaev–Zel'dovich (SZ) Effect: When CMB photons pass through the hot intracluster medium (T_e ≈ 10⁷–10⁸ K) of massive galaxy clusters, ~1% of the photons undergo inverse Compton scattering off high-energy thermal electrons, boosting photon energies and shifting the CMB spectrum upward in frequency—creating a temperature decrement below 217 GHz and an increment above 217 GHz (thermal SZ effect), independent of cluster redshift.
• Gravitational Lensing and the Integrated Sachs–Wolfe (ISW) Effect: Intervening dark matter filaments deflect CMB photon trajectories by ~2.5 arcminutes rms, while the late-time decay of gravitational potentials during dark-energy-dominated expansion (z < 1) imprints the late Integrated Sachs–Wolfe effect on large angular scales.

Related concepts include Dark Matter and Hertzsprung–Russell Diagram.

If the universe is expanding, where is the "surface of last scattering" located relative to Earth?

The Big Bang and recombination occurred everywhere in space simultaneously, not at a single point in an pre-existing void. Because CMB photons have been traveling at the speed of light c for 13.8 billion years since t_∗ = 378,000 years, the surface of last scattering that we observe today is a spherical shell centered on the observer in every direction, located at a present comoving proper distance (particle horizon radius) of roughly 45.7 billion light-years (14.0 Gpc). An observer in a distant galaxy 10 billion light-years away also sees a spherical surface of last scattering centered on their own position at z_∗ = 1089.8.

What is the "CMB Cold Spot" and does it violate standard cosmology?

The CMB Cold Spot is a large region in the southern Galactic hemisphere (constellation Eridanus, angular diameter ~5°–10°) identified by WMAP and confirmed by Planck where the temperature is roughly 70–150 μK colder than the average CMB sky (compared to typical rms fluctuations of ~18 μK on 5° scales). Rather than requiring exotic physics, statistical analyses of Planck data alongside galaxy redshift surveys indicate that the Cold Spot is consistent with a ~3σ Gaussian random tail fluctuation in the primordial density field, moderately amplified (~20 μK) by the late Integrated Sachs–Wolfe effect from the intervening Eridanus Supervoid at z ≈ 0.15–0.25.

1. Penzias, A. A., & Wilson, R. W. (1965). "A Measurement of Excess Antenna Temperature at 4080 Mc/s." The Astrophysical Journal, 142, 419–421. doi:10.1086/148307
2. Dicke, R. H., Peebles, P. J. E., Roll, P. G., & Wilkinson, D. T. (1965). "Cosmic Black-Body Radiation." The Astrophysical Journal, 142, 414–419. doi:10.1086/148306
3. Fixsen, D. J., Cheng, E. S., Gales, J. M., Mather, J. C., Shafer, R. A., & Wright, E. L. (1996). "The Cosmic Microwave Background Spectrum from the Full COBE FIRAS Data Set." The Astrophysical Journal, 473(2), 576–587. doi:10.1086/178173
4. Planck Collaboration, Aghanim, N., et al. (2020). "Planck 2018 results. VI. Cosmological parameters." Astronomy & Astrophysics, 641, A6. doi:10.1051/0004-6361/201833910