intermediate 10 min read
Physical sciences · Topic
Hubble's Law and Cosmic Expansion
differential equations · probability theory
Hubble's Law, $v = H_0 d$, encodes the expansion of the universe in a single proportionality constant measured across millions of light-years. Originally inferred from Doppler-shifted galaxy spectra in 1929, it now anchors the entire distance ladder and the $\Lambda$CDM concordance model. A persistent tension between values of $H_0$ derived from the early and late universe has emerged as one of the deepest puzzles in modern cosmology.

The Hubble–Lemaître Law

For a galaxy at proper distance $d$, the recession velocity due to cosmic expansion is

\[v = H_0\, d\]

where $H_0$ is the Hubble constant, with current best estimates near $70\ \text{km s}^{-1}\text{Mpc}^{-1}$. Hubble’s original 1929 value was $\approx 500\ \text{km s}^{-1}\text{Mpc}^{-1}$—a factor of seven too high due to a miscalibrated distance ladder.

The law is not a “velocity through space” but rather a statement about space itself expanding: every pair of points separated by distance $d$ recedes at rate $v = H_0 d$. The relationship follows directly from homogeneity and isotropy of the universe—galaxies are not moving from a central point but rather riding an expanding metric.

In terms of the scale factor $a(t)$, the proper distance evolves as

\[d(t) = a(t)\, \chi\]

where $\chi$ is the comoving coordinate (fixed to the cosmic grid). Differentiating:

\[\dot{d} = \dot{a}\, \chi = \frac{\dot{a}}{a}\, d \equiv H(t)\, d\]

The Hubble constant is thus the present-day value $H_0 = \dot{a}_0 / a_0$, conventionally written $H_0 = 100\, h\ \text{km s}^{-1}\text{Mpc}^{-1}$ with $h \approx 0.70$.

Cosmological Redshift

Light emitted at wavelength $\lambda_\text{em}$ and observed at $\lambda_\text{obs}$ defines the redshift

\[z = \frac{\lambda_\text{obs} - \lambda_\text{em}}{\lambda_\text{em}} = \frac{\Delta\lambda}{\lambda_\text{em}}\]

For a universe with scale factor $a(t)$, photon wavelengths stretch with the metric:

\[1 + z = \frac{a_0}{a_\text{em}} = \frac{1}{a_\text{em}}\]

(using the convention $a_0 = 1$). In the local universe ($z \ll 1$), this reduces to the Doppler formula $z \approx v/c$, so

\[v \approx c z = H_0 d \quad \Rightarrow \quad d \approx \frac{cz}{H_0}\]

For larger redshifts the relationship becomes geometry-dependent. The comoving distance to redshift $z$ is

\[\chi(z) = \frac{c}{H_0} \int_0^z \frac{dz'}{E(z')}\]

where $E(z) = H(z)/H_0 = \sqrt{\Omega_m(1+z)^3 + \Omega_r(1+z)^4 + \Omega_\Lambda}$ encodes the energy content.

Quantity Symbol Typical value
Hubble constant $H_0$ $70\ \text{km s}^{-1}\text{Mpc}^{-1}$
Hubble time $t_H = 1/H_0$ $\approx 14\ \text{Gyr}$
Hubble distance $D_H = c/H_0$ $\approx 4280\ \text{Mpc}$
CMB-derived $H_0$ (Planck) $H_0^\text{CMB}$ $67.4 \pm 0.5$
Local $H_0$ (SH0ES) $H_0^\text{local}$ $73.0 \pm 1.0$

The Cosmic Distance Ladder

Measuring $H_0$ requires calibrating distances independently of redshift. The distance ladder chains multiple overlapping methods:

Rung 1 — Geometric distances. Parallax gives precise distances to nearby stars. GAIA provides sub-microarcsecond parallaxes to $\sim 3\ \text{kpc}$.

Rung 2 — Cepheid variable stars. The period-luminosity (Leavitt) law:

\[M = -a \log_{10}(P/10\ \text{days}) + b\]

with $a \approx 2.81$ in the $I$-band. Cepheids reach to $\sim 50\ \text{Mpc}$ with HST or JWST.

Rung 3 — Type Ia supernovae (SNe Ia). These “standard candles” have a luminosity-decline-rate correlation (Phillips relation):

\[M_B^\text{max} = -19.3 + \alpha\, \Delta m_{15}(B)\]

where $\Delta m_{15}$ is the $B$-band decline in 15 days. SNe Ia reach $z \sim 1.5$, providing the kinematic evidence for accelerated expansion.

Rung 4 — Hubble flow. Once $d$ is calibrated and $z$ is measured spectroscopically, $H_0 = v/d$.

The distance modulus linking apparent magnitude $m$ to absolute magnitude $M$ and luminosity distance $d_L$:

\[\mu = m - M = 5\log_{10}\!\left(\frac{d_L}{10\ \text{pc}}\right)\]

Lookback Time and Comoving Distance

The lookback time to redshift $z$ is

\[t_L(z) = \frac{1}{H_0} \int_0^z \frac{dz'}{(1+z')\, E(z')}\]

For the concordance model ($\Omega_m = 0.3$, $\Omega_\Lambda = 0.7$):

$z$ Lookback time (Gyr) Comoving distance (Mpc)
0.1 1.3 420
0.5 5.1 1890
1.0 7.9 3300
2.0 10.3 5200
1100 13.8 14,000

The proper distance at the moment of emission differs from the comoving distance:

\[d_\text{proper}(t_\text{em}) = a(t_\text{em})\, \chi = \frac{\chi}{1+z}\]

The luminosity distance $d_L = (1+z)\chi$ and angular diameter distance $d_A = \chi/(1+z)$ differ by a factor of $(1+z)^2$, a consequence of photon time dilation and aberration.

The Hubble Tension

Two classes of measurement give inconsistent $H_0$ values at $\sim 5\sigma$ significance:

Early-universe (CMB) measurements use the acoustic peak positions in the CMB power spectrum—a standard ruler with angular scale $\theta_s = r_s / d_A$, where the sound horizon $r_s \approx 147\ \text{Mpc}$ is precisely calculated from pre-recombination physics. Planck 2018 gives $H_0 = 67.4 \pm 0.5\ \text{km s}^{-1}\text{Mpc}^{-1}$.

Late-universe (distance ladder) measurements use Cepheids anchored to geometric distances plus SNe Ia. The SH0ES collaboration gives $H_0 = 73.0 \pm 1.0\ \text{km s}^{-1}\text{Mpc}^{-1}$.

Proposed resolutions include:

  • Early dark energy modifying $r_s$ before recombination
  • Interacting dark energy with nonstandard equation of state
  • Systematic errors in Cepheid metallicity corrections or SNe Ia calibration
  • New physics at or after recombination

The tension persists across independent late-time probes (TRGB, megamasers, surface brightness fluctuations), suggesting a genuine cosmological discrepancy rather than a measurement artifact.

Statistical Estimation of $H_0$

Estimating $H_0$ from a catalog of galaxies requires propagating distance uncertainties. If each galaxy $i$ has observed recession velocity $v_i$ and distance estimate $d_i$ with uncertainty $\sigma_i$:

\[\hat{H}_0 = \frac{\sum_i v_i d_i / \sigma_i^2}{\sum_i d_i^2 / \sigma_i^2}\]

A full Bayesian treatment models the peculiar velocity contribution. Galaxies deviate from pure Hubble flow by peculiar velocities $v_\text{pec} \sim 300\ \text{km/s}$ due to local gravitational structure. The likelihood for galaxy $i$ is

\[\mathcal{L}_i(H_0) = \int P(v_\text{pec}) \, \mathcal{N}\!\left(v_i - H_0 d_i - v_\text{pec} \mid 0, \sigma_v^2\right) dv_\text{pec}\]

Gravitational wave standard sirens offer a completely independent route: binary mergers provide absolute distance from GW amplitude, and an electromagnetic counterpart provides $z$. The single event GW170817 gave $H_0 = 70^{+12}_{-8}\ \text{km s}^{-1}\text{Mpc}^{-1}$, with precision improving as the event catalog grows.