Skip to main content
Sandbox Physics

Redshift-Space Distortions

Is the cosmic web stretched toward us because distances are wrong, or because galaxies move? Mostly because galaxies move. A redshift mixes distance with velocity along our line of sight. Galaxies falling into large structures squash them along that line, and fast orbits inside clusters stretch each cluster into a finger pointing at us. A wrong distance scale also stretches the map, but it stretches every scale by the same factor, so fitting the pattern scale by scale tells the two apart.

Move 20,000 synthetic galaxies from true positions to redshift space, turn the line of sight, and fit growth, random motion and a distance error together.

Interactive modelRedshift-Space Distortions
Fitted growth rate ff0.72(set 0.76)0.72\allowbreak\quad(\text{set}\ 0.76)
Fitted distance stretch ss1(set 1)1\allowbreak\quad(\text{set}\ 1)
Fitted random velocity σv\sigma_v300 km s−1(set 300)300\ \mathrm{km\,s^{-1}}\allowbreak\quad(\text{set}\ 300)
Same fit with distances assumed right, s=1s=1f=0.72, σv=300 km s−1{f=0.72},\allowbreak\ {\sigma_v=300\ \mathrm{km\,s^{-1}}}

What the redshift map measures

Measured in your browser from the synthetic catalogue; no survey data are shown.

Monopole and quadrupole over real-space power P0(k)/P(k), P2(k)/P(k)P_0(k)/P(k),\ P_2(k)/P(k)
Pair correlation across and along the line of sight ξ(r⊥,r∥)\xi(r_\perp,r_\parallel)

Teal: monopoleAmber: quadrupolePoints: measured; solid: joint fitGray dashed: linear Kaiser for the set growth rateWhite dashed contours: same galaxies in real space

How the measurement works

Each mode of the redshift-space power is divided by the real-space power of the same galaxies in narrow bins of scale and angle. Only a simulation allows this; it removes most of the sampling scatter of one small box. A survey has no real-space copy, so its errors are much larger.

Read growth from the shape of a redshift map

BackgroundA galaxy survey turns each redshift into a distance with Hubble’s law. Galaxies also move, and the part of their velocity along our line of sight adds to the redshift, so the map is drawn in redshift space, not in real space.

Why it mattersHow fast galaxies fall into structures depends on how strongly gravity pulls, so the distortion measures the growth rate of structure. The same map is also distorted if the distance scale is wrong, and that has to be separated first.

Focus question
When a map looks stretched along the line of sight, how do you tell motion from a wrong distance?
One-sentence intuition
Coherent infall squashes large scales and random orbits stretch small ones, while a wrong distance stretches every scale by one factor.

Core mathematical model

Where a redshift puts a galaxy

s=x+(1+z) v⋅n^H(z) n^\mathbf{s}=\mathbf{x}+\frac{(1+z)\,\mathbf{v}\cdot\hat{\mathbf{n}}}{H(z)}\,\hat{\mathbf{n}}

Only the velocity along the line of sight n^\hat{\mathbf{n}} moves the galaxy, and only along that line. At redshift 0.5, 300 km s−1300\ \mathrm{km\,s^{-1}} shifts it by about 3.4 h−1Mpc3.4\,h^{-1}\mathrm{Mpc}.

Kaiser squashing on large scales

Ps(k,μ)=(b+fμ2)2P(k),β=fbP_s(k,\mu)=\bigl(b+f\mu^2\bigr)^2P(k),\qquad \beta=\frac{f}{b}

Here μ\mu is the cosine between the wavevector and the line of sight, bb the galaxy bias and ff the growth rate. Infall boosts clustering along the line of sight, which squashes structures in the map.

Fingers of God damp small scales

Ps(k,μ)=(b+fμ2)2P(k) e−k2μ2σ2,σ=(1+z) σvH(z)P_s(k,\mu)=\bigl(b+f\mu^2\bigr)^2P(k)\,e^{-k^2\mu^2\sigma^2},\qquad \sigma=\frac{(1+z)\,\sigma_v}{H(z)}

Random orbits smear each group along the line of sight. The Gaussian factor is an empirical model; it is trusted only on the large scales used in the fit.

A wrong distance scale stretches everything

kt=k1+(s2−1)μ2,μt=s μ1+(s2−1)μ2k_t=k\sqrt{1+(s^2-1)\mu^2},\qquad \mu_t=\frac{s\,\mu}{\sqrt{1+(s^2-1)\mu^2}}

If assumed distances along the line of sight are ss times the true ones, an observed mode at (k,μ)(k,\mu) is a true mode at (kt,μt)(k_t,\mu_t). This Alcock–Paczynski stretch is the same at every scale.

Common difficulties

Not every anisotropy is growth

Typical misconceptionAny squashing along the line of sight measures the growth rate.

Better mental modelA distance stretch below one also squashes the map, and random motion partly hides infall. Only a fit that frees all three can attribute the squashing to growth.

Run the experiment

  1. 01

    Watch galaxies move into redshift space

    Keep the default preset and watch the right window as the loop slides galaxies from true positions to redshift positions.

    What to observe: Long filaments across the line of sight sharpen, while compact groups grow streaks pointing at the observer.
  2. 02

    Separate the two effects

    Choose Coherent infall, then Fingers of God, and compare the correlation map.

    What to observe: Infall flattens the contours along the line of sight; random motion stretches the innermost contours along it.
  3. 03

    Fake a distortion with distances

    Choose Wrong distance: stretched, then switch to the fixed-distance readout.

    What to observe: The joint fit finds the stretch and no growth. With distances assumed right, the fit invents random motion instead.
  4. 04

    Turn the line of sight

    Drag the observer around the redshift-space window.

    What to observe: The streaks and the squashing turn with the observer, while the real-space window stays the same.

Related experiments and articles

What this page links to directly

See every article series on the content map →

Redshift-space distortions · Reading path

From a moving galaxy to a growth measurement

Articles on this path that use this experiment, followed by the other experiments they suggest.

  1. ArticleWhy do galaxy maps point at us?
  2. ArticleHow a redshift moves a galaxy
  3. ArticleMultipoles of a squashed map
  4. ArticleGrowth or a distance error?
  5. ArticleMeasure multipoles from a frozen catalogue
  6. ArticleHow Guzzo et al. measured growth from redshift distortions in 2008
Something broken, confusing, or missing on this page?Tell us
What kind of feedback?

Sent with this page’s address, its Lab ID, the site language and your device type. Please leave out personal details. Feedback is kept for up to 12 months.Privacy policy