Structure Formation and the Transfer Function

1. Linear Perturbation Theory

The evolution of a matter density perturbation $\delta = \delta\rho/\bar\rho$ in the Newtonian limit (sub-horizon scales) is governed by

$$\ddot\delta + 2H\dot\delta - \frac{3}{2}H^2\Omega_m\,\delta = 0\,.$$

This equation has a growing mode $D_+(t)$ and a decaying mode$D_-(t)$. During matter domination ($\Omega_m = 1$),$D_+ \propto a$ and $D_- \propto a^{-3/2}$. During radiation domination, sub-horizon perturbations in the dark matter grow only logarithmically:

$$\delta_{\rm DM}(a) \propto \ln(a)\quad (a < a_{\rm eq})\,.$$

This suppressed growth during radiation domination is the origin of the bend in the matter power spectrum at the scale of matter-radiation equality.

2. The Growth Factor

In a $\Lambda$CDM cosmology, the linear growth factor is well approximated by

$$D_+(a) = \frac{5\Omega_m}{2}\,H(a)\int_0^a \frac{da'}{[a'H(a')]^3}\,.$$

The growth rate parameter, important for redshift-space distortions, is

$$f \equiv \frac{d\ln D_+}{d\ln a} \approx \Omega_m(a)^{0.55}\,,$$

where the exponent 0.55 is an excellent approximation for $\Lambda$CDM (it would differ for modified gravity theories, making $f$ a key test of general relativity on cosmological scales).

3. The Matter Power Spectrum

The present-day matter power spectrum is related to the primordial spectrum by

$$P(k, z) = A\,k^{n_s}\,T^2(k)\,D_+^2(z)\,,$$

where $A$ is the primordial amplitude, $n_s \approx 0.965$ is the spectral index, $T(k)$ is the transfer function, and $D_+(z)$ is the linear growth factor normalized to unity today. The transfer function encodes all the microphysics occurring between horizon entry and the present:

$$T(k) \approx \frac{\ln(1+2.34q)}{2.34q}\left[1+3.89q+(16.1q)^2+(5.46q)^3+(6.71q)^4\right]^{-1/4}\,,$$

where $q = k/(\Omega_m h^2\;\text{Mpc}^{-1})$ is the BBKS fitting formula. The key scale is the equality wavenumber:

$$k_{\rm eq} = a_{\rm eq}\,H_{\rm eq} = \sqrt{2\Omega_m H_0^2/a_{\rm eq}} \approx 0.01\;\text{Mpc}^{-1}\,.$$

For $k \ll k_{\rm eq}$, $T(k) \approx 1$ and $P(k) \propto k^{n_s}$. For $k \gg k_{\rm eq}$, $T(k) \propto k^{-2}\ln k$ and$P(k) \propto k^{n_s-4}\ln^2 k$, reflecting the logarithmic growth during radiation domination.

4. Halo Mass Function: Press-Schechter Theory

The abundance of dark matter halos of mass $M$ is predicted by the Press-Schechter formalism. The variance of the density field smoothed on scale $R$ (enclosing mass$M = 4\pi\bar\rho R^3/3$) is

$$\sigma^2(R) = \frac{1}{2\pi^2}\int_0^\infty k^2\,P(k)\,|W(kR)|^2\,dk\,,$$

where $W(kR) = 3[\sin(kR)-kR\cos(kR)]/(kR)^3$ is the top-hat window function. The comoving number density of halos in the mass range $[M, M+dM]$ is

$$\frac{dn}{dM} = \sqrt{\frac{2}{\pi}}\frac{\bar\rho}{M^2}\frac{\delta_c}{\sigma}\left|\frac{d\ln\sigma}{d\ln M}\right|\exp\!\left(-\frac{\delta_c^2}{2\sigma^2}\right)\,,$$

where $\delta_c \approx 1.686$ is the critical linear overdensity for spherical collapse. The exponential suppression at high masses means that massive clusters are extremely sensitive probes of $\sigma_8$ and $\Omega_m$.

5. Axion Dark Matter and Small-Scale Structure

Ultra-light axions exhibit a quantum pressure (from the de Broglie wavelength of the condensate) that suppresses structure formation below the Jeans scale:

$$k_J \sim \left(m_a\,H_0\right)^{1/4}\left(\frac{\Omega_m}{a}\right)^{1/4}\,.$$

For the QCD axion ($m_a \sim \mu$eV), the Jeans length is astrophysically negligible ($\sim$ pc). However, “fuzzy dark matter” with$m_a \sim 10^{-22}$ eV has a de Broglie wavelength of galactic scale:

$$\lambda_{\rm dB} = \frac{2\pi}{m_a v} \sim 1\;\text{kpc}\left(\frac{10^{-22}\;\text{eV}}{m_a}\right)\left(\frac{200\;\text{km/s}}{v}\right)\,.$$

The transfer function for fuzzy DM is suppressed relative to CDM below the Jeans scale:

$$T_{\rm FDM}(k) = T_{\rm CDM}(k)\,\cos^3\!\left(\frac{x_J^3}{x_J^3 + x^3}\right)\,,\quad x = 1.61\,m_{22}^{1/18}\,\frac{k}{k_{J,\rm eq}}\,,$$

where $m_{22} = m_a/(10^{-22}\;\text{eV})$. This suppression has observable consequences:

6. Beyond Linear Theory

When $\delta \gtrsim 1$, perturbation theory breaks down and numerical N-body simulations become essential. The nonlinear power spectrum is related to the linear one via the halo model:

$$P_{\rm NL}(k) = P^{\rm 1h}(k) + P^{\rm 2h}(k)\,,$$

where the one-halo term $P^{\rm 1h}$ dominates on small scales (intra-halo correlations) and the two-halo term $P^{\rm 2h}$ dominates on large scales (inter-halo correlations). The characteristic scale marking the transition from linear to nonlinear evolution is defined by

$$\sigma^2(R_{\rm NL}) = 1\,,\qquad R_{\rm NL}(z=0) \approx 8\;h^{-1}\;\text{Mpc}\,,$$

which is why $\sigma_8 \equiv \sigma(8\;h^{-1}\;\text{Mpc})$ is adopted as a standard cosmological parameter. Current measurements give$\sigma_8 = 0.811 \pm 0.006$ (Planck 2018).

7. Baryon Acoustic Oscillations

The sound waves frozen into the matter distribution at recombination leave a characteristic bump in the correlation function at the sound horizon scale:

$$\xi(r) = \frac{1}{2\pi^2}\int_0^\infty k^2\,P(k)\,\frac{\sin(kr)}{kr}\,dk\,,$$

with a peak at $r = r_s(z_d) \approx 147$ Mpc. In Fourier space, the BAO appear as oscillations in $P(k)$ with period $\Delta k \approx \pi/r_s$. Galaxy surveys (SDSS, BOSS, DESI) measure this scale both transversely and radially, providing geometric distance constraints:

$$D_V(z) = \left[\frac{cz}{H(z)}\,d_A^2(z)\right]^{1/3}\,,$$

where $D_V$ is the volume-averaged distance. BAO measurements at multiple redshifts trace the expansion history and provide some of the tightest constraints on dark energy.

8. The S8 Tension

Weak gravitational lensing surveys measure the amplitude of structure growth through the parameter combination

$$S_8 \equiv \sigma_8\sqrt{\Omega_m/0.3}\,.$$

Several lensing surveys (KiDS, DES, HSC) report $S_8 \approx 0.76\text{--}0.78$, in $2\text{--}3\sigma$ tension with the Planck CMB inference$S_8 = 0.834 \pm 0.016$. If confirmed, this discrepancy could point to new physics suppressing structure growth — massive neutrinos, decaying dark matter, dark energy with $w \neq -1$, or modified gravity. Alternatively, systematic effects in photometric redshifts, intrinsic alignments, or baryonic feedback may resolve the tension.