Geometric galaxy-encounter rate (source code)

= Geometric galaxy-encounter rate
{title2=$\Gamma=\pi R^2N\langle v_{\mathrm{rel}}\rangle$}

In a dilute, homogeneous population with <number density> $n$, an effective encounter <cross section> $\Sigma=\pi R_{\rm enc}^2$ gives the rate per target <galaxy>
$$
\Gamma_{\rm enc}=n\Sigma\langle v_{\rm rel}\rangle.
$$
Here $R_{\rm enc}$ is the maximum encounter <impact parameter>, not automatically one <galaxy>'s visible <radius>. If the <cross section> depends on relative speed, average the product instead: $\Gamma_{\rm enc}=n\langle\Sigma(v_{\rm rel})v_{\rm rel}\rangle$. <Gravitational focusing> can create such speed dependence.

With independent encounters at a constant rate, a <Poisson process> gives
$$
\Pr(N_{\rm enc}\ge1)=1-e^{-\Gamma_{\rm enc}T}.
$$
The approximation $\Gamma_{\rm enc}T$ is valid only for rare encounters. This is an encounter <probability>; it is not automatically a <galaxy merger> <probability>. Passing <galaxies> can remain unbound, and a bound pair can take additional <time> to coalesce.

As a simple additional model, suppose each encounter independently produces a merger within the relevant window with a fixed conditional <probability> $p$. The <probability generating function> of the Poisson count gives
$$
\Pr(N_{\rm merge}=0)=\mathbb E[(1-p)^{N_{\rm enc}}]=e^{-p\Gamma_{\rm enc}T},
$$
so the corresponding merger <probability> is $1-e^{-p\Gamma_{\rm enc}T}$. A realistic $p$ depends on encounter geometry, relative <energy>, <angular momentum>, gas and the allowed delay. It cannot be inferred from the geometric encounter rate alone.

For an entire <volume> $V$ of identical potential targets, the total distinct encounter-pair rate is $nV\Gamma_{\rm enc}/2$, because summing the per-target rate counts each unordered pair twice. This pair correction is separate from the conditional chance of merging.