Posted for Barbara who is having computer issues.
@Bernard has provided the canonical reason for the misalignment:
minipages are vertically centered by default. But the suggestion for
correcting the alignment, using the option [t], doesn't have the
desired effect; the second equation is now just a bit higher than
the first.
@egreg has produced the desired result, along with a differently
formulated reason for the misalignment. But there is a more compact
way to get this result as long as you understand the reason for the
misalignment and the way that (La)TeX measures the height and depth
of a math expression. The \vphantom need contain just the
elements of the "taller" expression; it's not necessary to set the
whole thing twice. And if one of the equations is taller on the top
and the other one deeper on the bottom, a minimal \vphantom can
be applied as necessary to each, to even out the height and depth.
(That is left as an exercise to the reader.)
\documentclass{article}
\begin{document}
\thispagestyle{empty}
Using a \verb+\vphantom+:
\noindent
\begin{minipage}{.5\linewidth}
\begin{equation}
\alpha (u,v)
= \frac{N!A}{n_s \sqrt{1 - \left(\frac{N!A \rho}{n_s}\right)^2}}
\label{eq:alpha}
\end{equation}
\end{minipage}%
\begin{minipage}{.5\linewidth}
\begin{equation}
d = \alpha \rho z
\vphantom{\frac{N}{\sqrt{\left(\frac{N}{n_s}^2\right)}}}
\label{eq:disparity}
\end{equation}
\end{minipage}
\bigskip
using the optional \verb+[t]+ for top alignment:
\noindent
\begin{minipage}[t]{.5\linewidth}
\begin{equation}
\alpha (u,v)
= \frac{N!A}{n_s \sqrt{1 - \left(\frac{N!A \rho}{n_s}\right)^2}}
\end{equation}
\end{minipage}%
\begin{minipage}[t]{.5\linewidth}
\begin{equation}
d = \alpha \rho z
\end{equation}
\end{minipage}
\end{document}

minipages are vertically centred by default. Try\begin{minipage}[t]– Bernard Jun 18 '20 at 17:13