The problem is related to the Cannonball problem in Wikipedia. In more detail,
the book Unsolved Problems in Number Theory Second Edition by Richard Guy in section D3 starting on page 223:
The case "square pyramid = square" is Lucas's problem. Is $\, x=24,\ y=70 \,$
the only nontrivial solution of the diophantine equation
$$ y^2 = x(x+1)(2x+1)/6? $$
This was solved affirmatively by Watson, using elliptic functions, and by
Ljunggren, using a Pell equation in a quadratic field. Mordell asked if there was an elementary proof, and affirmative answers have been given by Ma,
by Xu & Cao, by Anglin and by Pinter.
The same equation in disguise is to ask if $\, (48,140) \,$ is the unique
nontrivial solution to the case "square = tetrahedron", since the previous
equation may be written $$ (2y)^2 = 2x(2x+1)(2x+2)/6, $$
though, as Peter Montgomery notes, this doesn't eliminate the possibility
of an odd square. A more modern treatment is to put $\, 12x = X-6,\ 72y = Y \,$
and note that $\, Y^2 = X^3 - 36X \,$ is curve $576H2$ in John Cremona's
tables. The point $\, (12,36) \,$ (which gives an odd square) serves as a
generator. There's an infinity of rational solutions, but the only nontrivial
integer solution to the original equation is given by the point
$\, (294,5040). \,$
The modern treatment that Richard Guy is writing about here uses the elliptic curve with Cremona label $\texttt{"576h2"}$ or LMFDB label 576.c3 which has two $2$-torsion point generators $(0,0)$ and $(6,0)$. The point $(0,0)$ added to $(-3,9)$ is $(12,36)$ where the latter two points are infinite order generators. The other integral points are $\ (-6,0), (-2,8), (18,72), (294,5040). \ $ This is directly related to the congruent number problem. In particular, finding rational right triangles with area $6$. Read about this in, for example, Karl Rubin on Right triangles and elliptic curves.
Another method is given in my work at
Weierstrass Elliptic Function Polynomials.
In brief, I construct four polynomial sequences $\, (w_n,x_n,y_n,z_n) \,$ such that
$\, (x_n/x_0)^2 - (y_n/y_0)^2 = (x_1^2-y_1^2) (w_n/w_1)^2, \,$
$\, (z_n/z_0)^2 - (y_n/y_0)^2 = (z_1^2-y_1^2) (w_n/w_1)^2, \,$
$\, (z_n/z_0)^2 - (z_n/z_0)^2 = (z_1^2-x_1^2) (w_n/w_1)^2 \,$ for all integer $\, n.$ They satisfy recursions such as
$\, w_{n+1}w_{n-1}x_0^2 = w_n^2 x_1^2 - x_n^2 w_1^2, \,$ and $\,
x_{n+1}x_{n-1}x_0^2 = x_n^2 x_1^2 -
(x_1^2-y_1^2)(x_1^2-z_1^2)(w_n/w_1)^2 x_0^4. $
To get solutions of $\, j^2 = 1 + 24m^2, \, k^2 = 1 + 48m^2 \,$ we let
$\, x_0 = y_0 = z_0 = w_1 = x_1 = 1 \,$ and
$\, y_1 = \sqrt{2}, \, z_1 = \sqrt{3}. \,$ This is almost exactly what I used
in my answer to MSE question 2556314. Let $\, t_n := w_n/\sqrt{24}. \,$ The solutions are given by
$\, j = y_{2n}/x_{2n}, \, k = z_{2n}/x_{2n}, \, m = t_{2n}/x_{2n}. \,$
For $\, n=0, \,$ we have a trivial solution
$\, (1)^2 = 1 + 24(0)^2, \, (1)^2 = 1 + 48(0)^2. \,$
For $\, n=1, \,$ we have a nontrivial solution
$\, (-5)^2 = 1 + 24(-1)^2, \, (-7)^2 = 1 + 48(-1)^2. \,$
For $\, n=2, \,$ we have
$\, (-1201/1151)^2 = 1+24(70/1151)^2, \, (-1249/1151)^2 = 1+48(70/1151)^2. \, $
For all $\, n>1 \,$ the solutions are non-integer rationals.
Now the cannonball connection is that solutions of $\, y^2 \!=\! x(x\!+\!1)(2x\!+\!1)/6 \,$ come from letting
$\, x\!=\!24m^2, \, x\!+\!1\!=\!j^2, \, 2x\!+\!1\!=\!k^2, \, y \!=\! 2mjk \,$ from solutions of
$\, j^2 \!=\! 1 \!+\! 24m^2, \, k^2 \!=\! 1 \!+\! 48m^2 .\,$
Thus, if only one nontrivial integer solution of this equation exists, then the cannonball problem has only one nontrivial solution and this has been proven by several authors and methods.