People have been able to determine the approximate date of the equinox since ancient times. The Babylonians even knew (to fairly good precision) the location of the First Point of Aries (the March equinox point) on the celestial sphere, where the ecliptic plane crosses the equatorial plane. Hipparchus was able to determine that the equinox point had precessed slightly in the intervening centuries.
Alexandrian mathematicians were aware that the daily path of the shadow cast onto a plane by a point is (approximately) a hyperbola. It's not exactly a hyperbola because the Sun's declination changes slightly over the course of the day. The great geometer Apollonius studied the hyperbola, and other conic sections.
Sundials often include such declination lines, which can be used to determine the date, and / or when the Sun enters the various signs of the Zodiac. Even quite ancient dials feature declination lines.
Here's an example of declination lines on a vertical sundial plate from Carl Sabanski's The Sundial Primer, which has a wealth of information about sundials.
The declination lines are the coloured curves. (The figure-8 shape is a noon analemma, which can be used to convert sundial time to mean time).
At the equinoxes, when the Sun's declination is zero, the declination curve degenerates to a straight line.
Of course, the declination curve looks fairly straight on the days near the equinox. However, the rate of change of the declination is at its fastest near the equinox, so you can make a pretty good estimate of the equinox day by making shadow point plots for a week or two and interpolating.
Also, on the equinox, the sunrise and sunset occur at 6 o'clock (solar time).
Determining the time of the equinox to higher precision requires more sophisticated techniques. There's some information about that at How was the First Point of Aries measured in ancient times?
The precise time of the equinox is tricky to calculate because the Earth's motion is affected by the Moon.