Computed from JPL DE441
How long do you have to wait for a total solar eclipse to coincide with Manhattanhenge: a syzygy with a street address?
Between 330,000 and 6.3 million years, according to how strictly the alignment is drawn — and not once in the fifteen thousand years of ephemeris that can be computed at all. Every near approach follows.
ΔT is under ten minutes here, so these are real forecasts: the date, the time and the size of the bite all hold.
1.92° above the horizon · ΔT 6 min · predicted
The one to circle: deep and well framed. Rival Earth-rotation models put it between 59% and 70%.
0.40° above the horizon · ΔT 6 min · predicted
The Sun half-swallowed by New Jersey. Beautiful, barely eclipsed.
The clock has drifted by tens of minutes. The eclipse happens and it is near the alignment, but the exact coverage is soft.
1.06° above the horizon · ΔT 45 min · approximate
Half a Sun, two days past that year’s full-sun day.
2.76° above the horizon · ΔT 25 min · approximate
High in the corridor but comfortably framed.
2.63° above the horizon · ΔT 59 min · approximate
A modest bite near the top of the season.
ΔT runs to hours or days. The eclipse is real; whether Manhattan was underneath it is not knowable.
3.11° above the horizon · ΔT 10025 min · speculative
The biggest bite in the whole search, and the least real.
1.47° above the horizon · ΔT 2864 min · speculative
More than half gone, low in the canyon.
1.32° above the horizon · ΔT 7221 min · speculative
A late-June evening, the Sun barely a degree up.
0.49° above the horizon · ΔT 11487 min · speculative
A canonical half-sun day — the classic composition, with a notch.