EN ▾
UTC — the one time scale every eclipse is quoted in
UTC is the single time scale this site computes in. Eclipse times are quoted in it because an event crossing a dozen zones needs one common reference.
Coordinated Universal Time is the time scale every instant on this site is computed in before it is shown to you. Eclipse times are quoted in UTC by long convention.
Why not local time
Because an eclipse is a single event, three hours or so from end to end, crossing a dozen time zones — and no local time describes it. Two observers a few hundred kilometres apart may be in different zones, or on different sides of a daylight-saving boundary, so quoting their contact times locally can make moments a few minutes apart look an hour and a few minutes apart, and moments an hour apart look simultaneous.
One scale removes that entirely. The timings panel shows both — UTC and your device's local time — with UTC first, because it is the one that can be compared with anything published anywhere else. The eclipse lists are UTC only.
The three letters
UTC is a compromise between the English word order and the French, which is why it matches neither. It is not translated anywhere on this site: the same three letters appear in every language, so a time on a page and a time in the app can be compared without decoding a label.
Two things it is worth knowing UTC is. It is an atomic scale, not an astronomical one: it is kept within 0.9 seconds of the Earth's actual rotation by the occasional insertion of a leap second, which is why it drifts from uniform time in one-second steps rather than smoothly. And it is what GMT became — for every purpose on this site the two are the same thing, and UTC is the name to use. It never observes daylight saving.
ΔT, and why the far future is vaguer than the near
The Earth's rotation is not perfectly uniform. The difference between it and a uniform time scale is called ΔT, and it is about seventy seconds today. It matters here because the Sun and Moon are computed on the uniform scale while the ground turns underneath on the rotational one: get ΔT wrong by a minute and every track moves about 25 km east or west.
This site corrects for it, and for the past fifty years it uses the measured values rather than a formula — so for any eclipse you might actually travel to, that correction is as good as the rest of the model.
Where it has not been measured, it cannot be. ΔT is not calculable from first principles; it is observed after the fact, because it depends on tides, on the slow rebound of the crust after the last ice age, and on the motion of the Earth's core. Predicting it means extrapolating a process nobody can see the future of, and the Earth has already embarrassed one careful forecast inside twenty years: predictions made in 2004 have the Earth several seconds slower by the 2020s than it turned out to be, because it stopped slowing shortly afterwards. A century ahead the uncertainty is a minute or so; by the year 3000 it is comfortably an hour, which is fifteen degrees of longitude.
So a far-future eclipse on this site has a reliable date, a reliable duration and a reliable shape, and an indicative longitude. The shadow will do what is drawn; where it does it is a forecast of the Earth's rotation as much as a prediction of the eclipse. That is a property of the question rather than of this site — every source you can consult has the same limit, and one that does not say so is hiding it.
Related
Updated 14 August 2026.