Earth age: how long a year and a day actually are
Your age in every astronomical definition of the year, in rotations of the planet, and in distance travelled — with an account of why none of those is a single number.
A tropical year is 365.242189 days, a sidereal year 365.256363, an anomalistic year 365.259636, and the calendar’s own long-run average exactly 365.2425. None of them is wrong. They measure four different returns, and the differences between them are what the rest of this page is about. Year lengths are J2000.0 means. They drift — the mean tropical year is shrinking by about 0.53 s per century — but across 1500–2200 the drift stays far below the precision shown.
- Tropical years
- 26.7631 tropical years Seasons returned since you were born
- Rotations
- 9,801.78 rotations True 360° turns against the stars
- Turns without a sunrise
- +26.76 extra spins Rotations the orbit added, one a year
- Carried around the galaxy
- 194.25 billion km At 230 km/s, the whole time
Six definitions of a year
Each row measures a different thing coming back round. The first four are measurements of the Earth; the last two are definitions people agreed on.
| Definition of a year | Length | Your age in them |
|---|---|---|
| Tropical year Equinox to equinox, against the precessing equinox. Slightly short of a full orbit, and the only one of these that keeps the seasons where they are. | 365.242189 days — 365d 05h 48m 45s | 26.7631 years |
| Sidereal year A true 360° around the Sun, against the fixed stars. Longer than the tropical year by 20m 24s, which is the whole of what precession costs. | 365.256363 days — 365d 06h 09m 10s | 26.7621 years |
| Anomalistic year Perihelion to perihelion. Longest of these, because the orbit’s long axis itself creeps forward about 11.6″ a year and Earth must chase it. | 365.259636 days — 365d 06h 13m 53s | 26.7618 years |
| Draconic (eclipse) year The Sun’s return to the same node of the Moon’s orbit. Nearly nineteen days short of the others, and the reason eclipse seasons walk backwards through the calendar. | 346.620076 days — 346d 14h 52m 55s | 28.2009 years |
| Julian year A defined unit, not a measurement of anything. The IAU’s year for scientific use, and the one a light-year is built from. | Exactly 365.25 days — 31,557,600 s | 26.7625 years |
| Gregorian mean year What the civil calendar averages once its leap rule has run a full cycle. Exact by construction, and 27 s longer than the year it is approximating. | Exactly 365.2425 days — 146,097 days per 400 years | 26.7631 years |
And the tropical year is not one number either
The mean tropical year is defined on a fictitious mean sun. The interval actually measured between two March equinoxes is different — and different again from two June solstices, and from each other — because Earth’s orbit is an ellipse and its speed around it varies.
| Interval measured | Length in days | Why it differs |
|---|---|---|
| December solstice to December solstice | 365.242740 | The longest: Earth is near perihelion and moving at its fastest. |
| March equinox to March equinox | 365.242374 | The one the Gregorian reform was aimed at, and the one most people mean. |
| Mean tropical year | 365.242189 | Defined on the mean sun. Equal to none of the four intervals above or below it. |
| September equinox to September equinox | 365.242018 | Between the two, on the way out towards aphelion. |
| June solstice to June solstice | 365.241626 | The shortest: Earth is near aphelion and moving at its slowest. |
The spread between the longest and the shortest is about 96 seconds. Starting from the December solstice, Earth is close to perihelion and moving at its fastest, so it loses very little time by not having to cover the small arc the equinox has precessed backwards, and that year runs long. Starting near aphelion the opposite holds.
The day is not one number either
Three rotation periods, all real. Two of them differ from each other by eight milliseconds, and both differ from the third by nearly four minutes. That last gap is the only one you can see, and it is the subject of the section after next.
| Rotation measured | Period | How many since your birth |
|---|---|---|
| Mean solar days Sunrises. The day the clock on the wall keeps, and the only one you can feel. | 24h 00m 00s — 86,400 s exactly, by convention | 9,775.013 rotations |
| Stellar days The true 360°: one full turn against the ICRF, the frame fixed by distant quasars. | 23h 56m 04.0989s — 86,164.0989 s | 9,801.775 rotations |
| Sidereal days Measured against the precessing equinox, which is why this one is 8.4 ms shy of a whole turn rather than equal to it. | 23h 56m 04.0905s — 86,164.0905 s | 9,801.776 rotations |
| Turns with no sunrise in them Sidereal days minus solar days: exactly one per year, and the reason is below. | One extra turn per tropical year | +26.76 full rotations |
How fast the planet is carrying you
Three motions in three different reference frames. They do not add up to a single speed, because each is measured against something different — the axis, the Sun, the galactic centre.
| Motion | Speed | Distance since your birth | Measured against |
|---|---|---|---|
| Rotation, at your birthplace Radius of the WGS84 parallel through it: 3,977.9 km | 1,044.3 km/h | 244,982,000 km | Earth’s axis |
| Orbit around the Sun A mean: the real speed runs from 29.29 km/s at aphelion to 30.29 at perihelion. | 107,218 km/h 29.7827 km/s | 25,153,310,510 km | The Sun |
| The Sun’s orbit of the galaxy Carrying the whole Solar System, and you with it. | 828,000 km/h 230 km/s | 194,249,057,921 km | The galactic centre |
One galactic orbit takes something like 230 million years, which makes your life so far 116.4 nano-galactic years of one. Even eighty years is about a third of a millionth of a lap. That period is the least certain number on this page by a wide margin — published estimates of the galactic year range from roughly 225 to 250 million years — so treat the figure as good to about a part in ten.
Why the tropical year is shorter than the sidereal one
Returning to the same season and returning to the same place are different journeys, and the shorter one is the seasonal one.
Earth’s rotational axis wobbles like a spinning top, one full circuit every 25,772 years. That is the precession of the equinoxes, and it drags the equinox point backwards along the orbit by about 50.29 arcseconds a year. The Sun therefore reaches the equinox slightly before Earth has completed a full 360°, and the tropical year comes out 20 minutes and 24 seconds short of the sidereal one.
Over eighty years that gap accumulates to more than 27 hours, which is to say your sidereal birthday and your ordinary birthday are more than a whole day apart by the end of a long life. The Gregorian calendar chases the tropical year and not the sidereal one, because what a calendar is for is keeping the harvest in the same month.
Why the anomalistic year is longer than both
The anomalistic year measures perihelion to perihelion — closest approach to closest approach. It is the longest of the four measured years because perihelion does not stay put either: the long axis of Earth’s orbit creeps forwards, by about 11.6 arcseconds a year, so Earth has to chase a target that is running away from it rather than towards it. That costs 4 minutes and 43 seconds a year against the sidereal year, in the opposite direction to the 20 minutes precession costs.
The extra rotation nobody notices
Live one year and you will see about 365 sunrises. In that same year the Earth turned 366 times.
The missing turn is spent on the orbit. While Earth spins, it also moves about a degree a day around the Sun, so after one full 360° rotation the Sun is not yet back overhead — the planet must turn for a further 3 minutes and 56 seconds to bring it there. That is the whole difference between the 23h 56m rotation and the 24h day. Go all the way round the Sun once and those extra fractions have added up to exactly one complete turn.
So everyone has lived through one more rotation of the Earth than they have seen sunrises, per year of their life, and the ribbon at the top of this page counts them.
This works out exactly, and it is worth saying why: the sidereal day and the tropical year are both measured against the same precessing equinox, so there are precisely 366.2422 sidereal days in 365.2422 solar days. The star-referenced pair balances too — 366.2564 stellar days per 365.2564 solar days in a sidereal year. Mixing a frame with the other, which is easy to do given the names, does not balance and is where most published versions of this fact go wrong.
A note on the names
The vocabulary is genuinely treacherous. The sidereal year is measured against the fixed stars, but the sidereal day is measured against the precessing equinox, which moves. The day that corresponds to a true 360° against the stars is the stellar day, and it is 8.4 milliseconds longer than the sidereal one. Every row of the tables above says what it is measured against, because the name alone will not tell you.
Latitude and how fast that makes you
Earth turns once every 23.934 hours, so a point on the equator is carried around a circle 40,075 km in circumference — 1,674.4 km/h, faster than sound. Further from the equator the circle is smaller and the speed falls with it, to nothing at all at the poles, where you turn on the spot once a day and go nowhere.
| Latitude | Radius of that parallel | Speed | Carried in 24 hours |
|---|---|---|---|
| The equator (0.000°) | 6,378.1 km | 1,674.4 km/h | 40,185 km |
| Singapore (1.290°) | 6,376.5 km | 1,673.9 km/h | 40,175 km |
| London (51.509°) | 3,977.9 km | 1,044.3 km/h | 25,062 km |
| Helsinki (60.170°) | 3,180.7 km | 835.0 km/h | 20,040 km |
| Reykjavík (64.135°) | 2,790.0 km | 732.4 km/h | 17,578 km |
| Either pole (90.000°) | 0.0 km | 0.0 km/h | 0 km |
Earth is an oblate spheroid rather than a sphere, so this is not simply the equatorial radius times the cosine of the latitude. The parallel you actually travel around has radius N cos φ, where N is the prime vertical radius of curvature — at London that is about 0.2% more than the spherical shortcut gives, which over a lifetime is a few hundred thousand kilometres.
What Unix time forgot
Every figure on this page starts from the difference between two Unix timestamps, and Unix time has a quiet convention buried in it: it counts every day as exactly 86,400 seconds, which means it pretends the leap seconds never happened. Twenty-seven of them have been inserted since 1972.
So a person born at the start of 1996 has lived through seven of them, and is genuinely about seven SI seconds older than any number here says. It is far below the precision anything is shown to, and it is not an error that can be corrected without abandoning ordinary timestamps — but on a page about how many different lengths a year has, it would be strange not to mention that the second-count is a convention too. So is the day: 86,400 seconds is a definition, and the real mean solar day drifts either side of it and is lengthening by roughly 1.5 milliseconds a century as tidal friction hands Earth’s angular momentum to the Moon.
Related
Dividing elapsed time by other planets’ orbital periods is the planetary age page, which is the same arithmetic and, as it says there, emphatically not a calendar. How the site treats birth times, time zones and the limits of every conversion is on the about page.
Where these numbers come from
Method. Elapsed SI seconds between the birth instant and now, divided by the length of the year or rotation named in each row. Year lengths are J2000.0 values in mean solar days of 86,400 s; rotation periods are IERS values in SI seconds. Surface speed is the IERS nominal angular velocity times the radius of the WGS84 parallel of latitude at the birthplace.
Limits. Elapsed time is computed from Unix timestamps, which pretend leap seconds never happened; a birth at the start of 1996 is therefore about seven SI seconds older than any figure here says, and one in 1972 about twenty-seven. Year lengths are means that drift slowly, and the mean solar day of exactly 86,400 s is a convention rather than a measurement of the Earth. The heliocentric distance uses Earth’s mean orbital speed, though the real speed varies between about 29.29 and 30.29 km/s over the year, so it is a mean-rate estimate and not a path length. Surface speed and distance ignore altitude above the ellipsoid. The galactic year is the least certain number on the page: published estimates run from about 225 to 250 million years, so the galactic figures are good to roughly a part in ten and no better.
Sources.
- IAU Working Group on Numerical Standards for Fundamental Astronomy (NSFA)
- IERS Conventions (2010), IERS Technical Note 36
- NASA Planetary Fact Sheet — Earth
- Meeus, J. and Savoie, D. (1992), “The history of the tropical year”, Journal of the British Astronomical Association 102, 40–42
Prose last reviewed 2026-08-23.