Julian Day Numbers: Astronomy's Continuous Day Count
A Julian Day (JD) is a continuous count of days, and fractions of a day, since noon Universal Time on 1 January 4713 BCE in the proleptic Julian calendar. Because it ignores months, years and calendar reforms, astronomers use it to measure the time between any two events with a single subtraction. It is unrelated to the Julian calendar except in name and in the choice of its starting point.
- Type
- Continuous day count
- Day 0 begins
- Noon, 1 January 4713 BCE (Julian) = 24 November 4714 BCE (Gregorian)
- Day begins
- At noon UT
- Midnight dates end in
- .5
- J2000.0
- JD 2,451,545.0 (1 January 2000, 12:00 TT)
- Unix epoch
- JD 2,440,587.5
Scaliger and the Julian Period
In 1583 the scholar Joseph Justus Scaliger published De emendatione temporum, a major work on historical chronology. In it he defined the Julian Period, a cycle of 7,980 years formed by multiplying three older calendar cycles. Scaliger chose 4713 BCE as year 1 of the period because it is the most recent year in which all three cycles were at their first year together, which places it before any recorded historical date.
Scaliger counted years, not days. The idea of numbering individual days from the start of the Julian Period was developed later by astronomers; John Herschel's Outlines of Astronomy (1849) did much to popularise it. Scaliger called the period Julian because it is built from Julian years. The frequently repeated story that he named it after his father, Julius Caesar Scaliger, is not supported by his own text.
| Cycle | Length | Purpose |
|---|---|---|
| Solar cycle | 28 years | Repeats the pattern of weekdays and dates in the Julian calendar |
| Metonic (lunar) cycle | 19 years | Repeats the phases of the Moon on the same dates |
| Roman indiction | 15 years | Fiscal cycle used to date documents |
| Julian Period | 28 × 19 × 15 = 7,980 years | Year 1 = 4713 BCE |
Why the day starts at noon
A Julian Day begins at noon Universal Time, not at midnight. For astronomers observing from Europe, this meant a whole night of observations fell on a single Julian Day number instead of being split across two calendar dates.
The consequence is that midnight always has a fractional part of .5. For example, the start of the civil day 1 January 2000 (00:00 UT) is JD 2,451,544.5, and noon on the same day is JD 2,451,545.0. The integer part at noon is called the Julian Day Number (JDN), so the JDN of 1 January 2000 is 2,451,545.
| Time | Julian Day |
|---|---|
| 00:00 | 2,451,544.5 |
| 06:00 | 2,451,544.75 |
| 12:00 | 2,451,545.0 |
| 18:00 | 2,451,545.25 |
Worked examples
The values below are the Julian Day at 00:00 UT on each date, which is what this site's converter displays. Add 0.5 to get the Julian Day Number of the day.
| Date (Gregorian unless noted) | JD at 00:00 UT | Note |
|---|---|---|
| 1 January 1 CE | 1,721,425.5 | Proleptic Gregorian |
| 15 October 1582 | 2,299,160.5 | First day of the Gregorian calendar |
| 17 November 1858 | 2,400,000.5 | Modified Julian Day 0 |
| 1 January 1970 | 2,440,587.5 | Unix epoch |
| 1 January 2000 | 2,451,544.5 | Half a day before J2000.0 |
| 1 January 2026 | 2,461,041.5 |
J2000.0 and astronomical epochs
Modern astronomy uses J2000.0 as its standard reference epoch. It is defined as JD 2,451,545.0 in Terrestrial Time (TT), which is 12:00 TT on 1 January 2000. Star catalogues, orbital elements and the equations of planetary theories are usually expressed relative to this moment.
Time since the epoch is often measured in Julian centuries of exactly 36,525 days: T = (JD − 2,451,545.0) / 36,525. A Julian year in this sense is exactly 365.25 days, regardless of any calendar.
A Julian Day is only meaningful together with its time scale. In the 2020s, a moment expressed as JD in Terrestrial Time is a little over a minute ahead of the same moment in Universal Time, a difference that matters for precise work such as eclipse and occultation predictions.
Why a single number is useful
Calendars are irregular: months differ in length, leap-year rules changed in 1582, and different countries adopted the Gregorian calendar at different times. A Julian Day avoids all of this. The interval between two observations is just the difference of their JD values, and periodic events such as variable-star maxima can be predicted by adding multiples of the period.
The weekday can also be found directly: (JDN + 1) mod 7 gives 0 for Sunday through 6 for Saturday. For JDN 2,451,545, the result is 6, and 1 January 2000 was indeed a Saturday.
- Variable star observers record times of maxima and minima in JD or heliocentric JD.
- Ephemerides and orbit software use JD or its variants as the internal time argument.
- Calendar converters, including this one, use a day count as the common ground between calendars.
Frequently asked questions
What is today's Julian Day?
Use the converter on this page: it shows the Julian Day for midnight UT at the start of the selected date. Add the fraction of the day elapsed since then for a specific time.
Why does a Julian Day start at noon?
So that a full night of astronomical observations in Europe falls under one day number rather than two.
What is the Julian Day of J2000.0?
J2000.0 is JD 2,451,545.0, which is 12:00 Terrestrial Time on 1 January 2000.
Is the Julian Day the same as the Julian calendar?
No. The Julian Day is a continuous count of days; it borrows only the starting point, defined in the Julian calendar, and Scaliger's choice of Julian years for his period.
What is the difference between JD and JDN?
JD is a real number that includes the fraction of the day since noon UT. The Julian Day Number is the integer JD at noon of a given day.
What is a 'Julian date' on food packaging?
That usually means an ordinal date, a day of the year from 1 to 366, which is a different thing from the astronomical Julian Day.
Sources and further reading
- Sean E. Urban and P. Kenneth Seidelmann (eds.), Explanatory Supplement to the Astronomical Almanac, 3rd ed. (University Science Books, 2013)
- Jean Meeus, Astronomical Algorithms, 2nd ed. (Willmann-Bell, 1998)
- Joseph Justus Scaliger, De emendatione temporum (Paris, 1583)
- Edward M. Reingold and Nachum Dershowitz, Calendrical Calculations (Cambridge University Press)