Like a wobbling top, the orientation of the Earth's axis is slowly but continuously changing, tracing out a conical shape in a cycle of approximately 25,765 years (the so called Great or Platonic year, and also the determining factor in the length of an astrological age). This movement is caused by the gravitational forces of the Sun and the Moon, and to a lesser extent other bodies, on the equatorial bulge of the spinning Earth. The term "precession" typically refers only to this largest periodic motion; other changes in the alignment of Earth's axis — nutation and polar motion — are very much smaller in magnitude.
In ancient times the precession of the equinox referred to the motion of the equinox relative to the background stars in the zodiac; this is equivalent to the modern understanding. It acted as a method of keeping time in the Great year.
Historically, Hipparchus is credited with discovering that the positions of the equinoxes move westward along the ecliptic compared to the fixed stars on the celestial sphere. The exact dates of his life are not known, but astronomical observations attributed to him date from 147 BC to 127 BC and were described in his writings, none of which are known to survive.
The precession of the Earth's axis has a number of observable effects. Firstly, the positions of the north and south celestial poles appear to move in circles against the (assumed fixed) backdrop of stars, completing one circuit in approximately 25,700 years. Thus, while today the star Polaris lies approximately at the north celestial pole, this will change over time, and other stars will become the "north star". As the celestial poles shift, there is a corresponding gradual shift in the apparent orientation of the whole star field, as viewed from a particular position on Earth.
Secondly, the position of the Earth in its orbit around the Sun at the solstices, equinoxes, or other time defined relative to the seasons, slowly changes. For example, suppose that the Earth's orbital position is marked at the summer solstice, when the Earth's axial tilt is pointing directly towards the Sun. One full orbit later, when the Sun has returned to the same apparent position relative to the background stars, the Earth's axial tilt is not now directly towards the Sun: because of the effects of precession, it is a little way "beyond" this. In other words, the solstice occurred a little earlier in the orbit. Thus, the tropical year, measuring the cycle of seasons (for example, the time from solstice to solstice, or equinox to equinox), is about 20 minutes shorter than the sidereal year, which is measured by the Sun's apparent position relative to the stars. Note that 20 minutes per year is approximately equivalent to one year per 25,700 years, so after one full cycle of 25,700 years the positions of the seasons relative to the orbit are "back where they started". (In actuality, other effects also slowly change the shape and orientation of the Earth's orbit, and these, in combination with precession, create various cycles of differing periods; see also Milankovitch cycles. The magnitude of the Earth's tilt, as opposed to merely its orientation, also changes slowly over time, but this effect is not attributed directly to precession.)
For identical reasons, the apparent position of the Sun relative to the backdrop of the stars at some seasonally fixed time, say the vernal equinox, slowly regresses a full 360° through all twelve traditional constellations of the zodiac, at the rate of about 50.3 seconds of arc per year (approximately 360 degrees divided by 25,700), or 1 degree every 71.6 years.
Virtually all Hipparchus' writings are lost, including his work on precession. They are mentioned by Ptolemy, who explains precession as the rotation of the celestial sphere around a motionless Earth. It is reasonable to assume that Hipparchus, like Ptolemy, thought of precession in geocentric terms as a motion of the heavens.
Instead of proceeding through the entire sequence of the zodiac, the equinoxes "trepidated" back and forth over an arc of 8°. The theory of trepidation is presented by Theon as an alternative to precession.
In the Middle Ages, Islamic and Latin Christian astronomers treated "trepidation" as a motion of the fixed stars to be added to precession. This theory is commonly attributed to the Arab astronomer Thabit ibn Qurra, but the attribution has been contested in modern times. Nicolaus Copernicus published a different account of trepidation in De revolutionibus orbium coelestium (1543). This work makes the first definite reference to precession as the result of a motion of the Earth's axis. Copernicus characterized precession as the third motion of the earth.
Over a century later precession was explained in Isaac Newton's Philosophiae Naturalis Principia Mathematica (1687) to be a consequence of gravitation (Evans 1998, p. 246). However, Newton's original precession equations did not work and were revised considerably by Jean le Rond d'Alembert and subsequent scientists.
Similar claims have been made that precession was known in Ancient Egypt prior to the time of Hipparchus, but these remain controversial. Some buildings in the Karnak temple complex, for instance, were allegedly oriented towards the point on the horizon where certain stars rose or set at key times of the year. A few centuries later, when precession made the orientations obsolete, the temples would be rebuilt. Note however that the observation that a stellar alignment has grown wrong does not necessarily mean that the Egyptians understood that the stars moved across the sky at the rate of about one degree per 72 years. Nonetheless, they kept accurate calendars and if they recorded the date of the temple reconstructions it would be a fairly simple matter to plot the rough precession rate. The Dendera Zodiac, a star-map from the Hathor temple at Dendera from a late (Ptolemaic) age, supposedly records precession of the equinoxes (Tompkins 1971). In any case, if the ancient Egyptians knew of precession, their knowledge is not recorded in surviving astronomical texts.
Michael Rice wrote in his Egypt's Legacy, "Whether or not the ancients knew of the mechanics of the Precession before its definition by Hipparchos the Bithynian in the second century BC is uncertain, but as dedicated watchers of the night sky they could not fail to be aware of its effects." (p. 128) Rice believes that "the Precession is fundamental to an understanding of what powered the development of Egypt" (p. 10), to the extent that "in a sense Egypt as a nation-state and the king of Egypt as a living god are the products of the realisation by the Egyptians of the astronomical changes effected by the immense apparent movement of the heavenly bodies which the Precession implies." (p. 56) Following Carl Gustav Jung, Rice says that "the evidence that the most refined astronomical observation was practised in Egypt in the third millennium BC (and probably even before that date) is clear from the precision with which the Pyramids at Giza are aligned to the cardinal points, a precision which could only have been achieved by their alignment with the stars. This fact alone makes Jung's belief in the Egyptians' knowledge of the Precession a good deal less speculative than once it seemed." (p. 31) The Egyptians also, says Rice, were "to alter the orientation of a temple when the star on whose position it had originally been set moved its position as a consequence of the Precession, something which seems to have happened several times during the New Kingdom." (p. 170) see also Royal Arch and the Precession of the Equinoxes
The notion that an ancient Egyptian priestly elite tracked the precessional cycle over many thousands of years plays a central role in the theories expounded by Robert Bauval and Graham Hancock in their 1996 book Keeper of Genesis. The authors claim that the ancient Egyptians' monumental building projects functioned as a map of the heavens, and that associated rituals were an elaborate earthly acting-out of celestial events. In particular, the rituals symbolised the "turning back" of the precessional cycle to a remote ancestral time known as Zep Tepi ("first time") which, the authors calculate, dates to around 10,500 BC.
The former professor of the history of science at MIT, Giorgio de Santillana, argues in his book, Hamlet's Mill, that many ancient cultures may have known of the slow movement of the stars across the sky; the observable result of the precession of the equinox. This 700 page book, co-authored by Hertha von Dechend, makes reference to approximately 200 myths from over 30 ancient cultures that hinted at the motion of the heavens, some of which are thought to date to the neolithic period.
Identifying alignments of monuments with solar, lunar, and stellar phenomena is a major part of archaeoastronomy. Stonehenge is the most famous of many megalithic structures that indicate the direction of celestial objects at rising or setting. Precession complicates the attempt to find stellar alignments, especially for very old sites. Many archaeological sites cannot be dated exactly, making it difficult or impossible to know whether a proposed alignment would have worked when the site was founded.
Why did Hipparchus need a lunar eclipse to measure the position of a star? The equinoctial points are not marked in the sky, so he needed the Moon as a reference point. Hipparchus had already developed a way to calculate the longitude of the Sun at any moment. A lunar eclipse happens during Full moon, when the Moon is in opposition. At the midpoint of the eclipse, the Moon is precisely 180° from the Sun. Hipparchus is thought to have measured the longitudinal arc separating Spica from the Moon. To this value, he added the calculated longitude of the Sun, plus 180° for the longitude of the Moon. He did the same procedure with Timocharis' data (Evans 1998, p. 251). Observations like these eclipses, incidentally, are the main source of data about when Hipparchus worked, since other biographical information about him is minimal. The lunar eclipses he observed, for instance, took place on April 21, 146 BC, and March 21, 135 BC (Toomer 1984, p. 135 n. 14).
Hipparchus also studied precession in On the Length of the Year. Two kinds of year are relevant to understanding his work. The tropical year is the length of time that the Sun, as viewed from the Earth, takes to return to the same position along the ecliptic (its path among the stars on the celestial sphere). The sidereal year is the length of time that the Sun takes to return to the same position with respect to the stars of the celestial sphere. Precession causes the stars to change their longitude slightly each year, so the sidereal year is longer than the tropical year. Using observations of the equinoxes and solstices, Hipparchus found that the length of the tropical year was 365+1/4−1/300 days, or 365.24667 days (Evans 1998, p. 209). Comparing this with the length of the sidereal year, he calculated that the rate of precession was not less than 1° in a century. From this information, it is possible to calculate that his value for the sidereal year was 365+1/4+1/144 days (Toomer 1978, p. 218). By giving a minimum rate he may have been allowing for errors in observation.
To approximate his tropical year Hipparchus created his own lunisolar calendar by modifying those of Meton and Callippus in On Intercalary Months and Days (now lost), as described by Ptolemy in the Almagest III.1 (Toomer 1984, p. 139). The Babylonian calendar used a cycle of 235 lunar months in 19 years since 499 BC (with only three exceptions before 380 BC), but it did not use a specified number of days. The Metonic cycle (432 BC) assigned 6,940 days to these 19 years producing an average year of 365+1/4+1/76 or 365.26316 days. The Callippic cycle (330 BC) dropped one day from four Metonic cycles (76 years) for an average year of 365+1/4 or 365.25 days. Hipparchus dropped one more day from four Callipic cycles (304 years), creating the Hipparchic cycle with an average year of 365+1/4−1/304 or 365.24671 days, which was close to his tropical year of 365+1/4−1/300 or 365.24667 days. The three Greek cycles were never used to regulate any civil calendar—they only appear in the Almagest in an astronomical context.
We find Hipparchus mathematical signatures in the Antikythera Mechanism, an ancient astronomical computer of the 2nd CBC. The mechanism is based on a solar year, the Metonic Cycle, which is the period the Moon reappears in the same star in the sky with the same phase (full Moon appears at the same position in the sky approximately in 19 years), the Callipic cycle (which is four Metonic cycles and more accurate), the Saros cycle and the Exeligmos cycles (three Saros cycles for the accurate eclipse prediction). The study of the Antikythera Mechanism proves that the ancients have been using very accurate calendars based on all the aspects of solar and lunar motion in the sky. In fact the Lunar Mechanism which is part of the Antikythera Mechanism depicts the motion of the Moon and its phase, for a given time, using a train of four gears with a pin and slot device which gives a variable lunar velocity that is very close to the second law of Kepler, i.e. it takes into account the fast motion of the Moon at perigee and slower motion at apogee. This discovery proves that Hipparchus mathematics were much more advanced than Ptolemy describes in his books, as it is evident that he developed a good approximation of Kepler΄s second law.
Mithraism is now recognized as having pronounced astrological elements, but the details are debated. One scholar of Mithraism, David Ulansey, has interpreted Mithras (Mithras Sol Invictus - the unconquerable sun) as a second sun or star that is responsible for precession. He suggests the cult may have been inspired by Hipparchus' discovery of precession. Part of his analysis is based on the tauroctony an image of Mithras sacrificing a bull, found in most of the temples. According to Ulansey, the tauroctony is a star chart. Mithras is a second sun or hyper-cosmic sun and or a constellation Perseus, and the bull is Taurus, a constellation of the zodiac. In an earlier astrological age, the vernal equinox had taken place when the Sun was in Taurus. The tauroctony, by this reasoning, commemorated Mithras-Perseus ending the "Age of Taurus" (about 2000 BC based on the Vernal Equinox - or about 11,500 BC based on the Autumnal Equinox).
The iconography also contains two torch bearing boys (Cautes and Cautopetes) on each side of the zodiac. Ulansey, and Walter Cruttenden in his book Lost Star of Myth and Time, interpret these to mean ages of growth and decay, or enlightenment and darkness; primal elements of the cosmic progression. Thus Mithraism is thought to have something to do with the changing ages within the precession cycle or Great Year (Plato's term for one complete precession of the equinox).
A consequence of the precession is a changing pole star. Currently Polaris is extremely well-suited to mark the position of the north celestial pole, as Polaris is a moderately bright star with a visual magnitude of 2.1 (variable), and it is located within a half degree of the pole.
On the other hand, Thuban in the constellation Draco, which was the pole star in 3000 BC, is much less conspicuous at magnitude 3.67 (one-fifth as bright as Polaris); today it is invisible in light-polluted urban skies.
The brilliant Vega in the constellation Lyra is often touted as the best north star (it fulfilled that role around 12000 BC and will do so again around the year AD 14000), however it never comes closer than 5° to the pole.
When Polaris becomes the north star again around 27800 AD, due to its proper motion it then will be farther away from the pole than it is now, while in 23600 BC it came closer to the pole.
It is more difficult to find the south celestial pole in the sky at this moment, as that area is a particularly bland portion of the sky, and the nominal south pole star is Sigma Octantis, which with magnitude 5.5 is barely visible to the naked eye even under ideal conditions. That will change from the eightieth to the ninetieth centuries, however, when the south celestial pole travels through the False Cross.
This situation also is seen on a star map. The orientation of the south pole is moving toward the Southern Cross constellation. For the last 2,000 years or so, the Southern Cross has nicely pointed to the south pole. By consequence, the constellation is no longer visible from subtropical northern latitudes, as it was in the time of the ancient Greeks.
The figures to the right attempt to explain the relation between the precession of the Earth's axis and the shift in the equinoxes. These figures show the position of the Earth's axis on the celestial sphere, a fictitious sphere which places the stars according to their position as seen from Earth, regardless of their actual distance. The first image shows the celestial sphere from the outside, with the constellations in mirror image. The second figure shows the perspective of a near-Earth position as seen through a very wide angle lens (from which the apparent distortion arises).
The rotation axis of the Earth describes, over a period of 25,700 years, a small circle (blue) among the stars, centered around the ecliptic north pole (the blue E) and with an angular radius of about 23.4°, an angle known as the obliquity of the ecliptic. The direction of precession is opposite to the daily rotation of the Earth on its axis. The orange axis was the Earth's rotation axis 5,000 years ago, when it pointed to the star Thuban. The yellow axis, pointing to Polaris, marks the axis now.
The equinoxes occur where the celestial equator intersects the ecliptic (red line), that is, where the Earth's axis is perpendicular to the line connecting the centers of the Sun and Earth. (Note that the term "equinox" here refers to a point on the celestial sphere so defined, rather than the moment in time when the Sun is overhead at the Equator, though the two meanings are related.) When the axis precesses from one orientation to another, the equatorial plane of the Earth (indicated by the circular grid around the equator) moves. The celestial equator is just the Earth's equator projected onto the celestial sphere, so it moves as the Earth's equatorial plane moves, and the intersection with the ecliptic moves with it. The positions of the poles and equator on Earth do not change, only the orientation of the Earth against the fixed stars.
As seen from the orange grid, 5,000 years ago, the vernal equinox was close to the star Aldebaran of Taurus. Now, as seen from the yellow grid, it has shifted (indicated by the red arrow) to somewhere in the constellation of Pisces.
Still pictures like these are only first approximations as they do not take into account the variable speed of the precession, the variable obliquity of the ecliptic, the planetary precession (whose center lies on a circle about 6° away from the poles) and the proper motions of the stars.
In popular science books, precession is often explained with the example of a spinning top. In both cases, the applied force is due to gravity. For a spinning top, this force tends to be almost parallel to the rotation axis. For the Earth, however, the applied forces of the Sun and the Moon are nearly perpendicular to the axis of rotation.
The Earth is not a perfect sphere but an oblate spheroid, with an equatorial diameter about 43 kilometers larger than its polar diameter. Because of the earth's axial tilt, during most of the year the half of this bulge that is closest to the Sun is off-center, either to the north or to the south, and the far half is off-center on the opposite side. The gravitational pull on the closer half is stronger, since gravity decreases with distance, so this creates a small torque on the Earth as the Sun pulls harder on one side of the Earth than the other. The axis of this torque is roughly perpendicular to the axis of the Earth's rotation so the axis of rotation precesses. If the Earth were a perfect sphere, there would be no precession.
The figure to the right explains how this process works. The Earth is given as a perfect sphere with the mass of the bulge approximated by a blue torus around its equator. The green arrows indicate the gravitational forces from the Sun on some extreme points. These tangential forces create a torque (orange), and this torque, added to the rotation (magenta), shifts the rotational axis to a slightly new position (yellow). Over time, the axis precesses along the white circle, which is centered around the ecliptic pole.
This torque is always in the same direction, perpendicular to the direction in which the rotation axis is tilted away from the ecliptic pole, so that it does not change the axial tilt itself. The magnitude of the torque from the sun (or the moon) varies with the gravitational object's alignment with the earth's spin axis and approaches zero when it is orthogonal.
Although the above explanation involved the Sun, the same explanation holds true for any object moving around the Earth, along or close to the ecliptic, notably, the Moon. The combined action of the Sun and the Moon is called the lunisolar precession. In addition to the steady progressive motion (resulting in a full circle in about 25,700 years) the Sun and Moon also cause small periodic variations, due to their changing positions. These oscillations, in both precessional speed and axial tilt, are known as the nutation. The most important term has a period of 18.6 years and an amplitude of less than 20 seconds of arc.
In addition to lunisolar precession, the actions of the other planets of the solar system cause the whole ecliptic to rotate slowly around an axis which has an ecliptic longitude of about 174° measured on the instantaneous ecliptic. This planetary precession shift is only 0.47 seconds of arc per year (more than a hundred times smaller than lunisolar precession), and takes place along the instantaneous equator. The sum of the two precessions is known as the general precession.
Because of gravitational disturbances by the other planets, the shape and orientation of Earth's orbit are not fixed, and the apsides (that is, perihelion and aphelion) slowly move with respect to a fixed frame of reference (i.e. the Earth's argument of periapsis slowly shifts). Therefore the anomalistic year is slightly longer than the sidereal year. It takes about 112,000 years for the ellipse to revolve once relative to the fixed stars.
Because the anomalistic year is longer than the sidereal year while the tropical year (which calendars attempt to track) is shorter, the two forms of precession add. It takes about 21,000 years for the ellipse to revolve once relative to the vernal equinox, that is, for the perihelion to return to the same date (given a calendar that tracks the seasons perfectly). The dates of perihelion and of aphelion advance each year on this cycle, an average of 1 day per 58 years.
Notice in the above figure that the areas swept during a specific season changes through time. Orbital mechanics require that the length of the seasons be proportional to the swept areas of the seasonal quadrants, so when the orbital eccentricity is extreme, the seasons on the far side of the orbit may be substantially longer in duration.
in arcseconds per Julian century, with T, the time in Julian centuries (that is, 36,525 days) since the epoch of 2000.
The rate of precession is the derivative of that:
The constant term of this speed corresponds to one full precession circle in 25,772 years.
The precession rate is not a constant, but is (at the moment) slowly increasing over time, as indicated by the linear (and higher order) terms in T. In any case it must be stressed that this formula is only valid over a limited time period. It is clear that if T gets large enough (far in the future or far in the past), the T² term will dominate and p will go to very large values. In reality, more elaborate calculations on the numerical model of solar system show that the precessional constants have a period of about 41,000 years, the same as the obliquity of the ecliptic. Note that the constants mentioned here are the linear and all higher terms of the formula above, not the precession itself. That is,
Theoretical models may calculate the proper constants (coefficients) corresponding to the higher powers of T, but since it is impossible for a (finite) polynomial to match a periodic function over all numbers, the error in all such approximations will grow without bound as T increases. In that respect, the International Astronomical Union chose the best-developed available theory. For up to a few centuries in the past and the future, all formulas do not diverge very much. For up to a few thousand years in the past and the future, most agree to some accuracy. For eras farther out, discrepancies become too large — the exact rate and period of precession may not be computed, even for a single whole precession period.
The precession of Earth's axis is a very slow effect, but at the level of accuracy at which astronomers work, it does need to be taken into account on a daily basis. Note that although the precession and the tilt of Earth's axis (the obliquity of the ecliptic) are calculated from the same theory and thus, are related to each other, the two movements act independently of each other, moving in mutually perpendicular directions.
Over longer time periods, that is, millions of years, it appears that precession is quasiperiodic at around 25,700 years; however, it will not remain so. According to Ward, when, in about 1,500 million years, the distance of the Moon, which is continuously increasing from tidal effects, has increased from the current 60.3 to approximately 66.5 Earth radii, resonances from planetary effects will push precession to 49,000 years at first, and then, when the Moon reaches 68 Earth radii in about 2,000 million years, to 69,000 years. This will be associated with wild swings in the obliquity of the ecliptic as well. Ward, however, used the abnormally large modern value for tidal dissipation. Using the 620-million year average provided by tidal rhythmites of about half the modern value, these resonances will not be reached until about 3,000 and 4,000 million years, respectively. Long before that time (about 2,100 million years from now), due to the increasing luminosity of the Sun, the oceans of the Earth will have boiled away, which will alter tidal effects significantly.