Solar eclipse

Each icon shows the view from the centre of its black spot, representing the Moon (not to scale)

Diamond ring effect at third contact—the end of totality—with visible prominences

Central eclipse is often used as a generic term for a total, annular, or hybrid eclipse.[14] This is, however, not completely correct: the definition of a central eclipse is an eclipse during which the central line of the umbra touches the Earth's surface. It is possible, though extremely rare, that part of the umbra intersects with the Earth (thus creating an annular or total eclipse), but not its central line. This is then called a non-central total or annular eclipse.[14] Gamma is a measure of how centrally the shadow strikes. The last (umbral yet) non-central solar eclipse was on April 29, 2014. This was an annular eclipse. The next non-central total solar eclipse will be on April 9, 2043.[15]

The visual phases observed during a total eclipse are called:[16]

• First contact—when the Moon's limb (edge) is exactly tangential to the Sun's limb.
• Second contact—starting with Baily's Beads (caused by light shining through valleys on the Moon's surface) and the diamond ring effect. Almost the entire disk is covered.
• Totality—the Moon obscures the entire disk of the Sun and only the solar corona is visible.
• Third contact—when the first bright light becomes visible and the Moon's shadow is moving away from the observer. Again a diamond ring may be observed.
• Fourth contact—when the trailing edge of the Moon ceases to overlap with the solar disk and the eclipse ends.

Geometry of a total solar eclipse (not to scale)

The diagrams to the right show the alignment of the Sun, Moon, and Earth during a solar eclipse. The dark gray region between the Moon and Earth is the umbra, where the Sun is completely obscured by the Moon. The small area where the umbra touches Earth's surface is where a total eclipse can be seen. The larger light gray area is the penumbra, in which a partial eclipse can be seen. An observer in the antumbra, the area of shadow beyond the umbra, will see an annular eclipse.[17]

The Moon's orbit around the Earth is inclined at an angle of just over 5 degrees to the plane of the Earth's orbit around the Sun (the ecliptic). Because of this, at the time of a new moon, the Moon will usually pass to the north or south of the Sun. A solar eclipse can occur only when a new moon occurs close to one of the points (known as nodes) where the Moon's orbit crosses the ecliptic.[18]

As noted above, the Moon's orbit is also elliptical. The Moon's distance from the Earth can vary by about 6% from its average value. Therefore, the Moon's apparent size varies with its distance from the Earth, and it is this effect that leads to the difference between total and annular eclipses. The distance of the Earth from the Sun also varies during the year, but this is a smaller effect. On average, the Moon appears to be slightly smaller than the Sun as seen from the Earth, so the majority (about 60%) of central eclipses are annular. It is only when the Moon is closer to the Earth than average (near its perigee) that a total eclipse occurs.[19][20]

	Moon		Sun
	At perigee (nearest)	At apogee (farthest)	At perihelion (nearest)	At aphelion (farthest)
Mean radius	1,737.10 km (1,079.38 mi)		696,000 km (432,000 mi)
Distance	363,104 km (225,622 mi)	405,696 km (252,088 mi)	147,098,070 km (91,402,500 mi)	152,097,700 km (94,509,100 mi)
Angular diameter[21]	33' 30" (0.5583°)	29' 26" (0.4905°)	32' 42" (0.5450°)	31' 36" (0.5267°)
Apparent size to scale
Order by decreasing apparent size	1st	4th	2nd	3rd

The Moon orbits the Earth in approximately 27.3 days, relative to a fixed frame of reference. This is known as the sidereal month. However, during one sidereal month, Earth has revolved part way around the Sun, making the average time between one new moon and the next longer than the sidereal month: it is approximately 29.5 days. This is known as the synodic month and corresponds to what is commonly called the lunar month.[18]

The Moon crosses from south to north of the ecliptic at its ascending node, and vice versa at its descending node.[18] However, the nodes of the Moon's orbit are gradually moving in a retrograde motion, due to the action of the Sun's gravity on the Moon's motion, and they make a complete circuit every 18.6 years. This regression means that the time between each passage of the Moon through the ascending node is slightly shorter than the sidereal month. This period is called the nodical or draconic month.[22]

Finally, the Moon's perigee is moving forwards or precessing in its orbit and makes a complete circuit in 8.85 years. The time between one perigee and the next is slightly longer than the sidereal month and known as the anomalistic month.[23]

The Moon's orbit intersects with the ecliptic at the two nodes that are 180 degrees apart. Therefore, the new moon occurs close to the nodes at two periods of the year approximately six months (173.3 days) apart, known as eclipse seasons, and there will always be at least one solar eclipse during these periods. Sometimes the new moon occurs close enough to a node during two consecutive months to eclipse the Sun on both occasions in two partial eclipses. This means that, in any given year, there will always be at least two solar eclipses, and there can be as many as five.[24]

Eclipses can occur only when the Sun is within about 15 to 18 degrees of a node, (10 to 12 degrees for central eclipses). This is referred to as an eclipse limit, and is given in ranges because the apparent sizes and speeds of the Sun and Moon vary throughout the year. In the time it takes for the Moon to return to a node (draconic month), the apparent position of the Sun has moved about 29 degrees, relative to the nodes.[2] Since the eclipse limit creates a window of opportunity of up to 36 degrees (24 degrees for central eclipses), it is possible for partial eclipses (or rarely a partial and a central eclipse) to occur in consecutive months.[25][26]

Fraction of the Sun's disc covered, f, when the same-sized discs are offset a fraction t of their diameter.[27]

During a central eclipse, the Moon's umbra (or antumbra, in the case of an annular eclipse) moves rapidly from west to east across the Earth. The Earth is also rotating from west to east, at about 28 km/min at the Equator, but as the Moon is moving in the same direction as the Earth's rotation at about 61 km/min, the umbra almost always appears to move in a roughly west–east direction across a map of the Earth at the speed of the Moon's orbital velocity minus the Earth's rotational velocity.[28]

The width of the track of a central eclipse varies according to the relative apparent diameters of the Sun and Moon. In the most favourable circumstances, when a total eclipse occurs very close to perigee, the track can be up to 267 km (166 mi) wide and the duration of totality may be over 7 minutes.[29] Outside of the central track, a partial eclipse is seen over a much larger area of the Earth. Typically, the umbra is 100–160 km wide, while the penumbral diameter is in excess of 6400 km.[30]

Besselian elements are used to predict whether an eclipse will be partial, annular, or total (or annular/total), and what the eclipse circumstances will be at any given location.[31]^{: Chapter 11}

Calculations with Besselian elements can determine the exact shape of the umbra's shadow on the Earth's surface. But at what longitudes on the Earth's surface the shadow will fall, is a function of the Earth's rotation, and on how much that rotation has slowed down over time. A number called ΔT is used in eclipse prediction to take this slowing into account. As the Earth slows, ΔT increases. ΔT for dates in the future can only be roughly estimated because the Earth's rotation is slowing irregularly. This means that, although it is possible to predict that there will be a total eclipse on a certain date in the far future, it is not possible to predict in the far future exactly at what longitudes that eclipse will be total. Historical records of eclipses allow estimates of past values of ΔT and so of the Earth's rotation. [31]^{: Equation 11.132}

The following factors determine the duration of a total solar eclipse (in order of decreasing importance):[32][33]

1. The Moon being almost exactly at perigee (making its angular diameter as large as possible).
2. The Earth being very near aphelion (furthest away from the Sun in its elliptical orbit, making its angular diameter nearly as small as possible).
3. The midpoint of the eclipse being very close to the Earth's equator, where the rotational velocity is greatest and is closest to the speed of the lunar shadow moving over Earth's surface.
4. The vector of the eclipse path at the midpoint of the eclipse aligning with the vector of the Earth's rotation (i.e. not diagonal but due east).
5. The midpoint of the eclipse being near the subsolar point (the part of the Earth closest to the Sun).

The longest eclipse that has been calculated thus far is the eclipse of July 16, 2186 (with a maximum duration of 7 minutes 29 seconds over northern Guyana).[32]

Between two and five solar eclipses occur every year, with at least one per eclipse season. Since the Gregorian calendar was instituted in 1582, years that have had five solar eclipses were 1693, 1758, 1805, 1823, 1870, and 1935. The next occurrence will be 2206.[42] On average, there are about 240 solar eclipses each century.[43]

The 5 solar eclipses of 1935

January 5	February 3	June 30	July 30	December 25
Partial (south)	Partial (north)	Partial (north)	Partial (south)	Annular (south)
Saros 111	Saros 149	Saros 116	Saros 154	Saros 121

Total solar eclipses are seen on Earth because of a fortuitous combination of circumstances. Even on Earth, the diversity of eclipses familiar to people today is a temporary (on a geological time scale) phenomenon. Hundreds of millions of years in the past, the Moon was closer to the Earth and therefore apparently larger, so every solar eclipse was total or partial, and there were no annular eclipses. Due to tidal acceleration, the orbit of the Moon around the Earth becomes approximately 3.8 cm more distant each year. Millions of years in the future, the Moon will be too far away to fully occlude the Sun, and no total eclipses will occur. In the same timeframe, the Sun may become brighter, making it appear larger in size.[44] Estimates of the time when the Moon will be unable to occlude the entire Sun when viewed from the Earth range between 650 million[45] and 1.4 billion years in the future.[44]

Eclipse glasses filter out eye damaging radiation, allowing direct viewing of the Sun during all partial eclipse phases; they are not used during totality, when the Sun is completely eclipsed

Pinhole projection method of observing partial solar eclipse. Insert (upper left): partially eclipsed Sun photographed with a white solar filter. Main image: projections of the partially eclipsed Sun (bottom right)

Viewing the Sun during partial and annular eclipses (and during total eclipses outside the brief period of totality) requires special eye protection, or indirect viewing methods if eye damage is to be avoided. The Sun's disk can be viewed using appropriate filtration to block the harmful part of the Sun's radiation. Sunglasses do not make viewing the Sun safe. Only properly designed and certified solar filters should be used for direct viewing of the Sun's disk.[70] Especially, self-made filters using common objects such as a floppy disk removed from its case, a Compact Disc, a black colour slide film, smoked glass, etc. must be avoided.[71][72]

The safest way to view the Sun's disk is by indirect projection.[73] This can be done by projecting an image of the disk onto a white piece of paper or card using a pair of binoculars (with one of the lenses covered), a telescope, or another piece of cardboard with a small hole in it (about 1 mm diameter), often called a pinhole camera. The projected image of the Sun can then be safely viewed; this technique can be used to observe sunspots, as well as eclipses. Care must be taken, however, to ensure that no one looks through the projector (telescope, pinhole, etc.) directly.[74] Viewing the Sun's disk on a video display screen (provided by a video camera or digital camera) is safe, although the camera itself may be damaged by direct exposure to the Sun. The optical viewfinders provided with some video and digital cameras are not safe. Securely mounting #14 welder's glass in front of the lens and viewfinder protects the equipment and makes viewing possible.[72] Professional workmanship is essential because of the dire consequences any gaps or detaching mountings will have. In the partial eclipse path, one will not be able to see the corona or nearly complete darkening of the sky. However, depending on how much of the Sun's disk is obscured, some darkening may be noticeable. If three-quarters or more of the Sun is obscured, then an effect can be observed by which the daylight appears to be dim, as if the sky were overcast, yet objects still cast sharp shadows.[75]

Solar eclipse of August 21, 2017

Baily's beads, sunlight visible through lunar valleys

Composite image with corona, prominences, and diamond ring effect

When the shrinking visible part of the photosphere becomes very small, Baily's beads will occur. These are caused by the sunlight still being able to reach the Earth through lunar valleys. Totality then begins with the diamond ring effect, the last bright flash of sunlight.[76]

It is safe to observe the total phase of a solar eclipse directly only when the Sun's photosphere is completely covered by the Moon, and not before or after totality.[73] During this period, the Sun is too dim to be seen through filters. The Sun's faint corona will be visible, and the chromosphere, solar prominences, and possibly even a solar flare may be seen. At the end of totality, the same effects will occur in reverse order, and on the opposite side of the Moon.[76]

A dedicated group of eclipse chasers have pursued the observation of solar eclipses when they occur around the Earth.[77] A person who chases eclipses is known as an umbraphile, meaning shadow lover.[78] Umbraphiles travel for eclipses and use various tools to help view the sun including solar viewing glasses, also known as eclipse glasses, as well as telescopes.[79][80]

The progression of a solar eclipse on August 1, 2008 in Novosibirsk, Russia. All times UTC (local time was UTC+7). The time span between shots is three minutes.

Photographing an eclipse is possible with fairly common camera equipment. In order for the disk of the Sun/Moon to be easily visible, a fairly high magnification long focus lens is needed (at least 200 mm for a 35 mm camera), and for the disk to fill most of the frame, a longer lens is needed (over 500 mm). As with viewing the Sun directly, looking at it through the optical viewfinder of a camera can produce damage to the retina, so care is recommended.[81] Solar filters are required for digital photography even if an optical viewfinder is not used. Using a camera's live view feature or an electronic viewfinder is safe for the human eye, but the Sun's rays could potentially irreparably damage digital image sensors unless the lens is covered by a properly designed solar filter.[82]

Eddington's original photograph of the 1919 eclipse, which provided evidence for Einstein's theory of general relativity.

The observation of a total solar eclipse of May 29, 1919, helped to confirm Einstein's theory of general relativity. By comparing the apparent distance between stars in the constellation Taurus, with and without the Sun between them, Arthur Eddington stated that the theoretical predictions about gravitational lenses were confirmed.[86] The observation with the Sun between the stars was possible only during totality since the stars are then visible. Though Eddington's observations were near the experimental limits of accuracy at the time, work in the later half of the 20th century confirmed his results.[87][88]

There is a long history of observations of gravity-related phenomena during solar eclipses, especially during the period of totality. In 1954, and again in 1959, Maurice Allais reported observations of strange and unexplained movement during solar eclipses.[89] The reality of this phenomenon, named the Allais effect, has remained controversial. Similarly, in 1970, Saxl and Allen observed the sudden change in motion of a torsion pendulum; this phenomenon is called the Saxl effect.[90]

Observation during the 1997 solar eclipse by Wang et al. suggested a possible gravitational shielding effect,[91] which generated debate. In 2002, Wang and a collaborator published detailed data analysis, which suggested that the phenomenon still remains unexplained.[92]

In principle, the simultaneous occurrence of a solar eclipse and a transit of a planet is possible. But these events are extremely rare because of their short durations. The next anticipated simultaneous occurrence of a solar eclipse and a transit of Mercury will be on July 5, 6757, and a solar eclipse and a transit of Venus is expected on April 5, 15232.[93]

More common, but still infrequent, is a conjunction of a planet (especially, but not only, Mercury or Venus) at the time of a total solar eclipse, in which event the planet will be visible very near the eclipsed Sun, when without the eclipse it would have been lost in the Sun's glare. At one time, some scientists hypothesized that there may be a planet (often given the name Vulcan) even closer to the Sun than Mercury; the only way to confirm its existence would have been to observe it in transit or during a total solar eclipse. No such planet was ever found, and general relativity has since explained the observations that led astronomers to suggest that Vulcan might exist.[94]

The Moon's shadow over Turkey and Cyprus, seen from the ISS during a 2006 total solar eclipse.

A composite image showing the ISS transit of the Sun while the 2017 solar eclipse was in progress.

Artificial satellites can also pass in front of the Sun as seen from the Earth, but none is large enough to cause an eclipse. At the altitude of the International Space Station, for example, an object would need to be about 3.35 km (2.08 mi) across to blot the Sun out entirely. These transits are difficult to watch because the zone of visibility is very small. The satellite passes over the face of the Sun in about a second, typically. As with a transit of a planet, it will not get dark.[95]

Observations of eclipses from spacecraft or artificial satellites orbiting above the Earth's atmosphere are not subject to weather conditions. The crew of Gemini 12 observed a total solar eclipse from space in 1966.[96] The partial phase of the 1999 total eclipse was visible from Mir.[97]

The solar eclipse of March 20, 2015, was the first occurrence of an eclipse estimated to potentially have a significant impact on the power system, with the electricity sector taking measures to mitigate any impact. The continental Europe and Great Britain synchronous areas were estimated to have about 90 gigawatts of solar power and it was estimated that production would temporarily decrease by up to 34 GW compared to a clear sky day.[98][99]

Eclipses may cause the temperature to decrease by 3 °C, with wind power potentially decreasing as winds are reduced by 0.7 m/s.[100]

In addition to the drop in light level and air temperature, animals change their behavior during totality. For example, birds and squirrels return to their nests and crickets chirp.[101]