A solar eclipse occurs when the Moon passes between Earth and the Sun, obscuring some or all of the Sun’s visible disk from a region of Earth’s surface. The alignment is possible only at new moon, but it does not occur every month because the Moon’s orbit is inclined to Earth’s orbital plane. Depending on the alignment and the apparent sizes of the two bodies, an eclipse may be partial, total, annular, or hybrid. A total eclipse briefly reveals the Sun’s faint outer atmosphere, normally concealed by the brilliance of its visible surface.
Solar eclipses are important in astronomy, historical chronology, and the study of the Sun and Earth’s environment. Their records document the development of mathematical prediction, provide evidence about long-term changes in Earth’s rotation, and mark several major scientific investigations, including nineteenth-century solar spectroscopy and the 1919 measurement of gravitational light deflection. Although their geometry can be calculated centuries in advance, the exact appearance and environmental effects at any location depend on circumstances ranging from lunar topography to local weather.
Orbital geometry and the lunar shadow
A solar eclipse is a particular kind of occultation: a nearer body hides a more distant one. Its occurrence depends on the relative positions of the Sun, Moon, and observer, rather than simply on the Moon lying somewhere between Earth and the Sun. At most new moons, the lunar shadow passes north or south of Earth because the lunar orbital plane is inclined by approximately to the ecliptic, the plane of Earth’s orbit around the Sun. An eclipse becomes possible when new moon occurs sufficiently close to one of the two points where these planes intersect, called the ascending and descending orbital nodes.
The Sun’s diameter is about 400 times the Moon’s diameter, while its distance from Earth is also roughly 400 times greater. Consequently, both appear approximately half a degree across. Their apparent diameters are not constant, however. For an approximately spherical object with physical radius , observed from a distance from its center, its angular diameter is
The Moon’s apparent diameter varies more strongly than the Sun’s because the Moon’s orbital eccentricity is greater than that of Earth’s orbit. Near lunar perigee, its closest orbital point to Earth, the Moon is more likely to cover the entire solar disk. Near apogee, its farthest point, it is more likely to appear too small to do so. Earth’s changing distance from the Sun modifies this relationship.
The Moon casts several geometrically distinct shadow regions. Within the umbra, the entire solar photosphere is hidden. Within the penumbra, only part of it is obscured. Beyond the tip of the umbra lies the antumbra, where the Moon appears wholly inside the solar disk but leaves a ring of sunlight around it. These regions arise because the Sun is an extended source rather than a point source. The umbra narrows away from the Moon, while the penumbra spreads outward.
The umbra’s length is comparable to the Moon’s distance from Earth, making the distinction between total and annular eclipses sensitive to orbital geometry. Earth’s curvature also matters: an observer beneath the Moon is closer to it than an observer viewing it near the horizon. The Moon’s irregular limb adds small-scale complexity. Mountains and valleys alter the precise moments when the last exposed portions of the solar surface disappear, producing differences that cannot be represented fully by treating the Moon as a perfectly smooth sphere.
Types of solar eclipse
A partial solar eclipse occurs where an observer lies in the lunar penumbra and the Moon covers only part of the Sun. Some eclipses are partial everywhere: the central shadow misses Earth entirely, while part of the penumbra crosses it. Such events commonly occur at high latitudes. Other partial eclipses are the surrounding phases or peripheral views of a total, annular, or hybrid event. The same astronomical event can therefore be total for one observer, partial for another, and entirely invisible from a third location.
A total solar eclipse occurs where the lunar umbra intersects Earth and the Moon completely covers the Sun’s photosphere, its normally visible surface. Totality takes place along a relatively narrow track, usually no more than a few hundred kilometres wide. At a fixed location, it commonly lasts only a few minutes, with an upper limit of approximately seven and a half minutes under especially favorable geometry. A location near the edge of the track experiences a shorter total phase than a location near the centerline, even when the two are geographically close.
An annular solar eclipse occurs when the antumbra reaches Earth. At maximum eclipse, the Moon appears surrounded by a bright ring, or annulus, of photospheric light. Annularity is not an incomplete version of totality caused by an off-center observer: it results from the Moon being apparently smaller than the Sun. Even on the central line, some photosphere remains visible. The faint solar corona therefore does not become conspicuous as it does during totality. Annular phases can last longer than total phases, sometimes exceeding ten minutes.
A hybrid solar eclipse changes between annular and total along its central track. Earth’s curvature is crucial: portions of the surface nearer the Moon can enter the umbra, while more distant portions lie beyond its tip in the antumbra. A hybrid event may begin as annular, become total, and end as annular, although the exact sequence depends on the geometry and which portions of the shadow encounter Earth. The eclipse of 20 April 2023 was hybrid, with totality visible along a narrow track crossing parts of Australia, Timor-Leste, and Indonesia.
Two numerical descriptions distinguish the extent of coverage. Eclipse magnitude expresses the fraction of the Sun’s apparent diameter covered along the line through the disk centers; for central eclipses, it is commonly stated as the ratio of lunar to solar apparent diameter. Obscuration is the fraction of the apparent solar disk’s area hidden by the Moon. These quantities are not interchangeable. Neither directly specifies perceived darkness, because the solar surface is not uniformly bright and the visual system adjusts to changing illumination.
Local progression and visible phenomena
At a location experiencing a central eclipse, the sequence is described by four principal contacts. First contact occurs when the lunar and solar limbs first appear to touch, beginning the partial phase. Second contact marks the start of totality or annularity, when the lunar disk has moved fully inside the solar disk’s apparent boundary or has just covered it completely. Third contact ends that central phase, and fourth contact marks the final separation. For a partial-only observer, only first and fourth contacts occur. Depending on the geometry, the complete event can extend over several hours.
As the Moon covers more of the Sun, the remaining photosphere becomes a narrowing crescent. Small openings between leaves or other objects can act as natural pinhole projectors, casting crescent-shaped images on the ground. These are images of the partly eclipsed Sun rather than crescent-shaped shadows of leaves. The changing illumination may appear unfamiliar before it becomes dramatically dark: shadows can become unusually sharp in one direction because the remaining luminous crescent acts as a narrow light source. The relationship between disk coverage and ground illumination is modified by solar limb darkening, in which the visible solar surface is dimmer near its edge.
Immediately before and after totality, sunlight can pass through valleys along the lunar limb, creating Baily’s beads. The phenomenon is named after Francis Baily, who described it prominently during the annular eclipse of 1836, although related observations predated him. A single surviving bright bead, seen beside the emerging corona, produces the diamond-ring effect. Because individual beads depend on lunar terrain, their timing and appearance differ between nearby observing sites. Near the path boundary, beads may persist through much of what would otherwise be the central phase.
During totality, the solar corona becomes visible as a faint, structured glow around the dark lunar silhouette. Its streamers, plumes, and loops reflect the organization of the solar magnetic field. Pink or reddish prominences may appear near the limb, and the thin chromosphere can briefly become visible at the beginning or end of totality. Bright planets and some stars may emerge, while the horizon often remains illuminated by distant regions outside the umbra.
Rapidly moving light and dark bands sometimes appear on pale surfaces shortly before or after totality. These shadow bands result from atmospheric turbulence acting on light from the extremely thin solar crescent. Their visibility is variable and their pattern is influenced by conditions in Earth’s atmosphere. They are distinct from the main lunar shadow, whose arrival and departure cause the much larger transition into and out of totality.
Eclipse seasons, frequency, and repeating cycles
Solar eclipses cluster in eclipse seasons, intervals when the Sun appears sufficiently close to a lunar orbital node for an eclipse to occur at new moon. The nodes do not remain fixed: their line regresses around the ecliptic in approximately 18.6 years. As a result, the Sun returns to the same node after about 346.6 days, an interval called the eclipse year. Successive eclipse seasons are separated by roughly half that interval, about 173 days, rather than by exactly six calendar months.
An eclipse season lasts approximately a month, with its precise limits depending on the geometry. Usually one solar eclipse occurs within it, but two can occur when new moons fall near opposite ends of the season. Earth experiences at least two and at most five solar eclipses in a calendar year, counting all types. Five is uncommon and does not imply five total eclipses. At any single place, partial eclipses are much more frequent than total ones because the penumbra covers a far larger area than the umbra.
The best-known recurrence period is the saros, approximately 6,585.3 days, or 18 years, 11 days, and 8 hours, with the calendar-day count affected by leap years. Its usefulness follows from the near-commensurability of three lunar periods:
The synodic month measures the cycle of lunar phases; the draconic month measures return to the same orbital node; and the anomalistic month measures return to perigee. After one saros, an eclipse therefore tends to recur at a similar phase, node position, and Earth–Moon distance.
A saros does not produce an identical eclipse at the same place. The additional eight hours allow Earth to rotate roughly another third of a turn, shifting the next eclipse’s track approximately in longitude, generally westward. Small mismatches between the three lunar periods also shift the track in latitude and gradually change the event’s character. A saros series commonly begins with small partial eclipses near one polar region, develops central eclipses over many centuries, and ends with partial eclipses near the opposite polar region. Individual series typically span more than a thousand years.
Three saros periods form an exeligmos, approximately 54 years and one month, which brings the time of day closer to its earlier value and can produce a more geographically similar recurrence. Even this is not an exact repetition. The narrowness of totality tracks means that the average interval between total eclipses at a fixed location is several centuries, while actual intervals vary widely. Some places experience two total eclipses within a few years; others wait far longer than the average.
Calculating eclipse paths and circumstances
Modern eclipse prediction begins with an ephemeris, a numerical description of the positions and motions of astronomical bodies. Calculations use gravitational models constrained by observations, including precise measurements of planetary motion and lunar laser ranging. The Moon’s orbit is especially sensitive to perturbations from the Sun, and high-accuracy predictions require more than an isolated two-body model. The positions of the observer, Earth’s rotation axis, and the apparent solar and lunar disks must all be expressed consistently.
A widely used mathematical framework employs Besselian elements, named after the nineteenth-century astronomer Friedrich Wilhelm Bessel. These describe the motion and dimensions of the lunar shadow in a fundamental plane passing through Earth’s center and perpendicular to the shadow axis. Transformations between this plane and Earth’s surface yield the eclipse’s central line, northern and southern limits, contact times, Sun altitude, and duration. This approach separates the moving shadow’s geometry from the coordinates of individual observing sites and provides a compact way to publish eclipse circumstances.
The path’s ground speed is determined by both the Moon’s orbital motion and Earth’s rotation. Central eclipse tracks generally move from west to east, although their orientation can curve strongly across maps and high-latitude geometries can be complicated. Shadow speeds near the middle of favorable tracks are commonly of the order of a kilometre per second or less, while speeds can be much higher near the beginning and end of a track. The width and local duration depend jointly on the shadow’s size, its intersection angle with the surface, and its speed relative to the ground.
Earth is not a sphere, and an observer’s elevation changes the horizon and slightly changes the apparent lunar position. Precise work therefore uses a reference ellipsoid, geographic coordinates, and local height. Lunar topography introduces further corrections: modern elevation maps of the Moon allow individual valleys and mountains to be incorporated into predictions of beads and contact times. Near the edge of totality, such details can determine whether a particular site experiences a continuous interval of complete photospheric coverage.
For eclipses in the distant past or future, an important uncertainty is Delta T, written , the difference between uniform Terrestrial Time and time based on Earth’s rotation. Earth’s rotation changes because of tidal interactions and exchanges of angular momentum within the Earth system. Consequently, the celestial alignment can be computed more confidently than the exact longitude beneath it thousands of years away. An error of one second in Earth’s rotation corresponds to about 15 arcseconds of longitude. Ancient eclipse descriptions thus both depend on and help constrain reconstructions of historical Earth rotation.
Historical records and explanations
Eclipses were recorded in several ancient astronomical traditions, but the reliability of particular reports varies. Some texts give enough information to identify a specific event; others contain imagery or retrospective narratives that cannot securely establish a date. Chinese court records, Babylonian astronomical texts, and later Greek and Islamic writings preserve observations and computational practices of differing kinds. Their historical value lies not only in identifying celestial events but also in documenting how societies organized observation, calendars, political ritual, and mathematical knowledge.
A notable chronological anchor is the Assyrian eclipse recorded during the eponym year of Bur-Sagale, conventionally identified with the solar eclipse of 15 June 763 BCE. It helps connect sequences of named officials and events to an absolute chronology. Another famous account concerns an eclipse during a war between the Medes and Lydians, described by Herodotus and often associated with 28 May 585 BCE. Herodotus credited Thales with predicting the year, but the method and precision of any such prediction are not established. The report does not demonstrate that Thales could calculate a narrow totality track in the modern sense.
Ancient Greek natural philosophy developed physical explanations of eclipses as alignments involving an opaque Moon. Aristotle discussed eclipses within a spherical-Earth cosmology, while Ptolemy presented mathematical procedures for eclipse calculation in the second century CE. Babylonian astronomers recognized recurring lunar and eclipse patterns, but recognizing a cycle was not equivalent to predicting local solar visibility accurately. Solar eclipses are especially demanding because an event visible in one region may be absent elsewhere, unlike a lunar eclipse, which is visible across much of Earth’s night side.
Astronomers working in the medieval Islamic world refined tables, observations, and computational methods inherited from earlier traditions. In early modern Europe, improved orbital theory, logarithmic calculation, and telescopic observations progressively sharpened predictions. Edmond Halley’s map of the total solar eclipse of 3 May 1715, dated in the Gregorian calendar, was an important example of publishing a predicted shadow track for a broad audience. Its comparison with observations also helped assess errors in the underlying astronomical tables.
Eclipse expeditions expanded greatly during the nineteenth century. Observers transported telescopes, clocks, cameras, and spectroscopes to remote sites, often coordinating through observatories and learned societies. In 1860, photographs by Warren De La Rue and Angelo Secchi helped establish that the prominences seen during totality belonged to the Sun rather than to the Moon. By the twentieth century, aircraft, radio communication, and later satellite navigation changed expedition logistics, while historical records acquired a second scientific role as evidence for the long-term rotational behavior of Earth.
Solar physics and tests of gravitation
Total eclipses provide a natural means of suppressing the photosphere’s glare and exposing the Sun’s outer layers. The corona is a very tenuous plasma whose temperature reaches millions of kelvin, much hotter than the approximately 5,800-kelvin effective temperature of the photosphere. Its visible continuum includes photospheric light scattered by free electrons and sunlight scattered by interplanetary dust. Emission lines carry information about highly ionized atoms, while the corona’s shape reveals how plasma is organized by magnetic structures. Eclipses therefore connect direct imaging with spectroscopy and the physics of the solar wind.
During the eclipse of 18 August 1868, Pierre Janssen observed a bright yellow spectral line in solar prominences. Norman Lockyer independently observed the line later that year using methods that did not require an eclipse, and he associated it with an unfamiliar solar constituent. It became linked to helium, which was identified terrestrially in 1895. Another puzzling emission feature, the green coronal line observed in 1869, was once assigned to a hypothetical element called “coronium.” Work by Bengt Edlén around 1940 explained prominent coronal lines as emissions from highly ionized known elements; the green line near 530.3 nanometres is associated with iron stripped of thirteen electrons.
The total eclipse of 29 May 1919 supplied a celebrated test of general relativity. Expeditions observed stars near the eclipsed Sun from Sobral in Brazil and Príncipe off western Africa. Teams associated with Arthur Eddington, Frank Dyson, and Andrew Crommelin compared apparent stellar positions with reference photographs. General relativity predicted a deflection of approximately 1.75 arcseconds for light grazing the solar limb, about twice the value obtained from a then-discussed Newtonian corpuscular calculation. The reported results favored Einstein’s prediction, though observational difficulties and the treatment of different photographic sets became subjects of later historical and statistical examination.
In the weak-field limit, the relativistic deflection angle is approximately
where is the gravitational constant, is the deflecting mass, is the speed of light, and is the ray’s impact parameter. The phenomenon is an instance of gravitational lensing. Subsequent radio and spacecraft measurements tested gravitational light propagation far more precisely than the 1919 plates, but the eclipse observations remain important in the experimental history of Albert Einstein’s theory.
Modern space observatories and coronagraphs observe the Sun without waiting for eclipses, yet totality retains scientific value. Artificial occulting disks and instrumentally scattered light can limit access to the innermost corona. Eclipse experiments can sample this region with specialized cameras, polarimeters, and spectrometers. Multiple stations along a track extend time coverage, allowing investigators to distinguish evolving solar features from differences caused by viewing geometry or instrumentation.
Effects on the atmosphere, living organisms, and technology
A solar eclipse briefly interrupts incoming solar energy over a moving region, creating a natural experiment in the response of Earth’s surface and atmosphere. The radiative forcing is predictable in time and space, although clouds complicate the amount of sunlight reaching the ground. Surface and near-surface air temperatures often fall, with the minimum occurring after maximum obscuration because soils, vegetation, buildings, and air retain heat. The size of the response depends on season, solar altitude, cloud cover, wind, soil moisture, and the characteristics of the observing site.
Reduced surface heating can weaken daytime convection and alter the lowest part of the atmosphere, the boundary layer. Turbulence and wind speed may decline, and local changes in wind direction can occur. These responses are not uniform enough to support a single universal “eclipse wind.” An eclipse can also launch atmospheric gravity waves: oscillations in which buoyancy acts as the restoring force, distinct from the gravitational waves of general relativity. Measurements must separate eclipse-related signals from ordinary weather variability, which can produce fluctuations of comparable magnitude.
The ionosphere responds because sunlight maintains much of its ionization. As the Moon blocks solar radiation, production of ions and electrons decreases in affected regions, while recombination continues. Different layers respond on different timescales, and the shadow’s effect varies with altitude and solar conditions. Changes can be measured through ionosondes, radio signals, and the total electron content inferred from satellite navigation transmissions. High-frequency radio propagation may shift as ionospheric reflection and absorption change. These effects resemble some aspects of the day–night transition but occur much more rapidly and with a moving, geographically restricted source.
Plants and animals can also respond to the abrupt change in illumination and temperature. Measurements have recorded temporary changes in photosynthesis and plant gas exchange, while observers have documented altered bird vocalization, insect activity, and behavior in some mammals. Responses vary among species and settings; an eclipse does not cause a universal, coordinated “nighttime” behavior. Distinguishing a direct response to darkness from reactions to cooling, human crowds, or unusual sounds requires controlled observations or comparison with uneclipsed sites.
For electricity systems, eclipses produce a geographically organized reduction and recovery in photovoltaic generation. The effect depends on installed solar capacity, cloud conditions, and the fraction of the Sun obscured across the service area. The European eclipse of 20 March 2015 became a prominent operational case because substantial photovoltaic capacity lay within its partial-eclipse zone. Unlike an unexpected power-plant outage, an eclipse’s astronomical timing is known in advance, although weather introduces uncertainty into the actual generation change. Such events provide data on how rapidly varying solar output interacts with the wider electrical grid.
Observation, instrumentation, and eye safety
Solar-eclipse observation combines ordinary solar astronomy with the exceptional conditions of totality. Before and after the total phase, and throughout partial or annular eclipses, some photosphere remains exposed. Its surface brightness remains high even when its visible area is small. The reduction in overall daylight therefore does not indicate that the remaining solar crescent is harmless to view. Solar viewing can cause retinal injury without an immediate pain response, and magnifying optical instruments concentrate additional radiation into the eye or onto a camera sensor.
Filters intended for direct visual solar observation differ from ordinary sunglasses, photographic neutral-density filters, and filters used inside astronomical instruments. ISO 12312-2 specifies requirements for filters for direct observation of the Sun. The standard does not by itself describe every requirement for telescope or camera systems, whose filters must account for concentrated light and instrument design. During a total eclipse, the photosphere is hidden only between second and third contacts; an annular eclipse never produces this condition. The distinction between totality and a very deep partial eclipse is consequently fundamental to discussions of eclipse safety.
Indirect imaging uses optics to form a solar image on a surface rather than requiring direct viewing. A small aperture produces a pinhole image, with larger projected images obtained at greater distances between aperture and screen, though brightness and sharpness impose practical limits. Telescope projection can produce larger, more detailed images but also concentrates heat within the instrument. Public observatories often use purpose-built filtered telescopes, projection arrangements, or electronic displays to present the event to groups while maintaining control over the optical system.
Photographic requirements change sharply across the eclipse. The photosphere, chromosphere, prominences, and outer corona span an enormous range of brightness, so a single exposure cannot record them all faithfully. Coronal images commonly combine multiple exposures, with processing used to display both bright inner structures and faint outer streamers. The resulting image represents a reconstruction from measured brightness values rather than a simple equivalent of unaided visual appearance. Accurate timestamps, exposure records, and instrument calibration are particularly important when images are used for scientific measurement.
Site selection and interpretation also depend on weather, terrain, and solar altitude. A location on the centerline may offer a longer theoretical totality but poorer prospects of clear sky than a site elsewhere within the path. Mountains can obstruct a low Sun, while thin cloud can obscure faint coronal structure even when contacts remain visible. Aircraft observations can avoid some cloud layers and change the observer’s motion relative to the shadow. During the eclipse of 30 June 1973, a modified Concorde aircraft followed the lunar shadow and extended the observed total phase to about 74 minutes, far beyond the duration possible from a fixed point on Earth’s surface.