Latitude and Longitude

Latitude and longitude are angular coordinates used together to specify locations on Earth. Latitude expresses position north or south of the equator, while longitude expresses position east or west of a designated prime meridian. Lines of equal latitude, called parallels, and lines of equal longitude, called meridians, form the geographical grid used on globes and maps. Unlike distances measured in metres, these coordinates describe angles, conventionally expressed in degrees. (rmg.co.uk)

Precise coordinates depend on how Earth’s shape and the coordinate reference system are defined. In geodesy, latitude ordinarily refers to the angle between the equatorial plane and the normal to a reference ellipsoid; it is not generally identical to the angle measured from Earth’s centre. A geodetic datum or reference frame establishes the basis for interpreting a coordinate pair, and ellipsoidal height can supply a third coordinate. These distinctions underpin surveying, navigation, and the integration of spatial information. (proj.org)

The development of latitude and longitude connects astronomy, cartography, and precision timekeeping. Determining longitude at sea stimulated both astronomical methods and marine chronometers, while the adoption of the Greenwich meridian in 1884 helped establish a common international reference. Satellite-based reference frames subsequently refined this framework, and standardized coordinate formats now allow geographical positions to be exchanged between digital systems. (rmg.co.uk)

Angular Coordinates and the Geographical Grid

Latitude and longitude are complementary angular coordinates for specifying horizontal position on Earth. Latitude, conventionally denoted by $\phi$, measures position north or south of the equator; longitude, denoted by $\lambda$, measures position east or west of a reference meridian. On a spherical globe, latitude is the angle between the equatorial plane and a line from the globe’s centre to the point, while longitude is the angle around the polar axis from the reference meridian to the point’s meridian. These are geographical coordinates, expressed in angular units rather than units of length. Precise terrestrial coordinates normally use an ellipsoidal model, whose definition of latitude differs from the elementary spherical construction. (oc.nps.edu)

The geographical grid is organized around the rotational poles and the equator. In the idealized globe model, the North and South poles are the two points where the rotation axis intersects the surface. The equatorial plane passes through the centre perpendicular to that axis, and its intersection with the surface forms the equator, at $0^\circ$ latitude. Other lines of constant latitude are called parallels: they encircle the globe east–west in planes parallel to the equatorial plane. Their circles become smaller toward the poles, where they reduce to points. Latitude extends from $90^\circ$ south at the South Pole to $90^\circ$ north at the North Pole; the equator separates the Northern and Southern hemispheres. (oc.nps.edu)

A meridian is a line of constant longitude running from one pole to the other. Unlike parallels, meridians converge at both poles. Together, the meridians and parallels form the graticule, intersecting at right angles on the globe. Away from the poles, one specified parallel and one specified meridian intersect at a single point: for example, $30^\circ$ N, $60^\circ$ E identifies the intersection of the parallel 30 degrees north of the equator with the meridian 60 degrees east of the longitude origin. The conventional longitude range is $180^\circ$ west to $180^\circ$ east, measured from the prime meridian at $0^\circ$; the two 180-degree limits identify the same meridian on the opposite side of the globe. At either pole, all meridians meet, so longitude does not distinguish additional locations. (oc.nps.edu)

The graticule should not be confused with a projected rectangular coordinate grid. A map projection represents geographical positions on a plane, where coordinates may be expressed as eastings and northings, usually in metres. These values refer to the projection’s coordinate axes and assigned origin rather than directly measuring angles from the equator and prime meridian. In the Universal Transverse Mercator system, for example, a zone’s central meridian is assigned an easting of 500,000 metres; a point one metre east of it has an easting of 500,001 metres. Such values belong to a particular projected grid, not to the global angular graticule. (usgs.gov)

A latitude–longitude ordered pair supplies two coordinates, not a complete three-dimensional location. It identifies horizontal position relative to a reference surface but does not specify how far above or below that surface a point lies. A third coordinate, height, is therefore needed: geographical position can be recorded as $(\phi,\lambda,h)$, where $h$ denotes height above the reference ellipsoid. Points at different heights can share the same latitude and longitude. The height reference must also be specified, since ellipsoidal height and elevation relative to a sea-level-based surface are different quantities. (repository.library.noaa.gov)

Latitude on a Sphere and an Ellipsoid

On a spherical model of Earth, latitude is the angle between the equatorial plane and the radius drawn to a point. The radius is also perpendicular to the sphere’s surface, so a definition based on the surface normal gives the same angle. Earth’s rotational flattening is represented more closely by an oblate reference ellipsoid: an ellipse rotated about its shorter, polar axis. Its equatorial semiaxis $a$ exceeds its polar semiaxis $b$. On this surface, the normal generally does not pass through the centre, making “angle above the equatorial plane” ambiguous unless the defining line is specified. A latitude must therefore be interpreted with its surface model and latitude convention identified. (oc.nps.edu)

Geodetic latitude, denoted $\phi$, is the angle between the equatorial plane and the outward normal to the reference ellipsoid. For a point above the ellipsoid, the defining direction is the normal through that point, extended from the ellipsoidal surface. Geocentric latitude, denoted $\phi_c$, instead uses the line joining the ellipsoid’s centre to the point. Both angles are positive northward and negative southward. Geodetic latitude is the convention ordinarily used for mapping and geographical coordinates; geocentric latitude describes the direction of the position vector from the centre. Their difference is geometrical, not an observational error. (oc.nps.edu)

For a point on the reference ellipsoid, the relationship is

Here $e$ is the ellipsoid’s first eccentricity and $f$ its flattening. The relationship follows from the geometry of a meridian ellipse: at cylindrical distance $p$ from the polar axis and height $z$ above the equatorial plane, the radius has slope $z/p$, whereas the surface normal has slope $a^2z/(b^2p)$. Thus, away from the equator and poles, geocentric latitude has a smaller absolute value than geodetic latitude. The angles coincide at the equator and, by the limiting geometry rather than substitution into the tangent formula, at both poles. For a sphere, $a=b$ and $e=0$, so they coincide everywhere. The surface-point formula is not generally valid for a point at nonzero ellipsoidal height. (oc.nps.edu)

The WGS 84 ellipsoid provides a numerical example: its defining equatorial semiaxis is $6{,}378{,}137$ metres and its inverse flattening is $298.257223563$. These give $e^2\approx0.00669438$. Substitution into the surface-point relationship shows that a point at geodetic latitude $45^\circ$ north has geocentric latitude approximately $44.8076^\circ$ north—a difference of about $0.1924^\circ$, or $11.55$ arcminutes. This calculated difference illustrates why the two latitude conventions cannot be interchanged even though Earth’s flattening is small. (earth-info.nga.mil)

Astronomical latitude, often written $\Phi$, uses a third direction: the upward local vertical, opposite to the direction of effective gravity indicated by a plumb line. It is the angle between that vertical and the equatorial plane, rather than between an ellipsoid normal and that plane. Uneven mass distribution makes the physical vertical depart from the smooth ellipsoid’s normal. This angular departure is the deflection of the vertical; its north–south component is $\xi=\Phi-\phi$. NOAA’s National Geodetic Survey describes surface deflections as typically a few arcseconds, sometimes reaching an arcminute. Astronomical and geodetic latitude therefore differ for a physical reason distinct from the purely geometrical difference between geodetic and geocentric latitude. (geodesy.noaa.gov)

Longitude, Prime Meridians, and Time

Longitude, denoted by $\lambda$, is the directed angle between a location’s meridian plane and a reference meridian plane, measured around the polar axis. On a sphere or an ellipsoid of revolution, each such plane contains the polar axis; longitude specifies its orientation in the equatorial plane, not an angle measured along a parallel at the location’s latitude. Eastward angles are conventionally positive and westward angles negative: $30^\circ$ E and $30^\circ$ W therefore represent $\lambda=+30^\circ$ and $\lambda=-30^\circ$. Unlike the equator, whose plane is determined by Earth’s rotational axis, no particular meridian is physically distinguished as the origin of longitude. A prime meridian must consequently be selected by convention. (oc.nps.edu)

Historical maps used several competing origins, including Paris, Ferro in the Canary Islands, and Pulkovo near Saint Petersburg. Relative to Greenwich, these meridians were approximately $2^\circ20'$ east, $17^\circ40'$ west, and $30^\circ20'$ east, respectively. Thus, a longitude on an old map cannot be interpreted unambiguously without identifying its reference meridian. Greenwich’s historic reference was defined by the Airy Transit Circle, installed at the Royal Observatory in 1851 under George Biddell Airy. The International Meridian Conference in Washington, D.C., in 1884, selected the meridian through this instrument as the common origin; the circumstances of international adoption belong to the broader history of coordinate standardization. (guides.loc.gov)

Longitude is directly related to local solar time because Earth’s rotation carries successive meridians past the direction of the Sun. A complete mean solar cycle corresponds to $360^\circ$ in 24 hours, giving

One degree therefore corresponds to four minutes of solar time, and one arcminute of longitude to four seconds of time. With east-positive longitude, the difference between local mean solar time and Greenwich mean solar time is $\lambda/(15^\circ/\mathrm{h})$, modulo 24 hours. For example, when Greenwich mean solar time is 12:00, it is 14:00 at $30^\circ$ E and 07:00 at $75^\circ$ W. Apparent solar time, indicated by a sundial, differs seasonally from mean solar time because of Earth’s orbital eccentricity and axial inclination; this difference, the equation of time, can reach about 16 minutes. (aa.usno.navy.mil)

Civil clock time is not a continuous function of longitude. Governments establish time-zone boundaries, which commonly follow administrative borders rather than meridians, and locations throughout a zone share a clock reading despite their different solar times. Daylight-saving adjustments introduce another distinction. The $15^\circ$-per-hour correspondence therefore expresses an astronomical relationship, not a rule from which a location’s legally observed time can be deduced. (aa.usno.navy.mil)

The historic Greenwich line is also distinct from the IERS Reference Meridian, the zero-longitude reference used by modern Earth-centred coordinate systems. At Greenwich, the latter passes approximately 102.5 metres east of the Airy Transit Circle; the instrument’s modern geodetic longitude is about $5.3''$ W. The historic astronomical determination depended on observations referenced to the local vertical, established by gravity. That vertical does not coincide exactly with the normal to an Earth-centred reference ellipsoid, a difference called deflection of the vertical. Stephen Malys and colleagues demonstrated in 2015 that its east–west component accounts for the Greenwich offset. Moving the reference plane’s ground trace eastward allows it to pass through Earth’s centre while preserving essentially the astronomical orientation used for timekeeping. Consequently, a GPS receiver at the historic courtyard line does not report exactly zero longitude. (rmg.co.uk)

Coordinate Notation and Digital Representation

Angular coordinates are commonly written in three forms: degrees–minutes–seconds (DMS), such as $40^\circ 30' 15''$; degrees and decimal minutes (DDM), such as $40^\circ 30.25'$; and decimal degrees (DD), such as $40.5041667^\circ$. One degree contains 60 angular minutes, and one minute contains 60 angular seconds, so one second equals $1/3600$ degree. These subdivisions describe angles, not intervals of time. DMS can include fractional seconds; for example, the United States National Geodetic Survey’s datasheets present geodetic coordinates as integer degrees, integer minutes, decimal seconds, and a hemisphere identifier. Its handheld-position submission interface also accepts degrees and decimal minutes. (geodesy.noaa.gov)

Hemisphere letters attach direction to an otherwise nonnegative angular magnitude: N and S identify latitude north and south of the equator, while E and W identify longitude east and west of the reference meridian. In the conventional signed representation, north and east are positive; south and west are negative. For nonnegative degree, minute, and second components $D$, $M$, and $S$, conversion to signed decimal degrees is

where $\sigma=+1$ for N or E and $\sigma=-1$ for S or W. The sign applies to the whole magnitude, rather than only to the degree component. For example, $73^\circ 59' 15''\mathrm{W}$ becomes $73^\circ 59.25'\mathrm{W}$, or $-(73+59/60+15/3600)=-73.9875^\circ$. Conversely, multiplying the fractional part of $73.9875$ by 60 gives $59.25$ minutes; multiplying the remaining $0.25$ minute by 60 gives 15 seconds. (geodesy.noaa.gov)

Coordinate order is a separate convention from angular notation. A latitude-first ordered pair is written $(\varphi,\lambda)$; National Geodetic Survey records likewise display latitude before longitude. An illustrative latitude-first pair $(40.5041667,-73.9875)$ therefore identifies a northern latitude and western longitude. Without axis labels or a specified format, the two numbers do not establish which axis comes first. Reversing this example still produces values within the permitted angular ranges, but describes a different location. (geodesy.noaa.gov)

GeoJSON, a geographical data-interchange format specified in RFC 7946 in August 2016, requires the opposite order: longitude first, latitude second. The illustrative pair above would be encoded as {"type":"Point","coordinates":[-73.9875,40.5041667]}. Its coordinates are JSON numbers in decimal degrees, without degree symbols or hemisphere letters. RFC 7946 specifies the World Geodetic System 1984 (WGS 84) datum and a geographical coordinate reference system equivalent to OGC CRS84; this fixes both the reference system and the horizontal axis order for interchange. (rfc-editor.org)

A GeoJSON position may include a third number representing height in metres above or below the WGS 84 reference ellipsoid. Thus [-73.9875,40.5041667,25] specifies an ellipsoidal height of 25 metres, not necessarily an elevation of 25 metres above mean sea level. Omitting this component does not assert a zero ellipsoidal height. RFC 7946 also discourages extending position arrays beyond three elements: additional values, such as timestamps or route measures, have no standardized interpretation and may be ignored by parsers. (rfc-editor.org)

Geodetic Datums and Reference Frames

A reference ellipsoid specifies a mathematical surface approximating Earth’s shape, usually through its semi-major axis $a$ and flattening $f=(a-b)/a$, where $b$ is the semi-minor axis. Its dimensions alone do not establish its placement relative to the physical Earth. A geodetic datum supplies that relationship, including the origin and orientation used to reference positions. A terrestrial reference frame realizes a reference system through numerical coordinates assigned to identifiable stations, often together with their velocities. These terms overlap in modern standards: a geodetic reference frame may itself be classified as a datum. A coordinate reference system combines the datum or frame with a coordinate system specifying axes, their order, and units. Unambiguous geographical coordinates therefore require the applicable datum or frame, coordinate conventions, and, for time-dependent positions, an epoch; any height also requires an identified height reference. (geodesy.noaa.gov)

GRS 80, WGS 84, NAD 83, and ITRF name related but distinct geodetic objects. The Geodetic Reference System 1980 supplies the GRS 80 ellipsoid used by the North American Datum of 1983. Its semi-major axis is $6\,378\,137$ metres and its inverse flattening is approximately $298.257222101$. WGS 84—the World Geodetic System 1984—is broader than an ellipsoid: it defines an Earth-centred, Earth-fixed system and associated reference-frame realizations used with the Global Positioning System. Its ellipsoid has the same semi-major axis as GRS 80 but an inverse flattening of $298.257223563$. These nearly identical dimensions do not make NAD 83 and WGS 84 interchangeable, because their realized origins, orientations, and evolution also matter. NAD 83 was first released in 1986; its original network adjustment incorporated approximately 250,000 stations. (geodesy.noaa.gov)

The International Terrestrial Reference Frame (ITRF) is a realization of the International Terrestrial Reference System, rather than a particular ellipsoid. ITRF2020 combines station-position time series from very long baseline interferometry, satellite laser ranging, global navigation satellite systems, and DORIS satellite tracking, together with surveyed connections between instruments at shared sites. Such realizations estimate station positions and velocities within a common global framework. NAD 83 realizations serve a different purpose: NAD 83(2011), for example, is referenced to the North American tectonic plate, while PA11 and MA11 are associated with the Pacific and Mariana plates. The distinction affects how motion caused by plate tectonics appears in reported coordinates. (itrf.ign.fr)

The same survey mark can consequently receive different latitude and longitude values in different datums without changing its physical location. Differences may arise from the ellipsoid, reference origin, orientation, scale, or distortions inherited from older survey networks. A datum transformation accounts for these differences: a three-dimensional Helmert transformation can apply three translations, three rotations, and a scale change, while spatial correction grids can represent variations that a single parameter set cannot adequately describe. Time-dependent transformations may additionally include rates of change. This is distinct from a map-projection conversion, which changes geographical coordinates into plane coordinates while retaining the underlying datum. Merely assigning a different datum label to existing numbers does not perform a transformation. (proj.org)

Precise records distinguish a realization’s name from the epoch to which particular coordinates refer. In “NAD 83(2011), epoch 2010.00,” the parenthetical number identifies the realization, whereas the epoch denotes January 1, 2010. Likewise, ITRF2020 has a reference epoch of 2015.0, not 2020. For a station with approximately uniform motion, coordinates can be propagated using $\mathbf{X}(t)=\mathbf{X}(t_0)+\mathbf{V}(t-t_0)$, where $\mathbf{X}(t_0)$ is its position at the reference epoch and $\mathbf{V}$ its velocity. ITRF2020 also supplies post-seismic and seasonal terms where applicable, because a constant velocity alone cannot represent all station motion. An observation date and a coordinate epoch need not coincide when observations have been reduced to a common reference date. (geodesy.noaa.gov)

Determining Position by Observation and Survey

Astronomical latitude is determined by measuring celestial directions relative to the local vertical or horizon. At sea, a sextant measures the altitude of the Sun or a star above the visible horizon; reductions account for instrumental error, the depression of the horizon caused by the observer’s height, atmospheric refraction, and, where applicable, parallax and the observed body’s apparent radius. At upper meridian transit, the corrected altitude $h$ and the body’s declination $\delta$—its angular distance north or south of the celestial equator—satisfy $90^\circ-h=|\phi_{\mathrm{A}}-\delta|$, where $\phi_{\mathrm{A}}$ is astronomical latitude. The body’s position north or south of the zenith resolves the sign: for example, an equatorial Sun observed due south at a corrected noon altitude of $50^\circ$ gives $40^\circ$ north latitude. In the Northern Hemisphere, Polaris provides another reference because its altitude approximates latitude, with corrections for its displacement from the celestial pole. These observations refer to the gravity-defined vertical, not directly to an ellipsoid normal. (thenauticalalmanac.com)

Astronomical longitude requires a comparison between locally determined time and time at a reference meridian. A celestial transit or altitude observation establishes local astronomical time; a chronometer supplies reference time carried from a known meridian, with its error and rate accounted for. When both readings represent the same time scale, an hour’s difference corresponds to $15^\circ$ of longitude. Thus local mean time of 14:00 when Greenwich mean time is 12:00 indicates $30^\circ$ east longitude; apparent solar time must first be reconciled with mean time. The lunar-distance method instead obtains reference time from the changing angular separation of the Moon and the Sun or a selected star. After correcting the observed separation for refraction and parallax, comparison with predicted separations gives reference time, which is then compared with local time. Nevil Maskelyne’s Nautical Almanac supplied lunar-distance predictions for 1767, making this astronomical clock practically accessible to navigators. (ngs.noaa.gov)

Land surveying transfers established coordinates through a geodetic control network, rather than requiring an independent celestial determination at every point. In triangulation, a measured baseline and observed angles establish successive triangle sides through trigonometry; initial coordinates and orientation anchor the network. Trilateration measures triangle sides directly, while a traverse propagates position through successive measured distances and angles. Redundant observations and network adjustment reconcile discrepancies and estimate coordinate uncertainty. Permanent survey monuments preserve identifiable points to which subsequent surveys can connect. In the United States, the Transcontinental Arc of Triangulation began in 1871; a later high-precision transcontinental traverse was completed in 1976. Such interconnected surveys made coordinates mutually consistent over extensive regions. (ngs.noaa.gov)

Satellite positioning substitutes radio-signal observations for celestial altitudes and terrestrial sight lines. A GPS receiver combines broadcast satellite positions with measured signal travel times. Multiplication by the speed of light produces apparent ranges, or pseudoranges, which also contain receiver-clock error and propagation effects. An unconstrained three-dimensional solution ordinarily requires at least four satellites to solve simultaneously for three spatial coordinates and the receiver’s clock offset. The resulting position is calculated in an Earth-centred, Earth-fixed frame and converted into latitude, longitude, and ellipsoidal height. GPS realizes WGS 84 through its satellite and monitoring infrastructure; the radio observations themselves are not latitude–longitude readings. (govinfo.gov)

Survey-grade satellite processing also connects new observations to reference stations with established coordinates. NOAA’s Online Positioning User Service, for example, processes uploaded GPS observations using continuously operating reference stations; antenna type and antenna height are part of the observation metadata. Network processing can first estimate coordinates in an international terrestrial frame and then transform them into a selected output frame. Consequently, the observation method and the reported datum are distinct: a satellite-observed point may be expressed in different reference frames without having been reoccupied. The published record must identify the frame realization and coordinate epoch as well as the numerical latitude and longitude. (ngs.noaa.gov)

Distances, Directions, and Coordinate Geometry

A difference in angular coordinates is not itself a distance. One degree of latitude corresponds to approximately 111 kilometres along a meridian, with modest variation caused by Earth’s ellipsoidal shape. Longitude has a stronger dependence on latitude: meridians converge toward the poles, so the distance covered by one degree of longitude decreases as absolute latitude increases. The U.S. Geological Survey gives illustrative values at $38^\circ$ north of about 364,000 feet (111 kilometres) per degree of latitude and 288,200 feet (88 kilometres) per degree of longitude. Consequently, equal-sized angular cells in a latitude–longitude grid do not have equal east–west dimensions at different latitudes. (usgs.gov)

On a sphere of radius $R$, these relationships follow from elementary geometry. A meridian is a circle of radius $R$, whereas the parallel at latitude $\varphi$ has radius $R\cos\varphi$. Applying the arc-length relation—radius multiplied by angle—gives the signed local displacement components

where both angular increments are in radians. Because the north and east directions are perpendicular, the surface line element is

For nearby points, evaluating the cosine at their mean latitude gives a local distance approximation. Choosing $R=6{,}371$ kilometres yields about 111.2 kilometres per latitude degree, while a longitude degree along the $60^\circ$ parallel spans about 55.6 kilometres. A local direction can likewise be obtained from $\alpha=\operatorname{atan2}(ds_{\mathrm{east}},ds_{\mathrm{north}})$, with $\alpha$ measured clockwise from north; this is a local approximation, not a general long-distance bearing formula. (oc.nps.edu)

Precise calculations replace the sphere with a reference ellipsoid and use geodetic latitude. Two distinct curvature radii are required: the meridional radius $M(\varphi)$, governing north–south displacements, and the prime-vertical radius $N(\varphi)$, governing east–west displacements. At the ellipsoid surface,

Both radii vary with latitude; neither should simply be replaced by the distance from Earth’s centre to the surface point. A finite meridian distance therefore requires integrating $M(\varphi)$ over the latitude interval. The length along a parallel is $N(\varphi)\cos\varphi$ times the traversed longitude angle, but that parallel arc generally is not the shortest surface route between its endpoints. (oc.nps.edu)

Surface distance also differs from straight-line distance through three-dimensional space. On a sphere, a shortest surface route follows the shorter arc of a great circle, whose plane passes through the sphere’s centre. If its endpoints subtend a central angle $\sigma$, the surface distance is $R\sigma$, whereas elementary geometry gives the chord length $2R\sin(\sigma/2)$. On an ellipsoid, shortest surface routes are segments of geodesics, rather than generally being planar great-circle arcs. The inverse geodesic problem determines distance and endpoint azimuths from two coordinate pairs; the direct problem determines an endpoint from a starting position, initial azimuth, and distance. Azimuth is measured clockwise from north, and its value generally changes along the route. These model-surface distances do not include terrain undulations. (ncl.ucar.edu)

Latitude and longitude can also be represented in a Cartesian coordinate system. For a unit sphere centred at the origin, let the $z$-axis point toward the north pole, the $x$-axis intersect the equator at zero longitude, and the $y$-axis intersect it at $90^\circ$ east. Resolving the equatorial projection and polar component gives

Multiplying these components by $R$ gives coordinates on a sphere of radius $R$. For two unit position vectors $\mathbf{u}$ and $\mathbf{v}$, their dot product gives $\cos\sigma=\mathbf{u}\cdot\mathbf{v}$, connecting Cartesian geometry directly to great-circle distance. These spherical formulas must not be applied unchanged to geodetic latitude on an ellipsoid, where the surface normal and centre-to-point direction generally differ. (ncl.ucar.edu)

Historical Development and International Standardization

In the third century BCE, Eratosthenes (c. 276–194 BCE) organized geographical information around selected parallels and meridians rather than a uniformly spaced angular grid. His north–south distances were probably expressed as meridian arc lengths in stadia; the circumference attributed to him, 252,000 stadia, corresponds to 700 stadia per degree. Hipparchus (c. 190–120 BCE) subsequently connected terrestrial position more closely with astronomy. Latitude information attributed to him includes the length of the longest day, the Sun’s noon altitude at the winter solstice, and ratios between a vertical gnomon and its shadow. These were alternative quantities from which latitude could be inferred. Neither scholar’s geographical treatise survives intact: much of the evidence comes through quotations and discussions in Strabo’s Geography. Consequently, reconstructions of their coordinate systems must distinguish explicitly transmitted measurements from inferred procedures and numerical inconsistencies introduced during transmission. (arxiv.org)

In the second century CE, Ptolemy made geographical coordinates the organizing framework of his Geography, supplying a catalogue of locations and instructions for representing them on maps. His longitudes were reckoned from a western reference meridian associated with the Fortunate Islands, rather than Greenwich. The precise modern identification of those islands and the geographical placement of his zero meridian remain subjects of historical investigation; the ancient name should not be treated as an unambiguous modern survey reference. Ptolemy’s transmitted coordinate tables thus provide a different kind of evidence from the fragmentary geographical writings of Eratosthenes and Hipparchus. (arxiv.org)

Longitude became an especially consequential maritime problem because navigators needed a dependable east–west position during voyages, not merely an estimate accumulated from course and distance. The British Longitude Act of 1714 offered financial rewards for a practical solution; its stated purposes included safer and quicker voyages, preservation of ships and lives, and benefits to trade. (rmg.co.uk) John Harrison pursued the timekeeping approach: a clock retaining reference time could be compared with locally determined time to establish longitude. His first marine timekeeper, H1, was presented in 1735 and tested on a voyage to Lisbon in 1736. Its interconnected balances addressed the difficulty of maintaining regular motion aboard a moving ship. After H2 and H3, Harrison developed H4, a watch-shaped instrument with a rapidly oscillating balance. In 1761 his son William carried H4 on a trial voyage to Jamaica. That trial did not immediately settle the reward question; a further Barbados trial in 1764 preceded the commissioners’ confirmation that H4 met the Act’s most stringent accuracy limits. (rmg.co.uk)

Astronomical longitude determination developed alongside marine timekeepers. In the lunar-distance method, the measured angular separation of the Moon from the Sun or a suitable star was compared with predicted separations at a reference meridian. The changing lunar position thereby supplied reference time without requiring a clock to preserve it throughout the voyage. Nevil Maskelyne, appointed Astronomer Royal in 1765, organized the calculations and publication needed to make this method more usable at sea. (rmg.co.uk) The first Nautical Almanac and Astronomical Ephemeris appeared in 1766 with data for 1767, including lunar distances tabulated at three-hour intervals. These tables reduced the navigator’s computational burden, while their Greenwich reference helped establish that meridian in navigational practice. (aa.usno.navy.mil)

The International Meridian Conference, held in Washington, D.C. in October 1884, selected the meridian through Greenwich Observatory’s transit instrument as the common initial meridian. The vote was 22 in favor and one against, with France and Brazil abstaining. Existing practice materially supported the choice: a shipping table presented to the conference assigned Greenwich-based navigation to 65 percent of ships and 72 percent of shipping tonnage—two distinct measures sometimes conflated in later accounts. The conference also recommended counting longitude east and west to $180^\circ$, and proposed a universal day beginning at mean midnight at Greenwich and numbered from zero to twenty-four hours. Its resolutions expressly preserved the use of local or other standard time where desirable; they did not impose a worldwide system of civil time-zone boundaries. (gutenberg.org)

Map Projections and Geospatial Applications

A map projection converts geographical coordinates $(\varphi,\lambda)$ on a reference sphere or ellipsoid into plane coordinates, commonly expressed as eastings and northings. Its forward equations have the general form $x=f(\varphi,\lambda)$, $y=g(\varphi,\lambda)$; an inverse conversion recovers latitude and longitude where the projection is invertible. The equations depend on parameters such as the central meridian, latitude of origin, scale factor, and coordinate offsets. The graticule consists of projected parallels and meridians, whereas a projected coordinate grid consists of lines of constant easting or northing, usually labelled in metres or feet. These are distinct frameworks: curved graticule lines can cross a rectangular coordinate grid. PROJ’s proj and invproj programs implement forward and inverse projection conversions; changing the underlying datum is a separate operation. (ngs.noaa.gov)

The normal cylindrical Mercator projection represents meridians as equally spaced vertical straight lines and parallels as horizontal straight lines whose spacing increases towards the poles. For a sphere of radius $R$, with central meridian $\lambda_0$, true scale at the equator, and angles in radians, its equations are $x=R(\lambda-\lambda_0)$ and $y=R\ln\tan(\pi/4+\varphi/2)$. Mercator is conformal: it preserves angles and infinitesimal shapes, not area. Its local linear scale factor is $1/\cos\varphi$, so at $60^\circ$ latitude lengths are enlarged by a factor of two and local areas by four relative to equatorial scale. A constant-bearing route, or rhumb line, appears straight; this property explains Mercator’s widespread use for nautical charts. (proj.org)

Conic and azimuthal projections arrange the graticule differently. In the normal Lambert conformal conic projection, meridians are straight lines converging towards a common point, while parallels are concentric circular arcs. In its two-standard-parallel form, scale is true along the selected parallels, making it useful for mid-latitude regions extending predominantly east–west; distortion increases away from those parallels. In a spherical polar azimuthal equidistant projection, meridians radiate from the pole and parallels form equally spaced concentric circles. Distances and azimuths measured from the centre are preserved, but distances between arbitrary off-centre points are not. Thus, preserving angles, areas, or selected distances requires different projection geometries rather than a single universally distortion-free layout. (pro.arcgis.com)

Applications frequently retain geographical coordinates for identifying positions while using projected coordinates for display or local calculations. NOAA’s nautical-chart reporting interface, for example, accepts a position as 36.729142N 122.204318W or in degrees and decimal minutes. For surveying and engineering, the United States State Plane Coordinate System, originally developed in the 1930s, supplies regional conformal projections and plane coordinates; NOAA’s SPCS83 software converts between NAD 83 latitude–longitude positions and State Plane eastings and northings. Satellite-navigation software likewise separates position determination from coordinate presentation. Garmin’s GPS12MAP manual provides latitude–longitude, UTM/UPS, and regional-grid display options: selecting a grid changes the representation of the position, not the satellite signal being received. (nauticalcharts.noaa.gov)

In a geographic information system, geographic coordinates also provide an exchange representation independent of a particular projected display. RFC 7946, published in August 2016, specifies GeoJSON positions in WGS 84 decimal degrees, ordered longitude first and latitude second. Scientific gridded datasets require additional structure. A regular latitude–longitude grid can use separate one-dimensional axes, but a projected or curvilinear grid may require two-dimensional auxiliary arrays lat(j,i) and lon(j,i) locating every horizontal grid point. The CF conventions illustrate this with a temperature array T(lev,yc,xc) having 18 vertical levels and a $64\times128$ horizontal grid; its coordinates attribute associates the measurements with the latitude and longitude arrays. Projected xc and yc coordinates can be retained alongside them, allowing software to address the same samples through either grid coordinates or geographical locations. (rfc-editor.org)

Precision, Singularities, and Interpretation Errors

Displayed decimal precision describes the resolution of a coordinate’s numerical representation, not its positional accuracy. Accuracy concerns agreement with the position being measured; additional digits do not establish smaller measurement uncertainty. For example, rounding latitude to five decimal places gives a step of $0.00001^\circ$, approximately 1.11 metres on the ground, and a maximum rounding displacement of approximately 0.56 metres in the north–south direction. These are calculated rounding limits, not guarantees about the original observation. Using a spherical approximation, the corresponding east–west step is $1.11\cos\varphi$ metres: approximately 1.11 metres at the equator, 0.56 metres at $60^\circ$ latitude, and 0.19 metres at $80^\circ$. Thus, the same number of longitude digits represents different ground resolutions at different latitudes. Adding trailing zeros to a metre-level observation cannot make it centimetre-accurate. (jcgm.bipm.org)

The poles are coordinate singularities because every meridian converges there. At latitude $+90^\circ$ or $-90^\circ$, changing longitude does not identify a different surface location. In the unit-sphere representation, $x=\cos\varphi\cos\lambda$ and $y=\cos\varphi\sin\lambda$ both vanish, leaving the same polar point for every $\lambda$. Longitude therefore has no unique geometrical value at the pole, even when a data format stores a conventional numerical value. Near a pole, a large longitude change can correspond to a small displacement; conversely, longitude can change rapidly along a short track passing close to the pole. This is a limitation of the angular coordinate representation, not a physical discontinuity in Earth’s surface. Earth-centred Cartesian coordinates avoid this particular ambiguity. (oc.nps.edu)

Longitude also has a representational discontinuity at the antimeridian: $+180^\circ$ and $-180^\circ$ denote the same meridian. Consequently, positions at $179^\circ$ east and $179^\circ$ west differ by $2^\circ$ along the shorter angular interval, rather than the $358^\circ$ obtained by ordinary subtraction. Their circular midpoint lies on the antimeridian, not at Greenwich. Digital geometries require explicit treatment of this wraparound. RFC 7946, published in August 2016, recommends cutting antimeridian-crossing GeoJSON geometries into parts that do not cross it, using multipart geometry types where appropriate. Its Fiji bounding-box example, [177.0, -20.0, -178.0, -16.0], spans $5^\circ$ of longitude although its eastern longitude is numerically smaller than its western longitude. Reordering those bounds produces the complementary $355^\circ$ span. (oc.nps.edu)

Coordinate-order and sign errors can remain syntactically valid. GeoJSON uses longitude followed by latitude: [-73, 40] means $73^\circ$ west, $40^\circ$ north. Reversing it to [40, -73] gives a different, still valid location. Dropping a western or southern minus sign likewise changes hemispheres without necessarily violating coordinate ranges. (rfc-editor.org) Incompatible geodetic datums cause a different error: numbers are interpreted against the wrong reference framework. NOAA’s National Geodetic Survey documents systematic shifts of dozens of metres across the United States between NAD 27 and the original 1986 realization of NAD 83. Changing a dataset’s datum label without transforming its coordinates does not apply those shifts. Precise records also identify the reference-frame realization and coordinate epoch, since coordinates can change with time. (geodesy.noaa.gov)

A third coordinate introduces another possible misinterpretation: ellipsoidal height is not interchangeable with elevation referred to a vertical datum. GeoJSON’s optional height is measured in metres relative to the WGS 84 reference ellipsoid. (rfc-editor.org) For compatible height reference systems, the conventional relationship is $H=h-N$, where $h$ is ellipsoidal height, $H$ is orthometric height, and $N$ is geoid height. NOAA gives the illustrative near-seashore case $H=+1$ metre and $h=-30$ metres, corresponding to $N=-31$ metres. A negative ellipsoidal height therefore does not necessarily indicate a position below sea level. Conversion depends on the applicable geoid model and datum definitions, rather than a universal constant offset. (geodesy.noaa.gov)

References

1. Coordinatesoc.nps.edu
2. What is latitude?oceanservice.noaa.gov
3. AT-SEA MONITORING PROGRAM MANUALrepository.library.noaa.gov
4. NOAA Reprint ofrepository.library.noaa.gov
5. Geocentric Latitude — PROJ documentationproj.org
6. NGA Geomatics - WGS 84earth-info.nga.mil
7. Geodesy for the Laymangeodesy.noaa.gov
8. National Geodetic Survey DEFLEC18geodesy.noaa.gov
9. Latitude and Longitude - Cartographic Resources for Genealogical Research: Eastern Europe and Russiaguides.loc.gov
10. Royal Museums Greenwich: Secondary School Guidermg.co.uk
11. The Equation of Timeaa.usno.navy.mil
12. What is the Prime Meridian, and why is it in Greenwich?rmg.co.uk
13. Why the Greenwich meridian movedlink.springer.com
14. INVERSE/FORWARD/INVERS3D/FORWRD3D Program Documentationgeodesy.noaa.gov
15. Data Item: Latitude and Longitudegeodesy.noaa.gov
16. DSWorld Upload Web: Add Single GPS Positiongeodesy.noaa.gov
17. RFC 7946: The GeoJSON Formatrfc-editor.org
18. datum namespace — PROJ documentationproj.org
19. PROJJSON — PROJ documentationproj.org
20. National Adjustment 2011 FAQgeodesy.noaa.gov
21. North American Datum of 1983 (NAD 83)geodesy.noaa.gov
22. ITRF | Itrf2020itrf.ign.fr
23. ITRF | Moreitrf.ign.fr
24. Geodetic transformation — PROJ documentationproj.org
25. Helmert transform — PROJ documentationproj.org
26. Conversions — PROJ documentationproj.org
27. The National Geodetic Survey Improves the National Spatial Reference System with Simultaneous Major Product Releasesgeodesy.noaa.gov
28. EXPLANATION OF NAVIGATION TABLESthenauticalalmanac.com
29. Chapter 28: Emergency Navigationthenauticalalmanac.com
30. Geodesy for the Laymanngs.noaa.gov
31. Appendix D: Navigational Coordinatesthenauticalalmanac.com
32. GMT and Longitude by Lunar Distance: Two Methods Compared From a Practitioner's Point of Viewcambridge.org
33. History of Royal Observatory Greenwichrmg.co.uk
34. Triangulation, Traverse and Horizontal Networksngs.noaa.gov
35. Network Adjustments — OPUS Projects User Guidebeta.ngs.noaa.gov
36. The Future of the Global Positioning Systemgovinfo.gov
37. NGA Geomatics — WGS 84earth-info.nga.mil
38. GPS and Earth Orientation Productsnga.mil
39. OPUS: the Online Positioning User Servicengs.noaa.gov
40. Real World Coordinates Page 2oc.nps.edu
41. Geodesic (Java implementation of GeographicLib 2.1 API)geographiclib.sourceforge.io
42. gc_latlonncl.ucar.edu
43. css2cncl.ucar.edu
44. Analysis of the latitudinal data of Eratosthenes and Hipparchusarxiv.org
45. The western coast of Africa in Ptolemy's Geography and the location of his prime meridianhgss.copernicus.org
46. BOARD OF LONGITUDE Explorersrmg.co.uk
47. Longitude found - the story of Harrison's timekeepersrmg.co.uk
48. Longitude found: Nevil Maskelyne and the lunar methodrmg.co.uk
49. History of The Nautical Almanacaa.usno.navy.mil
50. International Conference Held at Washington for the Purpose of Fixing a Prime Meridian and a Universal Day. October, 1884. Protocols of the Proceedingsgutenberg.org
51. State Plane Coordinate System (SPCS) - Tools - National Geodetic Surveyngs.noaa.gov
52. proj — PROJ documentationproj.org
53. Applications — PROJ documentationproj.org
54. Mercator — PROJ documentationproj.org
55. U.S. Office of Coast Survey: Nautical Cartographynauticalcharts.noaa.gov
56. Lambert conformal conic—ArcGIS Pro Documentationpro.arcgis.com
57. Map Projections Used by the USGSpubs.usgs.gov
58. U.S. Office of Coast Survey: ASSISTnauticalcharts.noaa.gov
59. SPCS83 User's Guidegeodesy.noaa.gov
60. GPS12MAP Owner's Manualstatic.garmin.com
61. RFC 7946: The GeoJSON Formatrfc-editor.org
62. Measurement accuracy — International Vocabulary of Metrologyjcgm.bipm.org
63. Measurement result — International Vocabulary of Metrologyjcgm.bipm.org
64. Converting GPS Height into NAVD88 Elevation with the GEOID96 Geoid Height Modelgeodesy.noaa.gov