Since a map is a representation, the original shape of the represented subject must first be defined. An important branch of cartography, geodesy studies the Earth shape and how it is related to its surface's features.
Geoid, from the Greek for "Earth-shaped", is the common definition of our world's shape. This recursive description is necessary because no simple geometric shape matches the Earth:
Taken in account, those factors greatly complicate the cartographer's job but, depending on the task, some irregularities can be ignored. For instance, although important locally, terrain levels are minuscule in planetary scale: the tallest land peak stands less than 9km above sea level, or nearly 1/1440 of Earth diameter; the depth of the most profound sea abyss is roughly 1/1150 diameter.
For maps covering very large areas, especially worldwide, the Earth may be assumed perfectly spherical, since any shape imprecision is dwarfed by unavoidable errors in data and media resolution. This assumption holds for most of this document. Conversely, for very small areas terrain features dominate and measurements can be based on a flat Earth.
For highly precise maps of smaller regions, the basic ellipsoidal shape can not be ignored. A geodetic datum is a set of parameters (including axis lengths and offset from true center of the Earth) defining a reference ellipsoid. For each mapped region, a different datum can be carefully chosen so that it best matches average sea level, therefore terrain features. Thus, data acquisition for a map involves surveying, or measuring heights and distances of reference points as deviations from a specific datum (a delicate task: due to mentioned irregularities, gravity - and therefore plumb bobs and levers - is not always aligned towards the center of the Earth).
Several standard datums were adopted for regional or national maps. International datums do exist, but can not fit any particular area as well as a local one.
|In this wooden sphere
(with an octant removed for clarity), the copper
arrows define the coordinate system's origins. One
arrow lies over the north-south geographical axis, while
the other is defined by convention.
The white surface "point" is located using two angles or coordinates, called the longitude () and latitude (). Every point has a single counterpart at maximum distance on the opposite side, called its antipode (not shown here).
|A graticule is a spherical
grid of coordinate lines over the planetary surface, comprising
circles on planes normal (perpendicular) to the
north-south axis, called parallels (red) and semicircular arcs
with that axis as chord, called
True to their name, no parallels ever cross one another, while all meridians meet at each geographic pole. Every parallel crosses every meridian at an angle of 90°. This is and other properties help assessing distortion.
A natural reference, the longest parallel divides the Earth in two equal hemispheres, north and south; thus its name, Equator. Four other important parallels are defined by astronomical constraints. The geographical north-south axis is actually tilted slightly less than 23.5° from the plane of the Earth's orbit around the sun. This accounts for the different seasons and different lengths of day and night periods throughout the year.
|Schematic cross section of Earth's orbit. AT is the axial tilt, about 23.5°|
Every year about December 21st, the solar rays fall vertically upon a parallel near 23.5°S. That is the longest day in the southern hemisphere (notice how most of it is exposed to the sun, so that date is called the southern summer solstice), but the shortest day in the northern hemisphere (therefore winter solstice); not only shorter daylight periods but a shallower angle of incidence of solar rays explain the lower temperatures north of Equator.
Near June 21st, a similar phenomenon happens along the parallel opposite North. By definition, these two parallels encircle the torrid or tropical zone; they are named after the zodiacal constellations where the sun is at those dates, thus Tropic of Capricorn (south) and Tropic of Cancer (north). In regions south of the Tropic of Capricorn the sun appears to run always north of the observer, even at noon; in places north of the Tropic of Cancer the sun runs always south of the observer, while in tropical regions the sun appears sometimes south, sometimes north, depending on the season.
Subtracting the axial tilt from 90° we get the latitudes of the Arctic (about 66.5°N) and Antarctic (about 66.5°S) polar circles. At December 21 the sun does not set at the Antarctic circle for a full day. Going south, we get even longer consecutive daylight periods, up to six months at the pole. There are correspondingly long nights at the Antarctic winter. Of course, the same occurs at the northern latitudes, with a six-month offset.
Points on the same parallel suffer similar rates of exposure to the sun, therefore are prone to similar climates (disregarding other factors like altitude, wind/sea conditions and terrain).
A point's latitude can be inferred from the sun's angle above the horizon at noon (the moment when the sun appears highest at the sky and a pole projects its shortest shadow). Sailors use instruments like the sextant for that job.
|East-west distances between points separated by one minute of solar time at different latitudes. Distance is zero at poles, where one "sees" every moment of the day at any time.|
Unlike the Equator, there's no easily defined
prime or "main"
meridian, which was fixed (mainly by political consensus) in 1884 over
the Royal Observatory in Greenwich, near London, UK. This
choice's only obvious advantage is setting the opposite
meridian (near or at the left or right edges of many world maps) away
from most inhabited areas. That opposite meridian is the
base of the international date line, which
separates world halves in two different days. Again, this line
is somewhat irregular in order to keep national
territories (mostly Pacific islands) in a single timezone.
Compared to finding a point's latitude, getting its longitude is a much more involved procedure, usually comparing the time separating the noon at the reference meridian and at the point in question.