chapter

one the earth and earth coordinates

The Earth as a sphere

The graticule
Parallels and meridians
Latitude and longitude
Prime meridians
The Earth as an oblate ellipsoid
Different ellipsoids
Geodetic latitude
Geodetic longitude
Determining geodetic latitude and longitude
Properties of the graticule
Circumference of the authalic and other spheres
Spacing of parallels
Converging meridians
Great and small circles
Quadrilaterals
Graticule appearance on maps
Small-scale maps
Large-scale maps
Geodetic latitude and longitude on large-scale maps
Horizontal reference datums
The Earth as a geoid
Vertical reference datums
Selected readings
 1
The earth and earth coordinates

Of all the jobs maps do for you, one stands out. They tell you where things are and let you communicate
this information efficiently to others. This, more than any other factor, accounts for the widespread
use of maps. Maps give you a superb locational reference system —a way of pinpointing the position of
things in space.

There are many ways to determine the location of a feature shown on a map. All of these begin with
defining a geometrical figure that approximates the true shape and size of the earth. This figure is
either a sphere or an oblate ellipsoid (slightly flattened sphere) of precisely known dimensions. Once
the dimensions of the sphere or ellipsoid are defined, a graticule of east–west lines called parallels and
north–south lines called meridians is draped over the sphere or ellipsoid. The angular distance of a
parallel from the equator and a meridian from what we call the prime meridian gives us the latitude and
longitude coordinates of a feature. The locations of elevations measured relative to an average gravity
or sea-level surface called the geoid can then be defined by three-dimensional coordinates.
6 Chapter 1 The earth and earth coordinates

The Earth as a sphere circumference. The story that has come down to us is of
Eratosthenes reading an account of a deep well at Syene
We have known for over 2,000 years that the earth is near modern Aswan about 500 miles (800 kilometers)
spherical in shape. We owe this idea to several ancient south of Alexandria. The well’s bottom was illuminated
Greek philosophers, particularly Aristotle (fourth by the sun only on June 21, the day of the summer
century BC), who believed that the earth’s sphericity solstice. He concluded that the sun must be directly
could be proven by careful visual observation. Aristotle overhead on this day, with rays perpendicular to the level
noticed that as he moved north or south the stars were ground (figure 1.1). Then he reasoned brilliantly that if
not stationary—new stars appeared in the northern the sun’s rays were parallel and the earth was spherical,
horizon while familiar stars disappeared to the south. a vertical column like an obelisk should cast a shadow
He reasoned that this could occur only if the earth were in Alexandria on the same day. Knowing the angle of
curved north to south. He also observed that departing the shadow would allow the earth’s circumference to be
sailing ships, regardless of their direction of travel, always measured if the north–south distance to Syene could be
disappeared from view hull first. If the earth were flat, determined. The simple geometry involved here is “if two
the ships would simply get smaller as they sailed away. parallel lines are intersected by a third line, the alternate
Only on a sphere would hulls always disappear first. His interior angles are equal.” From this he reasoned that
third observation was that a circular shadow is always the shadow angle at Alexandria equaled the angular
cast by the earth on the moon during a lunar eclipse, difference at the earth’s center between the two places.
something that would occur only if the earth were The story continues that on the next summer solstice
spherical. These arguments entered the Greek literature Eratosthenes measured the shadow angle from an obelisk
and persuaded scholars over the succeeding centuries in Alexandria, finding it to be 7°12΄, or 1/50 th of a circle.
that the earth must be spherical in shape. Hence, the distance between Alexandria and Syene
Determining the size of our spherical earth was a is 1/50 th of the earth’s circumference. He was told that
daunting task for our ancestors. The Greek scholar Syene must be about 5,000 stadia south of Alexandria,
Eratosthenes, head of the then-famous library and since camel caravans traveling at 100 stadia per day took
museum in Alexandria, Egypt, around 250 BC, made 50 days to make the trip between the two cities. From
the first scientifically based estimate of the earth’s this distance estimate, he computed the earth’s circum-
ference as 50 × 5,000 stadia, or 250,000 stadia. In Greek
times a stadion varied from 200 to 210 modern yards
(182 to 192 meters), so his computed circumference was
Interior angle
7°12’
somewhere between 28,400 and 29,800 modern statute
Pa un’s ne 2

miles (45,700 and 47,960 kilometers), 14 to 19 percent

ra
ll
S Ju

el
on

Obelisk at greater than the current value of 24,874 statute miles

lin
ra 1

Alexandria
es
ys

(40,030 kilometers).
5,
0
00
st

We now know that the error was caused by overesti-

ad
ia

mation of the distance between Alexandria and Syene,

Well at
Syene and to the fact that they are not exactly north–south of
each other. However, his method is sound mathemati-
cally and was the best circumference measurement until
the 1600s. Equally important, Eratosthenes had the idea
that careful observations of the sun would allow him to
determine angular differences between places on earth,
an idea that we shall see was expanded to other stars
and recently to the Global Positioning System (GPS), a
“constellation” of 24 earth-orbiting satellites that make it
possible for people to pinpoint geographic location and
7°12’ Interior angle
elevation with a high degree of accuracy using ground
Earth’s center
receivers (see chapter 14 for more on GPS).

Figure 1.1 Eratosthenes’s method for measuring the earth’s

circumference.
 The graticule 7

The graticule the equator to 90° at the poles. The letters N and S, such
as 45°N for Fossil, Oregon, are used to indicate north
Once the shape and size of the earth were known, and south latitude. Longitude is the angle, measured
mapmakers required some system for defining locations along the equator, between the intersection of the refer-
on the surface. We are again indebted to ancient Greek ence meridian, called the prime meridian, and the point
scholars for devising a systematic way of placing reference where the meridian for the feature of interest intersects
lines on the spherical earth. the equator. The numerical range of longitude is from 0°
to 180° east and west of the prime meridian, twice as long
Parallels and meridians as parallels. East and west longitudes are labeled E and W,
Astronomers before Eratosthenes had placed on maps so Fossil, Oregon, has a longitude of 120°W.
horizontal lines marking the equator (forming the circle Putting latitude and longitude together into what is
around the earth that is equidistant from the north and called a geographic coordinate (45°N, 120°W) pin-
south poles) and the tropics of Cancer and Capricorn points a place on the earth’s surface. There are several
(marking the northernmost and southernmost positions ways to write latitude and longitude values. The oldest
where the sun is directly overhead on the summer and is the Babylonian sexagesimal system of degrees (°),
winter solstices, such as Syene). Later the astronomer and minutes (΄), and seconds (˝), where there are 60 minutes
mathematician Hipparchus (190–125 BC) proposed in a degree and 60 seconds in a minute. The latitude
that a set of equally spaced east–west lines called parallels and longitude of the capitol dome in Salem, Oregon, is
be drawn on maps (figure 1.2). To these he added a set 44°56΄18˝N, 123°01΄47˝W, for example.
of north–south lines called meridians that are equally Sometimes you will see longitude west of the prime
spaced at the equator and converge at the north and meridian and latitude south of the equator designated
south poles. We now call this arrangement of parallels with a negative sign instead of the letters W and S. Lati-
and meridians the graticule. Hipparchus’s numbering tude and longitude locations can also be expressed in
system for parallels and meridians was and still is called decimal degrees through the following equation:
latitude and longitude.
Decimal degrees = dd + mm / 60 + ss / 3600
Latitude and longitude where dd is the number of whole degrees, mm is the
Latitude on the spherical earth is the north–south number of minutes, and ss is the number of seconds. For
angular distance from the equator to the place of interest example:
(figure 1.3). The numerical range of latitude is from 0° at
44°56΄18˝ = 44 + 56 / 60 + 18 / 3600 = 44.9381°

Parallels Meridians
North Pole

Tropic of Cancer

Equator

orn
Tropic of Capric

Figure 1.2 Parallels and meridians.

8 Chapter 1 The earth and earth coordinates

Pole in Greenwich, a suburb of London. This is called the

Greenwich meridian.
Fossil, Oregon
45°N, 120°W You may occasionally come across a historical map

Pr
using one of the prime meridians in table 1.1, at which

im
e
time knowing the angular difference between the

Me
r id
prime meridian used on the map and the Greenwich

ian
meridian becomes very useful. As an example, you
might see in an old Turkish atlas that the longitude of
45°

Center
Seattle, Washington, is 151°16΄W (based on the Istanbul
meridian), and you know that the Greenwich longitude of
120° Seattle is 122°17΄ W. You can determine the Greenwich
longitude of Istanbul through subtraction:
Equator

151°16΄ – 122°17΄ = 28° 59΄E

The computation is done more easily in decimal degrees
as described earlier:

Figure 1.3 Latitude and longitude on the sphere allow the 151.27 – 122.28 = 28.99 degrees
positions of features to be explicitly identified.

Decimal degrees are often rounded to two decimal places,

so that the location of the Oregon state capitol dome Amsterdam, Netherlands 4°53’01”E
would be written in decimal degrees as 44.94, –123.03. Athens, Greece 23°42’59”E
If we can accurately define a location to the nearest 1 Beijing, China 116°28’10”E
second of latitude and longitude, we have specified its Berlin, Germany 13°23’55”E
location to within 100 feet (30 meters) of its true loca- Bern, Switzerland 7°26’22”E
tion on the earth.
Brussels, Belgium 4°22’06”E
Copenhagen, Denmark 12°34’40”E
Prime meridians
Ferro, Canary Islands 17°40’00”W
The choice of prime meridian (the meridian at 0° used as
Helsinki, Finland 24°57’17”E
the reference from which longitude east and west is mea-
sured) is entirely arbitrary because there is no physically Istanbul, Turkey 28°58’50”E
definable starting point like the equator. In the fourth Jakarta, Indonesia 106°48’28”E
century BC Eratosthenes selected Alexandria as the Lisbon, Portugal 9°07’55”W
starting meridian for longitude, and in medieval times Madrid, Spain 3°41’15”W
the Canary Islands off the coast of Africa were used since Moscow, Russia 37°34’15”E
they were then the westernmost outpost of western civili- Oslo, Norway 10°43’23”E
zation. In the eighteenth and nineteenth centuries, many
Paris, France 2°20’14”E
countries used their capital city as the prime meridian
for the nation’s maps, including the meridian through Rio de Janeiro, Brazil 43°10’21”W

the center of the White House in Washington, D.C., for Rome (Monte Mario), Italy 12°27’08”E
early nineteenth-century maps (see table 1.1 for a list- St. Petersburg, Russia 30°18’59”E
ing of historical prime meridians). You can imagine the Stockholm, Sweden 18°03’30”E
confusion that must have existed when trying to locate Tokyo, Japan 139°44’41”E
places on maps from several countries. The problem was Washington, D.C., USA 77°02’14”W
eliminated in 1884 when the International Meridian
Conference selected as the international standard the Table 1.1 Prime meridians used previously on foreign
British prime meridian—defined by the north–south maps, along with longitudinal distances from the Greenwich
optical axis of a telescope at the Royal Observatory meridian.
 The earth as an oblate ellipsoid 9

The earth as an oblate from 0 at the center of the earth to a maximum of around
ellipsoid 1,040 miles per hour (1,650 kilometers per hour) at the
equator. This means that the total centrifugal force on
the column is quite large. From the pole to the equator
Scholars assumed that the earth was a perfect sphere there will be a steady increase in centrifugal force. We see
until the 1660s when Sir Isaac Newton developed the this at the 60th parallel where the column would be half
theory of gravity. Newton thought that mutual gravita- the earth’s radius in length and the velocity would be half
tion should produce a perfectly spherical earth if it were that at the equator.
not rotating about its polar axis. The earth’s 24-hour Newton noted that these outward centrifugal forces
rotation, however, introduces outward centrifugal forces counteract the inward pull of gravity, so the net inward
perpendicular to the axis of rotation. The amount of force decreases progressively from the pole to the equator.
force varies from zero at each pole to a maximum at the The column from the center of the earth to the equator
equator, obeying the following equation: extended outward slightly because of the increased cen-
trifugal force and decreased inward gravitational force. A
centrifugal force = mass × velocity 2 × distance from the
similar column from the center of the earth to the pole
axis of rotation
experiences zero centrifugal force and hence does not have
To understand this, imagine a very small circular disk the this slight extension. Slicing the earth in half from pole
diameter of a dinner plate (about 10 inches or 25 centi- to pole would then reveal an ellipse with a slightly shorter
meters) at the pole and two very thin horizontal cylin- polar radius and slightly longer equatorial radius; we call
drical columns, one from the center of the earth to the these radii the semiminor and semimajor axes, respec-
equator (along the equatorial axis) and the second per- tively (figure 1.5). If we rotate this ellipse 180 degrees
pendicular from the polar axis (the axis from the center about its polar axis, we obtain a three-dimensional solid
of the earth to the poles) to the 60th parallel (figure 1.4). that we call an oblate ellipsoid. In the 1730s, scientific
The disk at the pole has a tiny mass, but both the veloc- expeditions to Ecuador and Finland measured the length
ity and distance from the axis are very small, so the cen- of a degree of latitude at the equator and near the Arctic
trifugal force (the apparent force caused by the inertia
of the body that draws a rotating body away from the
North Pole
center of rotation) is nearly 0. The column to the equator
is the earth’s radius in length, and the velocity increases Semiminor axis
Polar radius

Zero centrifugal force

520 mph
60 Small ce
°N
-half ntrifugal
One radius force
Ea r t h ’s Equatorial radius
Zero centrifugal force

Semimajor axis

30° Equatorial radius is about

N 21 kilometers (13 miles), or 0.3%
longer than the polar radius

Large centrufiga 1,040

l force
radius
Earth’s mph

Equator

South Pole

Figure 1.5 The form of the oblate ellipsoid was determined

by measurements of degrees at different latitudes
beginning in the 1730s. Its equatorial radius was about
13 miles (21 kilometers) longer than its polar radius. The
Figure 1.4 A systematic increase in centrifugal force north–south slice through the earth’s center in this figure
from the pole to equator causes the earth to be an oblate is true to scale, but our eye cannot see the deformation
ellipsoid. because it is so minimal.
10 Chapter 1 The earth and earth coordinates

Equatorial
Name Date radius (km) Polar radius (km) Areas of use
WGS 84 1984 6,378.137 6,356.75231 Worldwide
GRS 80 1980 6,378.137 6,356.7523 Worldwide (NAD 83)
Australian 1965 6,378.160 6,356.7747 Australia
Krasovsky 1940 6,378.245 6,356.863 Soviet Union
International 1924 6,378.388 6,356.9119 Remainder of world not covered by older ellipsoids
(European Datum 1950)
Clarke 1880 6,378.2491 6,356.5149 France; most of Africa

Clarke 1866 6,378.2064 6,356.5838 North America (NAD 27)

Bessel 1841 6,377.3972 6,356.079 Central Europe; Sweden; Chile; Switzerland; Indonesia
Airy 1830 6,377.5634 6,356.2569 Great Britain; Ireland
Everest 1830 6,377.2763 6,356.0754 India and the rest of South Asia

Table 1.2 Historical and current oblate ellipsoids.

Circle, proving Newton correct. These and additional radius of 6,378.137 kilometers (3,963.191 miles) and a
meridian-length measurements in following decades polar radius of 6,356.752 kilometers (3,949.903 miles).
for other parts of the world allowed the semimajor and The oblate ellipsoid is important to us because
semiminor axes of the oblate ellipsoid to be computed parallels are not spaced equally as on a sphere, but vary
by the early 1800s, giving about a 13-mile (21-kilometer) slightly in spacing from the pole to the equator. This
difference between the two, only one-third of one is shown in figure 1.6, a cross section of a greatly flat-
percent. tened oblate ellipsoid. Notice that near the pole the
ellipse curves less than near the equator. We say that on
Different ellipsoids an oblate ellipsoid the radius of curvature (the mea-
During the nineteenth century better surveying sure of how curved the surface is) is greatest at the pole
equipment was used to measure the length of a
degree of latitude on different continents. From these
measurements slightly different oblate ellipsoids varying Pole

by only a few hundred meters in axis length best fit

Larger ra

the measurements. Table 1.2 is a list of these ellipsoids,

along with their areas of usage. Note the changes in
ellipsoid use over time. For example, the Clarke 1866
dius of c

ellipsoid was the best fit for North America in the

nineteenth century and hence was used as the basis for 15°
urvature

Equator

latitude and longitude on topographic and other maps 15°

Smalle
produced in Canada, Mexico, and the United States r rad
of curva ius
ture
from the late 1800s to about the late 1970s. By the
1980s, vastly superior surveying equipment coupled with
millions of observations of satellite orbits allowed us to
determine oblate ellipsoids that are excellent average fits
for the entire earth. Satellite data are important because
the elliptical shape of each orbit monitored at ground
receiving stations mirrors the earth’s shape. The most
recent of these, called the World Geodetic System of Figure 1.6 This north–south cross section through the
1984 (WGS 84), replaced the Clarke 1866 ellipsoid in center of a greatly flattened oblate ellipsoid shows that
North America and is used as the basis for latitude and a larger radius of curvature at the pole results in a larger
longitude on maps throughout the world. You’ll see in ground distance per degree of latitude relative to the
table 1.2 that the WGS 84 ellipsoid has an equatorial equator.
 The earth as an oblate ellipsoid 11

and smallest at the equator. The north–south distance Geocentric Geodetic Geocentric Geodetic
between two points on the surface equals the radius of 0° 0.000° 50° 50.126°
curvature times the angular difference between them. 5 5.022 55 55.120
For example, the distance between two points one 10 10.044 60 60.111
degree apart in latitude between 0° and 1° at the equa- 15 15.064 65 65.098
tor is 68.703 miles (110.567 kilometers), shorter than the 20 20.083 70 70.082
69.407 miles (111.699 kilometers) between two points
25 25.098 75 75.064
at 89° and 90° north latitude. This example shows that
30 30.111 80 80.044
the spacing of parallels decreases slightly from the pole
35 35.121 85 85.022
to the equator.
40 40.126 90° 90.000°
Geodetic latitude 45° 45.128°
Geodetic latitude is defined as the angle made by the Table 1.3 Geocentric and corresponding geodetic latitudes
horizontal equator line and a line perpendicular to the (WGS 84) at 5o increments.
ellipsoidal surface at the parallel of interest (figure 1.7).
Geodetic latitude differs from latitude on a sphere because
of the unequal spacing of parallels on the ellipsoid. Lines Geodetic longitude
perpendicular to the ellipsoidal surface only pass through It turns out that there is no need to make any distinc-
the center of the earth at the poles and equator, but all tion between geodetic longitude and geocentric longi-
lines perpendicular to the surface of a sphere will pass tude. While the definition of geocentric longitude is
through its center. This is why the latitude defined by mathematically different from geodetic longitude, the
these lines on a sphere is called geocentric latitude. end result is essentially the same. The angle between the
Defining geodetic latitude in this way means that line from a point on the surface of the earth to the center
geocentric and geodetic latitude are identical only at 0° of the earth and then to the prime meridian determines
and 90°. Everywhere else geocentric latitude is slightly the longitude. Because of the nature of the ellipsoidal
smaller than the corresponding geodetic latitude, as model, this turns out to be the same as the angle for
shown in table 1.3. Notice that the difference between geocentric longitude, which was described above.
geodetic and geocentric latitude increases in a sym-
metrical fashion from zero at the poles and equator to a
maximum of just under one-eighth of a degree at 45°.

Geocentric latitude Geodetic latitude

on a sphere on an oblate ellipsoid
d d

c c

45° 45°
a
a b b

Figure 1.7 Geocentric and geodetic latitudes of 45°. On a sphere, circular arc distance b–c is the same as circular arc
distance c–d. On the greatly flattened oblate ellipsoid, elliptical arc distance b–c is less than elliptical arc distance c–d. On
the WGS 84 oblate ellipsoid, arc distance b–c is 3,097.50 miles (4,984.94 kilometers) and arc distance c–d is 3,117.43 miles
(5,017.02 kilometers), a difference of about 20 miles (32 kilometers).
12 Chapter 1 The earth and earth coordinates

Determining geodetic dimmer star (Epsilon) is offset about 30 degrees below

latitude and longitude the center of the cross.
Geodetic longitude can be determined in a straight-
forward manner. As we saw earlier, the prime meridian
The oldest way to determine geodetic latitude and at 0° longitude passes through Greenwich, England.
longitude is with instruments for observing the positions Therefore, each hour difference between your time and
of celestial bodies. The essence of the technique is to that at Greenwich, called Greenwich mean time or
establish celestial lines of position (east–west, north– GMT, is roughly equivalent to 15° of longitude from
south) by comparing the predicted positions of celestial Greenwich. To determine geodetic longitude, compare
bodies with their observed positions. A handheld your local time with Greenwich mean time and multiply
instrument, called a sextant, historically was the tool by 15° of longitude for each hour of difference. The
used to measure the angle (or altitude) of a celestial body difficulty is that time is conventionally defined by broad
above the earth’s horizon (figure 1.8). Before GPS, it was time zones, not by the exact local time at your longitude.
the tool that nautical navigators used to find their way Your exact local time must be determined by celestial
using the moon, planets, and stars, including our sun. observations.
Astronomers study and tabulate information on the In previous centuries, accurately determining
actual motion of celestial bodies that helps to pinpoint longitude was a major problem in both sea navigation
latitude and longitude. Since the earth rotates on an axis and mapmaking. It was not until 1762 that a clock
defined by the north and south poles, stars in the north- portable enough to take aboard ship and accurate
ern hemisphere’s night sky appear to move slowly in a enough for longitude finding was invented by the
circle centered on Polaris (the North Star). The naviga- Englishman John Harrison. This clock, called a
tor needs only to locate Polaris to find north. In addition, chronometer, was set to Greenwich mean time before
because the star is so far away from the earth, the angle departing on a long voyage. The longitude of a distant
from the horizon to Polaris is the same as the latitude locale was found by noting the Greenwich mean time
(figure 1.9). at local noon (the highest point of the sun in the sky,
In the southern hemisphere, latitude is harder to found with a sextant). The time difference was simply
determine because four stars are used to interpolate due multiplied by 15 to find the longitude.
south. Because there is no equivalent to Polaris over the
south pole, navigators instead use a small constellation
called Crux Australis (the Southern Cross) to serve
the same function (figure 1.10). Finding south is more
complicated because the Southern Cross is a collection
of five stars that are part of the constellation Centaurus.
The four outer stars form a cross, while the fifth much
Courtesy of Dr. Bernie Bernard.

Center
of Earth

Figure 1.8 A sextant is used at sea to find latitude from the Figure 1.9 In the northern hemisphere, it is easy to
vertical angle between the horizon and a celestial body such determine your latitude by observing the height of Polaris
as the sun. above your northern horizon.
 Properties of the graticule 13

Courtesy of Dr. Yuri Beletsky.

Figure 1.10 The Southern Cross is used for
navigation in the southern hemisphere.

Properties of the graticule This sphere is a little over three-tenths of a percent larger
in surface area than the WGS 84 ellipsoid. These circum-
Circumference of the authalic and other ference values are all very close to each other, differing by
spheres less than two-tenths of a percent.
When determining latitude and longitude we some-
times use a different approximation to the earth than Spacing of parallels
the oblate ellipsoid. Using a sphere leads to simpler cal- As we saw earlier, on a spherical earth the north–south
culations, especially when working with small-scale ground distance between equal increments of latitude
maps of countries, continents, or the entire earth (see does not vary. However, it is important to know how
the next chapter for more on small-scale maps). On you want to define the distance. As we will see below,
these maps differences between locations on the sphere there are different definitions for terms you may take for
and the ellipsoid are negligible. The value of the earth’s granted, such as a “mile.”
spherical circumference used in this book is for what Using the WGS 84 authalic sphere circumference,
cartographer’s call an authalic sphere. The authalic latitude spacing is always 24,874 miles ÷ 360° or 69.09
(meaning “area-preserving”) sphere is a sphere with the statute miles per degree. Expressed in metric and nauti-
same surface area as a reference ellipsoid we are using. cal units, it is 111.20 kilometers and 60.04 nautical miles
The equatorial and polar radii of the WGS 84 ellipsoid per degree of latitude (see table C.1 in appendix C for the
are what we used to calculate the radius and circumfer- metric and English distance equivalents used to arrive at
ence of the authalic sphere that is equal to the surface these values.)
area of the WGS 84 ellipsoid. The computations involved You can see that there is about a 15 percent difference
are moderately complex and best left to a short com- in the number of statute and nautical miles per degree.
puter program, but the result is a sphere with a radius of Statute miles are what we use for land distances in the
3,958.76 miles (6,371.017 kilometers) and circumference United States, while nautical miles are used around
24,873.62 miles (40,030.22 kilometers). the world for maritime and aviation purposes. A statute
There are other properties of an oblate ellipsoid that mile is about 1,609 meters, while the international stan-
we can use to define the circumference of a spherical dard for a nautical mile is 1,852 meters exactly (about
earth. A rectifying sphere, for example, is one where the 1.15 statute miles). The original nautical mile was defined
length of meridians from equator to pole on the ellipsoid as 1 minute of latitude measured north–south along a
equals one-quarter of the spherical circumference. For meridian. Kilometers are also closely tied to distances
the WGS 84 ellipsoid, the rectifying sphere is of radius along meridians, since the meter was initially defined as
3,956.55 miles (6,367.449 kilometers) and circumfer- one ten-millionth of the distance along a meridian from
ence 24,859.73 miles (40,007.86 kilometers). We can the equator to the north or south pole.
also use the 3,964.038 mile (6378.137 kilometer) equa- We have seen that parallels on an oblate ellipsoid are
torial radius of the WGS 84 ellipsoid, giving a sphere of not spaced equally, as they are on a sphere, but decrease
circumference 24,901.46 miles (40,075.017 kilometers). slightly from the pole to the equator. You can see in
14 Chapter 1 The earth and earth coordinates

table C.2 (in appendix C) the variation in the length of 69.6

Miles per degree of latitude

a degree of latitude for the WGS 84 ellipsoid, measured
69.4 Authalic
along a meridian at one-degree increments from the equa- sphere
tor to the pole. You can see that the distance per degree of 69.2
latitude ranges from 69.407 statute miles (111.699 kilo-
69.0
meters) at the pole to 68.703 statute miles (110.567 kilo- Rectifying WGS 84
meters) at the equator. The graph in figure 1.11 shows how 68.8 sphere ellipsoid
these distances differ from the constant values of 69.09
68.6
and 69.07 statute miles (111.20 and 111.16 kilometers) 0 10 20 30 40 50 60 70 80 90
per degree for the authalic and rectifying spheres, respec- Geodetic latitude (deg.)
tively. For both spheres, the WGS 84 ellipsoid distances
per degree are about 0.3 miles (0.48 kilometers) greater
Figure 1.11 Distances along the meridian for one-degree
than the sphere at the pole and 0.4 miles (0.64 kilo-
increments of latitude from the equator to the pole on the
meters) less at the equator. The ellipsoidal and spheri-
WGS 84 ellipsoid, authalic sphere, and rectifying sphere.
cal distances are almost the same in the mid-latitudes,
somewhere between 45 and 50 degrees.

Converging meridians great circles. A great circle is the shortest route between
A quick glance at any world globe (or figure 1.2) shows any two points on the earth and hence great circle routes
that the length of a degree of longitude, measured east– are fundamental to long-distance navigation.
west along parallels, decreases from the equator to the Any circle on the earth’s surface that intersects the
pole. The precise spacing of meridians at a given latitude interior of the sphere at any location other than the
is found by using the following equation: center is called a small circle, and its circumference is
smaller than a great circle. You can see in figure 1.3 that
69.09 miles (or 111.20 kilometers) / deg. × cosine
all parallels other than the equator are small circles.
(latitude).*
The circumference of a particular parallel is given by the
At 45° north or south of the equator, for example, cosine following equation:
(45°) = 0.7071. Therefore, the length of a degree of lon-
24,874 miles (or 40,030 kilometers) × cosine (latitude)
gitude is
For example, the circumference of the 45th parallel is
69.09 × 0.7071, or 48.85 statute miles
(111.20 × 0.7071 or 78.63 kilometers) 24,874 × 0.7071 or 17,588 statute miles
(40,030 × 0.7071 or 28,305 kilometers)
This is roughly 20 miles (32 kilometers) shorter than the
69.09-mile (111.20-kilometer) spacing at the equator.
Quadrilaterals
Great and small circles Quadrilaterals are areas bounded by equal increments
A great circle is the largest possible circle that could be of latitude and longitude, 10° by 10° in figure 1.3, for
drawn on the surface of the spherical earth. Its circum- example. Since meridians converge toward the poles, the
ference is that of the sphere, and its center is the center shapes of quadrilaterals vary from a square on the sphere
of the earth so that all great circles divide the earth into at the equator to a very narrow spherical triangle at the
halves. Notice in figure 1.3 that the equator is a great pole. The equation cosine(latitude) gives the aspect ratio
circle dividing the earth into northern and southern (width/height) of any quadrilateral. A quadrilateral cen-
hemispheres. Similarly, the prime meridian and the 180° tered at 45°N will have an aspect ratio of 0.7071, whereas
meridian at the opposite side of the earth (called the a quadrilateral at 60°N will have an aspect ratio of 0.5. A
antipodal meridian) form a great circle dividing the map covering 10° by 10° will look very long and narrow
earth into eastern and western hemispheres. All other at this latitude.
pairs of meridians and their antipodal meridians are also

* Inexpensive engineering calculators can be used to compute trigonometric functions such as cosine. You can also find trig
function calculators on the Internet.
 Geodetic latitude and longitude on large-scale maps 15

Graticule appearance on Large-scale maps

maps Parallels and meridians are shown in different ways on
different types of maps. Topographic maps in the United
Small-scale maps States and other countries have tick marks showing the
World or continental maps such as globes and world atlas location of the graticule. All U.S. Geological Survey 7.5-
sheets normally use coordinates based on an authalic minute topographic maps, for example, have graticule
sphere. There are several reasons for this. Prior to using ticks at 2.5-minute intervals of latitude and longitude
digital computers to make these types of maps numeri- (figure 1.12). The full latitude and longitude is printed
cally, it was much easier to construct them from spheri- in each corner, but only the minutes and seconds of the
cal coordinates. Equally important, the differences in intermediate edge ticks are shown. Note the four “+”
the plotted positions of spherical and corresponding geo- symbols used for the interior 2.5-minute graticule ticks.
detic parallels become negligible on maps that cover so The graticule is shown in a different way on nautical
much area. charts (figure 1.13). Alternating white and dark bars
You can see this by looking again at table 1.3. Earlier spaced at the same increment of latitude and longitude
we saw that the maximum difference between spherical line the edge of the chart. Because of the convergence of
and geodetic latitude is 0.128° at the 45th parallel. Imag- meridians, the vertical bars on the left and right edges of
ine drawing parallels at 45° and 45.128° on a map scaled the chart showing equal increments of latitude are longer
at one inch per degree of latitude. The two parallels will than the horizontal bars at the top and bottom. Notice
be drawn a very noticeable 0.128 inches (0.325 centi- the more closely spaced ticks beside each bar, placed every
meters) apart. Now imagine drawing the parallels on a tenth of a minute on the chart in figure 1.13. These ticks
map scaled at one inch per 10 degrees of latitude. The are used to more precisely find the latitude and longitude
two parallels will now be drawn 0.013 inches (0.033 of mapped features.
centimeters) apart, a difference that would not even be Aeronautical charts display the graticule another way.
noticeable given the width of a line on a piece of paper. The chart segment for a portion of the Aleutian Islands
This scale corresponds to a world wall map approxi- in Alaska (figure 1.14) shows that parallels and meridians
mately 18 inches (46 centimeters) high and 36 inches are drawn at 30-minute latitude and longitude intervals.
(92 centimeters) wide. Ticks are placed at 1-minute increments along each grati-
cule line, allowing features to be located easily to within
a fraction of a minute.

123°22’30” 20’ 17’30” 123°15’

44°37’30” 44°37’30”
Geodetic latitude and
longitude on large-scale
maps
35’ 35’
You will always find parallels and meridians of geodetic
latitude and longitude on detailed maps of smaller extents.
This is done to make the map a very close approximation
to the size and shape of the piece of the ellipsoidal earth
32’30” 32’30” that it represents. To see the perils of not doing this, you
only need to examine one-degree quadrilaterals at the
equator and pole, one ranging from 0° to 1° and the
second from 89° to 90° in latitude.
You can see in table C.2 (in appendix C) that the
44°30’ 44°30’ ground distance between these pairs of parallels on the
123°22’30” 20’ 17’30” 123°15’ WGS 84 ellipsoid is 68.703 and 69.407 miles (110.567
and 111.699 kilometers), respectively. If the equatorial
Figure 1.12 Graticule ticks on the Corvallis, Oregon, 7.5- quadrilateral is mapped at a scale such that it is 100 inches
minute topographic map. (254 centimeters) high, the polar quadrilateral mapped
16 Chapter 1 The earth and earth coordinates

Reproduced with the permission of the Canadian Hydrographic Service.

Courtesy of the National Aeronautical Charting Office.

Figure 1.13 Graticule bars on the edges of a nautical chart
segment. Figure 1.14 Graticule ticks on a small segment of an
aeronautical chart.

at the same scale will be 101 inches (256.5 centimeters) by surveyors when they determined the exact location
long. If we mapped both quadrilaterals using the authalic of a point) like figure 1.15 on the ground on top of a
sphere having 69.09 miles (111.20 kilometers) per degree, hill or other prominent feature. From the 1920s to the
both maps would be 100.6 inches (255.5 centimeters) early 1980s these control points were surveyed relative
long. Having both maps several tenths of an inch (or to the surface of the Clarke 1866 ellipsoid, together
around a centimeter) longer or shorter than they should forming what was called the North American Datum
really be is an unacceptably large error for maps used of 1927 (NAD 27). Topographic maps, nautical and
to make accurate measurements of distance, direction, aeronautical charts, and many other large-scale maps
or area. of this time period had graticule lines or ticks based on
this datum. For example, the southeast corner of the
Horizontal reference datums Corvallis, Oregon, topographic map first published in
To further understand the use of different types of 1969 (figure 1.16) has an NAD 27 latitude and longi-
coordinates on detailed maps of smaller extents, we tude of 44°30΄N, 123°15΄W.
must first look at datums—the collection of very By the early 1980s, better knowledge of the earth’s
accurate control points (points of known accuracy) shape and size and far better surveying methods led to
surveyors use to georeference all other map data (see the creation of a new horizontal reference datum, the
chapter 5 for more on control points and georeferenc- North American Datum of 1983 (NAD 83). The
ing). Surveyors determine the precise geodetic latitude NAD 27 control points were corrected for survey-
and longitude of horizontal control points spread ing errors where required, then these were added to
across the landscape. You may have seen a horizontal thousands of more recently acquired points. The geo-
control point monument (a fixed object established detic latitudes and longitudes of all these points were
 Geodetic latitude and longitude on large-scale maps 17

Europe in the early 1900s faced another problem—

separate datums for different countries that did not mesh
into a single system for the continent. Military map users
in World War II found different latitudes and longitudes

Courtesy of the National Oceanic and Atmospheric Administration.

for the same ground locations on topographic maps along
the borders of France, Belgium, the Netherlands, Spain,
and other countries where major battles were fought. The
European Datum of 1950 (ED 50) was created after
World War II as a consistent reference datum for most
of western Europe, although Belgium, France, Great
Britain, Ireland, Sweden, Switzerland, and the
Netherlands continue to retain and use their own national
datums. Latitudes and longitudes for ED 50 were based
on the International Ellipsoid of 1924. Users of GPS
receivers will find that, moving westward through
Europe from northwestern Russia, the newer Euro-
pean Terrestrial Reference System 1989 (ETRS 89)
longitude coordinates based on the WGS 84 ellipsoid
Figure 1.15 Horizontal control point marker cemented in
gradually shift to the west of those based on the Interna-
the ground.
tional Ellipsoid of 1924. In Portugal and western Spain
the WGS 84 longitudes are approximately 100 meters
to the west of those found on topographic maps based
determined relative to the Geodetic Reference System on ED 50. Moving southward, WGS 84 latitudes gradu-
of 1980 (GRS 80) ellipsoid, which is essentially identi- ally shift northward from those based on ED 50, reach-
cal to the WGS 84 ellipsoid. ing a maximum difference of around 100 meters in the
The change of horizontal reference datum meant that Mediterranean Sea.
the geodetic coordinates for control points across the Great Britain and Ireland are examples of countries
continent changed slightly in 1983, and this change had that continue to use ellipsoids defined in the nineteenth
to be shown on large-scale maps published earlier but still century to best fit their region. Topographic maps in
in use. On topographic maps the new position of the map both nations use the Airy 1830 ellipsoid as the basis for
corner is shown by a dashed “plus” sign, as in figure 1.16. the Ordnance Survey Great Britain 1936 (OSGB 36)
Many times the shift is in the 100 meter range and must datum for geodetic latitude and longitude coordinates.
be taken into account when plotting on older maps the Along the south coast of England, WGS 84 latitudes
geodetic latitudes and longitudes obtained from GPS are about 70 meters to the south of those based on
receivers and other modern position finding devices. OSGB 36. This southward shift gradually diminishes
to zero near the Scottish border and then becomes a
northerly difference that reaches a maximum value of
Courtesy of the U.S. Geological Survey.

around 50 meters at the northern extremes of Scotland.

In Ireland, WGS 84 longitudes are around 50 meters
to the east of their OSGB 36 equivalents and gradually
increase to a maximum difference of around 120 meters
along the southeast coast of England.

Figure 1.16 Southeast corner of the Corvallis, Oregon,

topographic map showing the difference between its
NAD 27 and NAD 83 position.
18 Chapter 1 The earth and earth coordinates

The earth as a geoid

When we treat the earth as a smooth authalic sphere
Ellipsoid

Elevation (H)
or oblate ellipsoid, we neglect mountain ranges, ocean height (h)
trenches, and other surface features that have vertical Ellipsoid
WGS 84
relief. There is justification for doing this, as the earth’s Geoid
surface is truly smooth when we compare the sur- height (N)
Local mean
sea level
face undulations to the 7,918-mile (12,742-kilometer) Local geoid
diameter of the earth based on the authalic sphere. The
greatest relief variation is the approximately 12.3-mile Global geoid
Ocean Same gravity strength
(19.8-kilometer) difference between the summit of Global mean as at mean sea level
Mt. Everest (29,035 feet or 8,852 meters) and the sea level

deepest point in the Mariana Trench (36,192 feet or

Land
11,034 meters). This vertical difference is immense on
our human scale, but it is only 1/640 th of the earth’s Figure 1.17 The geoid is the surface where gravity is the
diameter. If we look at the difference between the earth’s same as at mean sea level. Elevations traditionally have been
average land height (2,755 feet or 840 meters) and ocean measured relative to the geoid, but modern GPS-determined
depth (12,450 feet or 3,795 meters), the average rough- heights are relative to the WGS 84 ellipsoid.
ness is only 4,635m/12,742km, or 1/2,750 th of the diam-
eter. It has been said that if the earth were reduced to the
diameter of a bowling ball, it would be smoother than Later, surveyors could determine elevation by making
the bowling ball! gravity measurements at different locations on the land-
The earth’s global-scale smoothness aside, knowing the form and relating them to the strength of gravity at the
elevations and depths of features is very important to us. point used to define MSL. Gravity differences translate
Defining locations by their geodetic latitude, longitude, into elevation differences.
and elevation gives you a simple way to collect elevation Mean sea level is easy to determine along coastlines, but
data and display this information on maps. The top of what about inland locations? What is needed is to extend
Mt. Everest, for example, is located at 27°59΄N, 86°56΄E, mean sea level across the land. Imagine that the mean sea
29,035 feet (8,852 meters), but what is this elevation rela- level is extended under the continental land masses, which
tive to? This leads us to another approximation of the is the same thing as extending a surface having the same
earth called the geoid, which is a surface of equal gravity strength of gravity as mean sea level (figure 1.17). This
used as the reference for elevations. imaginary equal gravity surface doesn’t form a perfect
ellipsoid, however, because differences in topography and
Vertical reference datums earth density affect gravity’s pull at different locations.
Elevations and depths are measured relative to what is The slightly undulating nearly ellipsoidal surface that
called a vertical reference datum, an arbitrary surface best fits mean sea level for all the earth’s oceans is called
with an elevation of zero. The traditional datum used for a global geoid. The global geoid rises and falls approxi-
land elevations is mean sea level (MSL) (see chapter 6 mately 100 meters above and below the oblate ellipsoid
for more on mean sea level). Surveyors define MSL as the surface in an irregular fashion. World maps showing land
average of all low and high tides at a particular starting topography and ocean bathymetry use land heights and
location over a metonic cycle (the 19-year cycle of the water depths relative to the global geoid surface.
lunar phases and days of the year). Early surveyors chose The mean sea level datum based on the geoid is so
this datum because of the measurement technology of the convenient that it is used to determine elevations around
day. Surveyors first used the method of leveling, where the world and is the base for the elevation data found on
elevations are determined relative to the point where nearly all topographic maps and nautical charts. But be
mean sea level is defined, using horizontally aligned tele- aware that the local geoid used in your area is probably
scopes and vertically aligned leveling rods (see chapter 6 slightly above or below (usually within two meters) the
for more on leveling). A small circular monument was global geoid elevations used on world maps. This differ-
placed in the ground at each surveyed benchmark eleva- ence is caused by mean sea level at one or more locations
tion point. A benchmark is a permanent monument that being used as the vertical reference datum for your nation
establishes the exact elevation of a place. or continent, not the average sea level for all the oceans.
 The earth as a geoid 19

In the United States, for example, you may see since the retreat of glaciers at the end of the last ice age,
elevations relative to the National Geodetic Vertical or subsidence from sedimentation and the extraction of
Datum of 1929 (NGVD 29) on older topographic natural resources like oil and water.
maps. This datum was defined by the observed heights GPS has created a second option for measuring eleva-
of mean sea level at 26 tide gauges, 21 in the United tion (see chapter 14 for more on GPS). GPS receivers
States and 5 in Canada. It also was defined by the set calculate what is called the ellipsoidal height (h), the
of elevations of all benchmarks resulting from over distance above or below the surface of the WGS 84 ellip-
60,000 miles (96,560 kilometers) of leveling across soid along a line from the surface to the center of the
the continent, totaling over 500,000 vertical control earth (figure 1.17). An ellipsoidal height is not an eleva-
points. In the late 1980s, surveyors adjusted the 1929 tion, since it is not measured relative to the mean sea level
datum with new data to create the North American datum for your local geoid. Therefore, you must convert
Vertical Datum of 1988 (NAVD 88). Topographic GPS ellipsoidal height values to mean sea level datum ele-
maps, nautical charts, and other cartographic products vations (H) before you can use them with existing maps.
made from this time forward have used elevations You do this by subtracting the geoid height (N) at each
based on NAVD 88. Mean sea level for the continent point from the ellipsoid height (h) measured by the GPS
was defined at one tidal station on the St. Lawrence receiver using the equation H = h – N. The look-up table
River at Rimouski, Quebec, Canada. NAVD 88 was needed to make this conversion usually is stored in your
a necessary update of the 1929 vertical datum since GPS receiver’s computer. In the conterminous United
about 400,000 miles (650,000 kilometers) of leveling States, geoid heights range from a low of –51.6 meters in
had been added to the NGVD since 1929. Additionally, the Atlantic Ocean to a high of –7.2 meters in the Rocky
numerous benchmarks had been lost over the decades— Mountains (figure 1.18). Worldwide, geoid heights vary
and the elevations at others had been affected—by from –105 meters just south of Sri Lanka to 85 meters
vertical changes caused by rising of land elevations in Indonesia.

Courtesy of the National Geodetic Survey.

Figure 1.18 Geoid heights in the United States and vicinity (from the National Geodetic Survey GEOID2003 model).
20 Chapter 1 The earth and earth coordinates

Selected readings
Greenhood, D. 1964. Mapping. Chicago: University of
Chicago Press.
Iliffe, J. C. 2000. Datums and map projections for remote
sensing, GIS and surveying. Caithness, Scotland: Whittles
Publishing.
La Condamine, C. M. de. 1747. A succinct abridgement of
a voyage made within the inland parts of South America
as it was read to the Academy of Science, Paris, April 28,
1745. London.
Maling, D. H. 1992. Coordinate systems and map projections.
2d ed. New York: Pergamon Press.
Maupertius, P. L. M. de. 1738. The figure of the earth deter-
mined from observations made by order of the French king,
at the polar circle (translation). London.
Meade, B. K. 1983. Latitude, longitude, and ellipsoidal height
changes NAD-27 to oredicted NAD-83. Surveying and
Mapping 43: 65–71.
Robinson, A. H., et al. 1995. Basic geodesy. In Elements of
cartography. 6th ed. New York: John Wiley & Sons, Inc.
Smith, J. R. 1988. Basic geodesy. An introduction to the his-
tory and concepts of modern geodesy without mathematics.
Rancho Cordova, Calif.: Landmark Enterprises.
Snyder, J. P. 1987. Map projections–a working manual. U.S.
Geological Survey Professional Paper 1395. Washington,
D.C.: U.S. Government Printing Office.
Sobel, D. 1995. Longitude: The true story of a lone genius who
solved the greatest scientific problem of his time. New York:
Walker & Co.
Wallis, H. M., and A. H. Robinson, eds. 1987. Cartographical
innovations. London: Map Collector Publications.
Wilford, J. N. 1981. The mapmakers. New York: Alfred A.
Knopf.
U.S. Department of the Army. 2001. Grids. In Map reading.
FM 3–25.26 Washington, D.C.: Department of the Army.