Mapmaking - The Science of Navigation - page 98

bad as those that result from the much more familiar Mercator projection.
The map of figure 4.4 also distorts distances except along the north-south
direction and along the equator. It also distorts directions except along the
cardinal compass points (north, south, east, and west). Because of these
distortions, it is rarely used. Its main advantage accrues to the mapmaker,
not the map user: it is easy to draw.
Three Common Types of Projections—and an Oddball
There are dozens, hundreds, of named map projections. Why so many?
Because each has an advantage over its fellows, in one way or another, as
we will see. Maps are used for very di√erent purposes by di√erent people.
A navigator in an Age-of-Sail ship, plowing the open oceans, required
nautical charts that served a very di√erent purpose from the maps used
today to monitor the spread of pandemic diseases. A di√erent map would
be better at showing the world's ocean currents; yet another map will be
better suited to displaying airline routes; and so on. 9
CYLINDERS, CONES, AND PLANES
The majority of maps fall into one of three basic families: azimuthal projec-
tions , conical projections , and cylindrical projections . These three categories
are illustrated in figure 4.5. The trick is to wrap a piece of paper around a
globe. We cannot stretch the paper, so we cannot wrap the globe com-
pletely. By wrapping the paper around the equator, as in figure 4.5c, we
have formed it into a cylinder. From this configuration, if we can map the
globe onto the paper by performing a projection , then every point on the
globe is systematically placed upon the paper, and when we straighten out
the paper, we have a map. So how do we perform a projection from spheri-
cal globe to cylindrical paper? There are many di√erent ways, as we will
soon see.
Conical projections adopt the configuration of globe and paper as shown
in figure 4.5b. Paper can be twisted into a cone without stretching it,
so this configuration works. We project somehow from globe to cone,
unfold, and we have a conical-projection map. For the azimuthal projec-
tions, we don't need to twist the paper at all: it remains a flat plane,
9. Snyder (1993) describes almost 200 projections in his technical history of the subject.
Readers who crave the detailed math can find it in, e.g., Dent (1998) and Yang, Snyder, and
Tobler (2000). A much lighter, humorous introduction is provided by Gonzalez and Sherer
(2004).