Robinson Projections
|
| Figure 1. The Robinson projection.
|
The Robinson projection was developed by Arthur H. Robinson in 1963.
It was the first major map projection to be commissioned by a large
private corporation; Rand McNally hired Robinson (who was a
professor in the Geography Department at the University of
Wisconsin in Madison from 1945 until he retired in 1980)
to develop the projection because they were not satisfied with
the ability of existing projections to create intuitively
appealing depictions of the entire world. Rand McNally still makes
extensive use of the Robinson projection, and the National Geographic
Society uses it as well (although the Society seems to be using
the projection a bit less frequently now than it did in the 1980s).
Robinson called this the orthophanic projection (which means
"right appearing"), but this name never caught on. In at least one
reference book, this projection is termed the Pseudocylindrical
Projection with Pole Line, which is highly descriptive (the "pole
line" bit comes from the fact that the North and South Poles on a
Robinson projection are shown as lines and not points -- see Figure
1), but so unwieldily that it is not surprising that this name also
failed to gain much acceptance. Robinson projection is
unquestionably the name of choice.
The Robinson projection is highly unique. Unlike all other projections,
Dr. Robinson did not develop this projection by developing new geometric
formulas to convert
latitude
and
longitude
coordinates from the surface of the
Model of the Earth
to locations on the map. Instead, Dr. Robinson used a huge number of
trial-and-error computer simulations to develop a table that allows a
cartographer
to look up how far above or below a Robinson map's
equator
a particular
line of latitude
will be located, and then to estimate (via a simple interpolation
process) where along this line a particular
longitude
will fall. To this day, no other projection uses this approach to build
a map.
- Form:
The Robinson projection can best be described as being
pseudocylindrical,
but given its unique method of development, it does not fall perfectly
into any known
form
category.
- Case:
The Robinson projection is basically
secant,
with
lines of tangency
running along the 38° 0' 0"N and 38° 0' 0"S
lines of latitude,
but once again, given its unique method of development, you could
certainly make a compelling argument that the Robinson projection has
no meaningful
case.
- Aspect:
Robinson projections have
normal aspects.
- Variation Within Robinson Projections:
Robinson projections differ from one another in the locations of their
central meridians.
- Distortions
- Shearing:
The Robinson projection is not
conformal;
shapes are distorted more than they would be in a truly
conformal
projection. However, shapes are not distorted very badly within
about 45° north or south of the
equator
or within about 45° of the map's
central meridian.
Beyond these limits, shape distortion can be quite severe.
- Tearing:
Robinson maps show
lines of latitude
as parallel straight lines and
lines of longitude
as nonparallel lines that become increasingly curved as you move farther
away from the map's
central meridian.
The poles are shown as straight lines (0.5322 times as long as the
equator)
and the east and west edges of the map are curved lines 180° from
the map's
central meridian.
These east and west edges are markedly less curved than are the
edges of other
pseudocylindrical
projections (the result being that the Robinson projection generally
suffers from less
shearing
than do other
pseudocylindrical
projections). Tearing occurs along the edges of a Robinson map.
The Robinson projection is not well suited to building
interrupted maps.
- Compression:
Robinson projections are not
equivalent;
they do suffer from
compression.
However, the amount of area distortion is generally low within about
45° of the
equator.
- Equivalence:
Robinson projections are not
equivalent;
they do suffer from
compression.
However, the amount of area distortion is generally low within about
45° of the
equator.
- Conformality:
The Robinson projection is not
conformal;
shapes are distorted more than they would be in a truly
conformal
projection. However, shapes are not distorted very badly within
about 45° north or south of the
equator
or within about 45° of the map's
central meridian.
Beyond these limits, shape distortion can be quite severe.
- Equidistance:
The Robinson projection is not
equidistant;
there is no point or points from which all distances are shown
accurately.
- Azimuthality:
The Robinson projection is not
azimuthal;
there is no point or points from which all directions are shown
accurately.
- Uses: The Robinson projection is very unique. Its only
real purpose is to create visually appealing maps of the entire world; it
is rarely used for any other purpose. It is a compromise projection; it does
not eliminate any type of distortion, but it keeps the levels of all types of
distortion relatively low over most of the map. It is still used in atlases
and other reference works; but being as specialized as it is, it is doubtful
if the Robinson projection will ever find extremely widespread use.