Lab 5: GIS Projections

Figure 1:


Figure 2:


Figure 3:


Lab 5 demonstrated how different map projections may cause variations in a map's characteristics. Although projections cause some discrepancies, they are the most efficient way of representing the real three dimensional world on a two dimensional surface. Even the most accurate representation (a globe) is imperfect because it models the Earth as a perfect sphere when it is a crude ellipsoid. Nevertheless, projections are invaluable because they help people objectify their spatial surroundings.

Different map projections have different strengths and thus are used for a variety of purposes. Conformal map projections, for example, preserve angles and relative shapes. Likewise, equal area projections preserve relative areas on a map and equidistant projections preserve distances.Since each projection type preserves only a few aspects (i.e. distance, angles, and area), the type of projection should be chosen only after the purpose of the map is defined. Further, the view may also change between map projections. Notice the view difference between the "Equidistant Conic" and the "Plate Carree" images (Figure 2). Although they are both equidistant projections, the perspective they take are obviously different. The equidistant conic takes more of a bird's eye approach from the pole whereas the plate carree excludes the poles.

Although three map projection types exist (conformal, equidistant, and equal area), two maps of the same type are often appear very different. For example, the "Mercator" and "Gall Stereographic" projections are both conformal maps which means they preserve local angles (Figure 3). However, they are mainly different in that the Mercator emphasizes the upper and lower thirty degrees, evidenced by the disproportionate sizes of Greenland and Antarctica. In contrast, the Gall Stereographic projection gives each 30 degrees of longitude equal space on the map. Even in the Gall Stereographic projection, however, Greenland and Antarctica are still disproportionate when compared to an equal area projection.

The disparity in distances between Washington D.C. and Kabul, Afghanistan were interesting to note. The measured distances between the two cities ranged from 6,973 miles to 10,082 miles depending on the map projection used. At first glance, it was surprising that the Plate Carree and Equidistant Conic projections (two equidistant projections) varied so widely in distance: 10,079 miles to 6,972 miles. On further inspection, I realized that simply drawing a straight line between the two points created two entirely different routes that could not be compared due to their difference in view. For example, the line connecting Washington D.C. and Kabul in the Equidistant Conic projection passed through the Arctic Circle whereas as in the Plate Carree, the line crossed the Atlantic Ocean. An important lesson from this lab is that maps should not be taken at 'face value'. A map's specifications (i.e. the type of projection), similar to any set of data, should be considered when being viewed.