Summary graphics for integrated visualization in dynamic cartography

Dynamic cartography calls for a single, more cognitively friendly graphic that summarizes salient relationships in an animated sequence of maps. The theory of human information processing suggests that because the human eye-brain system does not instantaneously process patterns from short-term memory through to long-term memory, information presented toward the end of a dynamic cartographic sequence retards the memory and comprehension of information presented earlier in the sequence. Four types of summary graphics hold particular promise as animation supplements: (1) the centrographic time-series map/used since the late 19th century to portray the westward march of the center of the U.S. population; (2) the biplot, a joint, twodimensional representation of time units and places based upon two principal components; (3) canonical trend-surface analysis, which might extract one or two salient spatial trends, the canonical loadings of which can be plotted in a time-series graph showing when each trend was particularly prominent; and (4) the time-series correlation graph, which reveals temporal variation in the apparent influence of given trends or regionalizations on a particular bivariate correlation. Because an animated sequence of maps can promote understanding of an otherwise complicated summary graphic, the cartographic animation and its summary graphic can be complementary.