Editorial

Abstract

Introduction

The Issues

Dynamic mapping tools

cdv

Alternatives to choropleth maps

Conclusions

Future directions

Acknowledgements

References

Case Studies Index

3. Dynamic mapping tools and cartographic data visualization

This section will stress highly interactive visualisation as a way to use the issues noted in (2) above to the map user's advantage. Current technology permits rapid map production, and so single, or more often multiple, transient maps can be developed as requested to highlight very specific user-led queries. Graphics are now available instantly at the press of a button, and they embrace new dynamic properties that can add enormously to their functionality and fitness for use. In consequence, a trend has developed away from a single optimum means of representation towards multiple alternatives, where symbols appear in different ways at different times to stress particular messages. This has had an impact on cartography. Rather than being restrictive shackles that must be searched through for an aesthetic, multipurpose, or optimal solution, map design parameters can be altered in real-time to assess confidence in patterns through the extent to which they are influenced by changes in the method of representation. Muehrcke (1990) referred to this property as map stability. Map users can focus their attention on different aspects of a spatial data set at different times with maps that vary as required. Local variations and minor fluctuations can be afforded the visual power of the full range of symbolism available, if and when requested.

Maps can thus be used an integral part of the research process, as dynamic visual interfaces to a collection of spatial information which respond when they are interrogated for information. This shift in map use was recognized by DiBiase (1990) who identified private map use as distinct from that occurring in the public realm and termed the cyclical map-based process of visual exploration and synthesis of ideas visual thinking. It corresponds with a use of graphical tools early in the research process across the sciences termed visualization and acknowledged by McCormick, DeFanti & Brown (1987). The use of dynamic maps for visual thinking is called cartographic visualization.

Early examples of transient private maps which embraced dynamic techniques were provided by Ferreira and Wiggins (1990) whose density dial allowed users to slice through a variable's range at a level determined by a dial and view the resultant classification on a map. Such a dial enables users to determine the sensitivity of the classification, and by varying the scheme they can determine the level at which a variety of patterns occur.

Such transient symbolism can be used to simply highlight symbols in a nominal manner when they are selected for investigation. Stuetzle (1988) was an early pioneer of what is called brushing, where corresponding symbols in two or more views are identified by identical distinct symbolism. Monmonier (1989) extended the technique to link choropleths with statistical plots, adding a geographic component to scatter plots, and vice versa. Just as transient maps eliminate the constraints associated with producing single representations, linking effectively removes the restriction of having to use the spatial dimensions of the page to show geographical locations. Maps can use the spatial arrangement of symbols on a page to reflect geographic relationships whilst at the same time using locations on linked views to show statistical variation.

Dynamic maps enable the focus of interest to be varied spatially by changing scale, or location, with zoom and pan functionality, and statistically by selecting subsets from statistical views of data. For example, the problem of simultaneous contrast can be overcome by selecting a data sub-set to extend the variation in symbolism and by highlighting symbols on a statistical dot plot which uses location to display the same values.

In order to publicise and promote these techniques and changes in map use a JISC New Technologies Initiative Project, ARGUS, produced software, documentation and data sets for cartographic visualization as demonstrators and in research and teaching. The resources developed constitute a Visualization Toolkit for Teaching and Learning in the Spatial Sciences (see http://midas.ac.uk/argus) which consists of:

• A series of tutorials on cartographic techniques
• Example data for a specified region in the public domain
• Software for visualizing two-dimensional data sets
• Software for visualizing three-dimensional data sets
• Research papers on methods and techniques

Two-dimensional census data sets were visualized with a Cartographic Data Visualizer, cdv, specifically developed by the project to demonstrate a variety of new methods and opportunities for interacting with enumerated data, such as those outlined above. The following section introduces some of these techniques.