Recent advances in technology allow cartographic images to be transformed into dynamic interactive computerized viewing environments. With maps less restricted to the format of a static image, human interaction may play an expanded role in the exploration of data contained in such dynamic maps. This study tests the ability of map users to make accurate comparisons from the information presented in a series of interactive and dynamic three-dimensional (3D) maps.
The use of web based virtual reality for mapping is a development of the past few years. Easy access to such maps is also recent. Yet, little is known about the ability of the average map user to correctly visualize the space represented on the screen, nor about the viewer’s use of the interactive navigation tools.
The concept here is to consider a basic population map (see Figure 1.1) and apply to it some of the features available in the 3D interactive viewing environment of the Virtual Reality Modeling Language (VRML). The research examines demographic mapping which uses the three-dimensional and dynamic features available on the World Wide Web with VRML. It tests visual perception of objects within such a mapping environment with the goal being to determine appropriate and useful applications of those features.
Figure
1.1. Basic population map of
Hawai`i (Juvik and Juvik 1998)
The initial idea is to transform flat population circles, which have the size characteristic of radius, into three-dimensions. If the circle is transformed into a sphere, it would still be limited to one size characteristic. By creating a cylinder, the shape has an additional size characteristic – height, and it also maintains the same shape over the same area at all heights. The cylinders are placed on end so that, if viewing the map from above, the cylinders located directly below appear as circles. To take advantage of this 3D mapping environment, however, the maps are viewed by looking across the landscape, rather than planimetrically. This allows the height of the cylinder as well as the diameter to represent information about the population. The VRML environment also allows for animation to be applied to these cylinders. In the concept proposed here, rotation, which includes the properties of axis of direction, direction of rotation, and rate of rotation, is applied to the cylinders.
As will be discussed in Chapter 2, there is a need to limit the number of parameters applied to symbols on a map, in this case the cylinders. The concept proposed here is to use the following variables: location, color, height, and rotation. In this case, location is used to represent where the data belongs. Color is used to divide the cylinders into categorical groups. Height, rather than the conventional diameter, is used to show the population size due to the different basic viewpoint of the mapping environment. Rotation is used to show change in population. If the rotation is clockwise, the population is increasing, and if counter-clockwise, the population is decreasing. A faster rate of rotation indicates a higher rate of change. Height and rotation represent the central aspect of this study, although the actual test focuses on height estimation.
Participants were asked to view a series of three-dimensional population maps in different interactive levels and then make estimates of the populations represented by the cylinders. In considering the three-dimensional and dynamic aspects of this mapping environment, analysis in this research focuses on the accuracy of user responses. These responses are examined in regard to three major capabilities that are provided in the VRML environment. These include the capability to have a base map either be flat or represent topographic terrain, the capability to rotate some symbols while leaving other symbols stationary, and the capability to offer different kinds of interaction between the user and the computerized map. These user interaction types include no interaction, pre-programmed interaction, and participant controlled interaction. Length of time spent within each interaction level are examined as well as participant use of the navigation tools. The objective of this research is to make a determination on which of these factors, or combination thereof, are most effective for the proposed mapping environment.