A Classification of Cartographic Animations: Towards a Tool for the Design of Dynamic Maps in a GIS Environment

Menno-Jan Kraak & Arjen Klomp

Back to Table of Contents

Introduction

The need in the GIS environment to deal with processes as a whole, and no longer with single time-slices also has its influences on the visualization aspects. To visualize models or planning operations is no longer efficient with static paper maps. However, the on-screen map does offer opportunities to work with moving and blinking symbols, and is very suitable for animations. They provide a strong method of visual communication, especially while they can deal with real data, as well as abstract and conceptual data. Animations not only tell a story or explain a process, they also have the capability to reveal pattern or relations which would not be evident if one would look at individual maps.
Cartographers have been tempted by animation ever since the sixties. Work has been done by Thrower (1961) Cornell and Robinson (1966 ) and Tobler (1970 ). However, the period only allowed for the non digital cartoon approach. During the eighties technological developments gave a second impulse to cartographic animation (see moellering, 1980; mounsey; 1982, and taylor, 1984). Currently a third move on animations is on going, created by questions from the GIS-environment (langran, 1989 ; monmonier, 1990 ; dibiase, 1991 1982 ; weber and buttenfield, 1993; openshaw et all, 1994 ). A historic overview is given by Campbell and Egbert ( 1990 ).
(Cartographic) animations are about change. Peterson (1995) expresses this as "what happens between each frame is more important then what exists on each frame". This should worry cartographers, since they mainly developed tools and rules for the design of static maps. How can we deal with this new phenomena? Is it possible to provide the producers of cartographic animations with a set tools and rules to create 'good' animations? For instance: "If your data are ......, and your aim is ......, then use the variables ......". To be able to do so, and to take advantage of knowledge of computer graphics developments and the 'Hollywood' scene, the nature and characteristics of cartographic animations have to be understood. However, the problem is that 'understanding' animations alone will not be of much help, since the environment where they are used, the purpose of their use, and the users themself will greatly influence 'performance'. Questions that arise here are 'What influence has an available interface to interact with the animation,?', 'Is an animation in a GIS environment used to explore, analyse or present spatial data?' and 'Are the viewers producers of the animation as well?' Figure 1 somehow summarizes these questions. Knowledge on map reading, in itself already limited, needs to be extended as well. How will the viewer tackle the problems of the many individual maps offered in a quick sequence? And, even when provided with interaction tools how would those influence the narrative story of the cartographic animation?
It will be clear that this paper will not be able to answer all these questions. It intends to provide a classification of cartographic animations which links up to existing cartographic theory, and suggest extension of this theory. Before the classification will be explained and demonstrated some terminology in respect to animation has to be clarified.

Cartographic animations

As mentioned before, animations can be very useful to clarify trends and processes, as well as to explain or provide insight into spatial relations. Based on which criteria can animations be classified for further study. Criteria could be application or visualization objective. The first (for instance socioeconomic and earth sciences) is very arbitrary, the second (exploration, analysis, or presentation) is less useful because the animations can be used for more then one objective. It is more natural to start the spatial data's components, location, attribute and time, in relation to display date. Display time can be described as the moment a viewer of an animation actually sees the images. A direct relation does exist between the individual frames and the display time as seen in figure 2. Cartographic animations are often categorized in temporal and non-temporal animations. In the first category display time is used to visualize real world time in a temporal sequence. In the second category display time is applied to explain spatial relations by presenting individual images in logical sequence.
When dealing with a temporal animation a direct relation between display time and world time exists. World time is the time scale of reality, the moment an event takes place in the real world. Examples of these animations are those of the Dutch coastline from Roman times until today, boundary changes in Africa since the second World War, or the changes of yesterday's weather. Time units can be seconds, weeks, or years. Temporal animation can also deal with time aggregates, for instance display weekly cycles. The GIS-environment distinguishes another type of time as well, database-time, e.g., the moment and real world event are registered in the database. Those three different types of time were already recognized, although not explicitly, by the cartographer working on topographic maps. Think of topographic map updates, where a difference of several years between world time, database time, and display time (respectively the moment a new road is built, the aerial photograph taken, and the final map printed). Temporal animations show change, change of the locational or attribute component of spatial data. As mentioned before, it is important the user can influence the flow of the animation.
Display time in non-temporal animations is not directly linked with world time. The dynamics of the map are used to show spatial relations or to clarify geometrical of attribute characteristics of spatial phenomena. Examples of non-temporal animations are:
-Display of choropleths with different classification methods used (peterson, 1993 );
-Other methods suggested (dransch, 1993 ) are to display a particular data set by changing the graphic representation, for instance subsequently show the same data in a dot map, a choropleth, a stepped statistical surface or an isoline map;
-Map with blinking symbols to attract attention to a certain location on the map. This could be used to indicate the accuracy of the mapped data (fisher, 1994 );
-A simulated flight through the landscape - a changing viewpoint of the user (Moellering's (1984) real time surface exploration);
-Some also include the pan and zoom concept (Dorling's (1992) animating space).

Classification of cartographic animations

The division in temporal and non-temporal animations is often taken for granted, and for many a cartographic animation is directly linked to temporal animations. However, the category non-temporal animation has much to offer, as can be seen in the list of examples at the end of the previous section. An attempt to categorize the non-temporal animation was made by Dransch (1992 ). It is her classification that will be elaborated and adapted here. Next to the temporal animations she recognizes animations through successive build-up, animations of changing representations, animations of changing content, and animations of changing combinations of data. An example of the first is the addition of subsequent map layers, This type leads the viewer through a theme, to help understand spatial and contextual coherence. An example of the second is to show a data set in different graphic representations, such as an isoline map, an smooth statistical surface or a dot map to provide the viewer a comprehensive impression of the same data set. The third shows different topics belonging to the same subject, and fourth categorize presents different interrelated subjects. The last two are somehow aggregations and not used here.
This leaves three types of cartographic animations, which are the core of the classification presented here. It is the relation between spatial data's components and display time which distinguishes them from each other.
Time series
These animations show change of spatial pattern patterns in (world) time. Since time is plotted against display time a transition between individual frames implies change in the data's locational and/or attribute component. Figure 3 shows two examples of this category, one with change in the locational component, and another with the change of attribute data. a town's build-up area. The other, of the attribute data, with change of the number of inhabitants of several neighbourhoods. However, temporal animations also include time-aggregates, indirect representations of world time.
Successive build-up
The coherence or of a data set (spatial internal relations), is shown by a division in subsets. In the centre of figure 3 two examples of animation of the successive build-up can be found. Throughout the animation spatial data's temporal component is fixed while location and/or attribute is/are plotted against display time. Changes in location or attribute take place and can effect each other.
Changing representations
This type of animation offers the viewer an extensive look at a particular data set. In this type of animations all three components are fixed. Graphic or data manipulations are plotted against display time. The changes observed by the viewer are due to manipulations with the data set from which graphics are produced or changes in the graphic representation itself. Figure 3 gives an example of each. The last five non- animation examples given in the previous section belong to this sub-category, including the fly-throughs and the pan and zoom concept, which give the viewer a look at the same data set in the same graphic representation but from a different perspective.

Towards design tools

The above classification should be looked upon from a cartographic design perspective. This because each category could require a different design approach. How can one design an animation to make sure the viewer indeed understands the trend or phenomena. The traditional graphic variables are used to represent the spatial data in each individual frame. Bertin, the first to write on graphical variables, had a negative approach to dynamic maps. He stated in his work (1967) : '....however, movement only introduces one additional variable, it will be dominant, it will distract all attention from the other (graphical) variables'. Recent research, however, proved this is not the case. Here we should remember that technological opportunities offered at the end of the sixties were limited compared with those of today. Koussoulakou and Kraak (1992) found that viewer of animations would not necessarily get a better or worse understand of the contents of an animation when compared with static maps. Dibiase et al (1992) found that movement would give the traditional variable new energy.
In this framework Dibiase introduced three so-called dynamic variables: duration, order, and rate of change. In 1994 MacEachren added frequency, display time and synchronization to this list:
The time at which some display change is initiated.
The length of time nothing changes at the display.
Frequency is the dual of duration. Either can be defined in terms of the other.
The sequence of frames or scenes.
The difference in magnitude of change/unit time (or m/d) for each of a sequence of frames or scenes.
Synchronization (phase correspondence) refers to the temporal correspondence of two or more time series.
In the animation literature the so-called animation variables have surfaced (hayward, 1984 ). They include size, position, orientation, speed scene, colour, texture, perspective (viewpoint), shot (distance), and sound. The last one is not considered here but can have an important impact (see kryger, 1994 ). These variables are shown in figure 4 in relation to the graphic and dynamic variables. From this figure it can be seen that Bertin's graphic variables each have a match with one of the animation variables. From the dynamic variables only order and duration have a match.
Animations have a narrative character. They tell a story. The dynamic variables can be seen as additional tools to design an animation. With those one can control all visual manipulations. Most prominent of these variables are duration and order, which have a strong impact on the animation's narrative character (see also figure 4). They define the order of the images, and the time they are visible. Since the other dynamic variables are a functions of or dependent order and duration (display date and frequency) or more or less data dependent (rate of change) or special (synchronization - multiple temporal data sets), research concentrates on effects of duration and order. In literature describing animations the same pattern can be found: In sixteen animation descriptions found only four explicitly use other variables then duration and order.
Duration and order can not only be used to influence the narrative character of an animation (see figure 5). It is also possible to directly use the dynamic variable to represent spatial data (their representational character). An example of this use is duration to show uncertainty. Those parts of the map with high certainty are stable and those with a low certainty appear chaotic by blinking dots (fisher, 1994 ). Another application of the dynamic variables is called dynamic symbolization. Here they are used to strengthen the graphical variables. Maps with blinking symbols (use of frequency) are good examples. All three levels can be present in an animation, and have to be understood.
In the scheme in figure 6 the cartographic animation categories are plotted against the dynamic variable duration and order. From it becomes clear that order is determined by the animation type. It somehow becomes a design metaphor for the aspect of the data to be visualized. Duration is used to put relative emphasis on specific parts of the animation, and can be seen as a clear design variable on the narrative level.

Conclusions

To better understand the nature of cartographic animation in relation to their design the different types of animation have been categorized in three groups: time series, successive build-up and changing representations. Sample animation are available via the author. To be able to suggest design rules the classification was linked to cartographic theory by mean of the dynamic variables duration and order.
The validation of the classification will be proved/disapproved while testing the effect of the dynamic variables in relation to the user interface in a particular design environment. These research results will contribute to a interactive design tool for dynamic maps.
References
Bertin, J. (1967) Semiology Graphique. Den Haag:
Mouton.
Campbell, C.S. & S.L Egbert (1990) Animated cartography / thirty years of scratching the surface. Cartographica 27, (2), 24-46.
Cornell, B. & A.h. Robinson (1966) Possibilities for computer animated films in cartography. The Cartographic Journal 3, (1), 79-82.
Dibiase, d. et al. (1991) Animated cartographic visualisation in earth system science. Proceedings 15th ICA Conference. Bournemouth, 1, 223-232.
DiBiase, D., A.M. MacEachren, J.B. Krygier, and C. Reeves (1992) Animation and the role of map design in scientific visualization. Cartography and Geographic Information systems, 19, (4), 201-214.
Dorling, D. (1992) Stretching space and splicing time: from cartographic animation to interactive visualisation. Cartography and Geographic information Systems 19, (4), 215-227.
Dransch, D. (1993) Kartographische Animation. In: Grunreich, D. (ed.) Kartographische Schriften, Kartographie und Geo-Informationssysteme, grundlagen, entwicklungsstand und trends. Bonn: Kirschbaum Verlag, pp. 3-15.
Egbert, S.L. and T.A. Slocum (1992) "EXPLOREMAP": An exploration system for choropleth maps. Annals of the Association of American Geographers 82 (2), 275-288.
Fisher, P. 1994, Animation and sound for the visualization of uncertain spatial information. In: Visualization in Geographical Information Systems edited by H.M. Hearnshaw and D.J. Unwin 181- 185 (London: J. Wiley).
Hayward, S. (1984) Computers for animation. Norwich: Page Bros. Ltd.
Koussoulakou, A. & M.J. Kraak (1992) Spatio-temporal maps and cartographic communication. The Cartographic Journal, 29, (2), 101-108.
Kraak, M.J. (1994) Interactive Modelling Environment for 3d Maps, Functionality and Interface issues. In: Visualization in Modern Cartography. Pergamon Press, (eds. F.Taylor & A. MacEachren).
Kraak, M. J. & A.M. MacEachren (1994) Visualization of spatial data's temporal component. In Advanced in GIS Research (eds: Waugh, T.C. & R.G. Healey). Proceedings
Spatial Data Handling, Edinburgh, 1, 391-409
Kraak, M.J. & F.J. Ormeling (1996) Cartography, visualization of spatial data. London: Longman.
Krygier, J. (1994) Sound and cartographic visualiza-
tion. In: MacEachran, A.M. and D.R.F.Taylor 1994 Visualization in modern cartography. Modern Cartography series. Oxford: Pergamon.
Langran, G. (1989) A review of temporal database research and its use in GIS applications. International Journal of Geographic Information Systems 3, (3), 215-232.
MacEachren, A. M. (1994) How Maps Work: Issues in Representation & Design, New York: Guilford Press.
Moellering, H. (1984) Real maps, virual maps and interactive cartography. In Spatial statistics and models (eds: G.L. Gaile & C.J. Wilmott) . Dordrecht: Reidel Publishing, 109-132.
Moellering, H. (1980) The real-time animation of three-dimensional maps. The American Cartographer 7 , (1), 67-75.
Monmonier, M. (1990) Strategies for the visualization
of geographic time-serie data. Cartographica 27, (1), 30-45.
Monmonier, M. (1989) Geographic Brushing: enhanching exploratory analysis of the scatterplot matrix. Geographical analysis . 21, (1), 81-84.
Mounsey, H. (1982) Cartography of time-changing phenomena - the animated map. Unpublished PhD thesis University of Durham.
Openshaw, S, D. Waugh & A. Cross (1994) Some ideas about the use op map animation as a spatial analysis tool. In: Visualization in Geographical Information Systems edited by H.M. Hearnshaw and D.J. Unwin 131- 138 (London: J. Wiley).
Peterson, M.P. (1993) Interactive cartographic animation. Cartography and Geographical Information Systems 20, (1), 40-44.
Peterson, M.P. (1995) Interactive and animated cartography. Englewood cliffs: Prentic Hall.

Taylor, D.R.F. (1994) The creation and design of maps for Videotex systems. Technical papers of the Austra CartoOne Seminar. Perth, 277-290.
Thrower, N. (1961) Animated Cartography in the United States. International Yearbook of Cartography, 20-28.
Tobler, W.R. (1970) A computer movie: simulation of population change in the Detroit region. Economic Geography 46, (20), 234-240.
Weber, R. & B.P. Buttenfield (1993) A cartographic animation of average yearly surface temperatures for the 48 contiguous United states: 1897-1986. Cartography and GIS 20, (3), pp.