David Howard
Introduction
Geographic visualization (GVIS) is becoming recognized as an important tool in the explorative stages of research. The power of GVIS is rooted in the ability of the human eye-brain system to recognize patterns in the visual field (MacEachren and Ganter 1990). GVIS encourages the creation and inspection of multiple views of data sets, thus allowing researchers to search for patterns that might merit fruitful investigation.
One approach to the study of environmental change is integrated assessment. Integrated assessment focusses on the interactions between the physical world and human actions. These interactions involve both those human actions that alter aspects of the environment and the physical manifestations of environmental systems that affect what humans value.
The study of global environmental change is an increasingly important topic in many social sciences. Geographers, physical scientists, economists, anthropologists, etc. are all working (sometimes together) on explanations for the changes that are currently happening in our environment and the effect that these changes will have on environmental conditions in the future. Global environmental change is not new; the climatic conditions of the world have not been stable in the past and there is no reason to expect that they would have settled into totally stable patterns in the present. However, current environmental changes are different from those in the past in two ways: 1) they are occurring at much faster rates, on the order of decades and centuries rather than millenia; and 2) the causes of the changes are largely anthropogenic in origin (NRC, 1992).
Global environmental change involves much more than physical changes in the environment. Human impacts on the environment and impacts of the environment on human lives, possessions and values are also wrapped up in the study of changes of the environment we live in. This integrated view of both human systems and environmental systems is crucial to understanding both the changes in the environment and the impacts of environmental change.
Integrated assessment is an paradigm for the study of environmental change that incorporates human dimensions as well as physical changes. The integrated assessment model (4.5K) 'begins' with human causes of environmental changes. The causes part of the model is broken into two components: social drivers of human activities and the human activities themselves. This breakdown recognizes that human activities do not happen in a cultural vacuum but are seated in social norms, societal rules, and other drivers. These human activities spark climate change which causes, for instance, hydrologic and water resources change. At this point in the model various changes could be studied, such as average temperature change or change in atmospheric composition. Whatever the specific area of interest, any of these environmental changes have socioeconomic impacts which provoke human responses. The responses can be adaptations at either the individual or societal level or they can be policy directed at mitigating the effects of the environmental change. The responses then directly affect the causes of environmental change and indirectly cause more change in the climate and the environment and thus further socioeconomic impacts (Yarnal, 1996).
One way to start studying global environmental change is to examine environmental change at smaller spatial scales. The Susquehanna River basin covers most of central and eastern Pennsylvania as well as small sections of both Maryland and New York. It serves as a good region to study hydrologic and water resource change because, as a river basin, it is relatively easy to keep track of the inputs to the system and the outputs from the system. Another reason that this region is ideal stems from the fact that most of the area falls within one state. Thus there are less complications due to differing state policies affecting the basin. This simplifies the study of the human factors of environmental change in the area. Hopefully the lessons learned by studying the Susquehanna River basin will be generalizable to other drainage basins across the East Coast of the United States.
The Susquehanna River Basin Integrated Assessment (SRBIA) is still in the data gathering stages. (The physical side of the SRBIA project can be examined at the SRBEX (Susquehanna River Basin Experiment) site.) Once data collection is complete, exploration and analysis will begin. The data being gathered include physical measurements of rainfall, streamflow, stream pollution, water storage, etc. as well as measurements of human impacts such as land use, use of fertilizer, industrial pollutant production, etc. Also being collected are data relating to the social drivers. One example of this type of data is information about laws and policies that affect human activities and the dates that these rules were put into effect (Yarnal, in press).
Geographic visualization tools will be very helpful when the researchers begin to explore the data that has been gathered and look for relationships. The relationships between environmental systems and human systems are still indefinite and thus tools that help analysts to both formulate and confirm hypotheses are especially valuable. This conforms to the first three stages of DiBiase's (1990) model (15K) of visualization, exploration, synthesis and confirmation.luable in the future. The visualization system created in conjunction with the SRBIA project will be another example of Koussalakou's third type of visualization.
MacEachren (MacEachren, 1992) has raised two questions related to the visualization of geo-referenced data and data reliability:
In addition, the exploration of temporal data has been proposed as an important area for geographic analysis (Peuquet, 1994). The data gathered in the SRBIA project offers opportunities to examine both temporal and reliability information. Both the temporal aspects of the relationships and the reliability of the data and of data manipulations are of interest to the SRBIA analysts. The prototype visualization system will provide tools for the exploration of both, although it will focus on reliability data.
User interface design is a crucial aspect of the construction of visualization environments because the interface needs to be powerful and flexible enough to satisfy the needs of the scientists using it (Lindholm and Sarjakoski, 1994). However, the interface should also be easy to use; the analyst should not have to spend precious time learning a non-intuitive interface. The best interface is one that is not even noticed by the user because it is easy to initiate any operation and contains every operation that is desired. Mark and Gould (1991) suggest that interface designers need to focus on human-problem or human-data interactions rather than the traditional human-computer interaction perspective. Their philosophy of user interface design is that "Σthe prime objective should be to enhance user intaction with geographic information and with geographic problem-solving, rather than with software or hardware."
User interface design research has a long history in computer science and engineering research. For example, Carroll and Thomas (1982) reported on the role of metaphor in computer interfaces and Polson and Lewis (1990) discussed a theoretical foundation for the design of easily learned interfaces. In cartography, interest in user interfaces is more recent. Much of the recent research has been in the specific area of improving interfaces for GIS. Mark (1992) discusses different interaction paradigms and matches them with the ways that humans interact with their own environment. For example, the direct manipulation paradigm is the result of a metaphorical mapping from the way humans perceive haptic space--based on experiences of touch and movement through space--onto the way humans interact with digital data in a computer.
Currently, it seems that the most frequent use of environmental data exploration is to inform decision-makers about environmental processes and human effects upon those processes. Because of this use of visualization to aid in the making of policy decisions, increased attention is being paid to issues of environmental data reliability. The National Center for Geographic Informtion and Analysis (NCGIA), concerned with the quality of cartographic materials used in GIS-facilitated decision-making, issued an initiative on the "Visualization of Data Quality" (Beard and Buttenfield 1991). This initiative was an appeal to develop methods for graphically incorporating reliability information directly into maps. Numerous cartographers and others responded to the NCGIA's initiative. Beard and Mackaness (1993), Goodchild et al (1994), MacEachren (1992), and McGranaghan (1993), among others, all have suggested methods for depicting reliability of point, line and area data. Monmonier (1993) also contributed to the endeavor with ideas about the use of graphic scripts to discern limitations and statistical characteristics of mapped data. Meanwhile, Fisher (1993) and Krygier (1994) have explored methods of expressing reliability through sound.
In order to build upon this body of work about the depiction of reliability, it is necessary to explore the question of whether or not the user of a geographic visualization be able, or willing, to use reliability information when it is provided.
The visualization system will be developed in concert with the project members. This will allow the system design to parallel the needs of the main users. System design will follow a three-level hierarchical design proposed by Howard and MacEachren (in press). The three levels are the conceptual level, the operation level and the implementation level.
The conceptual level is concerned with the whole system acting as an interface between the system users and information, rather than with specifics of the connections between a user and the system. Howard and MacEachren (in press) offer four questions to help system designers at this level:
At the operation level of interface design, operations are defined that help to meet conceptual level goals. These operations are not necessarily singular tools; they are usually a set of tools or functions that all linked in some way. For instance, with a goal of studying temporal variation in data sets, an animation operation would be useful. This animation operation might involve several tools such as flickering between two time-slices to compare them, animations of all of the time-slices to find changes that happen over time, and an animation of a system diagram that would show the varying inputs and outputs for each part of a hydrologic system.
Finally, the implementation level involves the look and 'feel' of the system. At this level the designer is concerned with taking the conceptual goals and the operations that were delineated in the first two levels and turning them into a working system on the hardware and software available.
The goals of the SRBIA visualization system are to allow the researchers involved with the project to explore multiple physical and human data sets in order to search for patterns and relationships that may help explain some of the interactions between physical and human systems. The researchers will also be able to explore the temporal aspects of the various data sets and their relationships. Lastly, reliability data will be made available so that researchers will have an idea about the certainty with which they can make predictions and suggestions about further action.
The goals will be reached by providing users with a visualization system that allows the inspection of singular and multiple data sets in both separate and combined ways. Users will be able to select data sets for display and choose among several display methods in order to examine multiple views of the data sets and the possible relationships between them. Through this exploration of the physical and human data sets, it is hoped that analysts will be able to increase their understanding of the ways that humans influence environmental change and how that change affects the ways humans live.
The primary users of the system will be the members of the SRBIA project. These are mainly physical and social scientists at Penn State. Some of the members of the project are grad students in geography and others are faculty members associated with the Earth Systems Science Center.
Buckley (submitted) suggests evaluating the expertise of the users based upon their familiarity with (1) the nature of the data, (2) the content of the data, (3) the visual format, and (4) the media. With respect to these four characteristics, the users of the SRBIA visualization system can be assumed to be experts with respect to both the nature and content of at least some of the data and some of the users will be experts regarding the visual format because of their familiarity with maps and mapmaking. However, although many of the users will be familiar with the hardware, they will know nothing of the software and thus their expertise with the media is questionable.
I will be evaluating the usefulness of the prototype system concentrating on the user interface and the reliability representations. In the evaluation of this system, I will employ the expert/novice paradigm. This approach "examines how experience, prior knowledge and training in a specific domain affect perception, comprehension and problem solving in that domain." (McGuinness, 1994) McGuinness presents several questions that are relevant to expert/novice studies:
Expert-novice studies date from the mid-1960's when de Groot (1965) studied differences in the thought processes of master and non-experienced chess players. His research led to the conclusion that the number of ideas does not change between experts and novices but that the quality of ideas differs. Following this research, several studies were done about the memory structures of experts and novices (Chase and Simon, 1973; Egan and Schwartz, 1979) and it was found that experts could remember meaningful configurations from within their field of expertise (i.e., a mid-game configuration of pieces for chess masters or electric circuit diagrams for expert electricians) much better than could novices. However, expert and novice memory for random configurations is the same. This indicates that experts are able to form units or layers when thinking about or memorizing patterns from situations in which they are knowledgable. Expert novice map reading studies indicates that novices look at the features on a map while experts look 'beneath' the features to find patterns and relationships (Williamson and McGuinness, 1990). These studies also indicate that, like master chess players and expert electricians, experiences map readers remember maps differently and better than novices (Chang et al, 1985; Gilhooly et al, 1988). The body of research on the differences between experts and novices indicates that this distinction is important in geographic visualization. Memory of spatial patterns, comparison of remembered patterns with displays, and the ability to discover relationships are all seemingly crucial abilities for the use of GVIS. An expert-novice approach highlights the experience that gives people these abilities and evaluates the importance of this experience.e subjects' cognitive processes. By asking subjects to speak their thoughts aloud as they use the system and recording these thoughts, I can analyze what subjects were thinking about as they used RVIS. Crampton (1992) used this procedure to great effect when studying expert and novice orienteers.
The SRBIA project will serve as an important step in the journey to understanding global environmental change and human-physical interactions. However, in order to make sense of the massive amounts of data being collected, they need a way to easily explore the multiple data sets and look for relationships between them. The prototype visualization system will provide this and therefore will hopefully aid researchers with hypothesis generation and confirmation.
Beard, K., and Mackaness, W. (1993). "Visual Access to Data Quality in Geographic Information Systems." Cartographica, 30(2 & 3), 37-45.
Beard, M. K., and Buttenfield, B. P. (1991). "NCGIA Research Initiative 7: Visualization of Spatial Data Quality." 91-26, NCGIA.
Buckley, A. M. (submitted). "Visualization of Multivariate Geographic Data for Exploration." Draft submitted for Young Scholar's Institute in GIS; Berlin, Germany; June 24-August 1, 1996.
Carroll, J. M., and Thomas, J. C. (1982). "Metaphor and the Cognitive Representation of Computing Systems." IEEE Transactions on Systems, Man, and Cybernetics, SMC-12(2), 107-116.
Chang, K., J. Antes and T. Lenzen (1985) "The Effect of Experience on Reading Topographic Relief Information: Analysis of Performance and Eye Movements", The Cartographic Journal, Vol. 22, pp. 88-94.
Chase, W. and H. Simon (1973) "Perception in Chess", Cognitive Psychology, Vol. 4, pp. 55-81.
Crampton, J. (1992). "New Directions in the Information Era - A Reply to Taylor (1991)." The Cartographic Journal, 29(December 1992), 145-150.
de Groot, A. (1965) Thought and Choice in Chess, Morton, The Hague.
DiBiase, D. (1990). "Visualization in the Earth Sciences." Earth and Mineral Sciences, 59(2), 13-18.
DiBiase, D., Reeves, C., MacEachren, A. M., Krygier, J. B., Wyss, M. v., Sloan, J. L., and Detweiler, M. C. (1994). "A Map Interface for Exploring Multivariate Paleoclimate Data." .
Egan, D. and B. Schwartz (1979) "Chunking in Recall of Symbolic Drawings", Memory and Cognition, Vol. 7, pp. 149-158.
Fisher, P. F. (1993). "Visualizing Uncertainty in Soil Maps by Animation." Cartographica, 30(2 & 3), 20-27.
Gilhooly, K. J., Wood, M., Kinnear, P. R., and Green, C. (1988). "Skill in Map Reading and Memory for Maps." The Quarterly Journal of Experimental Psychology, 40A(1), 87-107.
Goodchild, M., L. Chih-Chang and Y. Leung (1994) "Visualizing Fuzzy Maps" in Hearnshaw, H. and D. Unwin (eds.) Visualization in Geographic Information Systems, John Wiley and Sons, New York, pp. 141-149.
Howard, D. L. and A. M. MacEachren. (in press). "Interface Design for Geographic Visualization: Tools for Representing Reliability." Cartography and Geographic Information Systems.
Koussoulakou, A. (1994). "Spatial-Temporal Analysis of Urban Air Pollution." Visualization in Modern Cartography, A. M. MacEachren and D. R. F. Taylor, eds., Elsevier Science, Oxford, 243-268.
Krygier, J. B. (1994) "Sound and Geographic Visualization" in MacEachren, A. M. and D. R. F. Taylor (eds.), Visualization in Modern Cartography, Elsevier Science, Oxford, pp. 149-166.
Lindholm, M. and T. Sarjakoski (1994) "Designing a Visualization User Interface" in MacEachren, A. M. and D. R. F. Taylor (eds.), Visualization in Modern Cartography, Elsevier Science, Oxford, pp. 167-184.
MacEachren, A. M. (1992). "Visualizing Uncertain Information." Cartographic Perspectives, 13, 10-19.
MacEachren, A. M., and Ganter, J. H. (1990). "A Pattern Identification Approach to Cartographic Visualization." Cartographica, 27(2), 64-81.
Mark, D. M. "Spatial Metaphors for Human-Computer Interaction." , 104-112.
Mark, D. M., and Gould, M. D. (1991). "Interacting with Geographic Information: A Commentary." Photogrammetric Engineering and Remote Sensing, 57(11), 1427-1430.
McGranaghan, M. (1993). "A Cartographic View of Spatial Data Quality." Cartographica, 30(2 & 3), 8-19.
McGuinness, C. (1994) "Expert/Novice Use of Visualization Tools" in MacEachren, A. M. and D. R. F. Taylor (eds.), Visualization in Modern Cartography, Elsevier Science, Oxford, pp. 185-199.
Monmonier, M. (1993). "Exploring the Quality of Enumeration-Area Data with Graphic Scripts." Cartographica, 30(2 & 3), 28-36.
National Research Council, ed. (1992). Global Environmental Change: Understanding the Human Dimensions, Chapters 1,2 and 3 (P. Stern, O. Young, and D. Druckman, editors). Washington, DC: National Academy Press.
Peuquet, D. (1994). "It's About Time: A Conceptual Framework for the Representation of Temporal Dynamics in Geographic Information Systems" Annals of the Association of American Geographers 84(3), 441-461.
Polson, P. G., and Lewis, C. H. (1990). "Theory-Based Design for Easily Learned Interfaces." Human-Computer Interaction, 5, 191-220.
Rhyne, T., Bolstad, M., Rheingans, P., Peterson, L., and Shackelford, W. (1993). "Visualizing Environmental Data at the EPA." IEEE Computer Graphics and Applications, 13(2), 34-38.
Yarnal, B. (1996). Personal communication.
Yarnal, B. (in press). "The Policy Relevance of Global Environmental Change Research." Global and Planetary Change.
