CaGIS Vol. 27, No. 4 (2000)

Introduction - Terry Slocum

With this issue, I officially take over the editorship of Cartography and Geographic Information Science (CaGIS) from Bob Cromley. It is my intention and the goal of everyone involved in publishing CaGIS to provide, in our journal, a fair representation of the research context in which cartographers and geographic information scientists from around the world are working. As in the past, the journal will continue to publish papers with a cartographic emphasis, but we will also be interested in papers that might be labeled "pure GIS." The key is that papers should provide theoretical advancements in cartography and GIScience.

Several special issues are planned for the coming two years. For 2001, Alan MacEachren and Menno-Jan Kraak will co-edit Research Issues in Geographic Visualization, and Robert McMaster will edit Generalization Issues in Cartography. For 2002, David Woodward and Mark Monmonier will co-edit the History of 20th Century Cartography. Other special issues we are considering include GIS and public health, cognition, community mapping, geocomputation, open GIS, internet-based teaching materials, and visualization (covering completed research as opposed to research issues). I encourage every cartographer to contribute to our effort to make Cartography and Geographic Information Science one of the leading voices of our discipline and the broader field of spatial data representation and use.

In the transition period between June 1999 and October 2000, I have received more than 30 manuscripts for publication in the journal. The three appearing in this issue are an example of the topics readers may expect to find appearing in subsequent issues. Because this is the last issue in volume 27, I would also like to thank reviewers who have given graciously of their time to complete reviews.

Reviewers in 1999-2000

Marc Armstrong, David Bennett, Cynthia Brewer, Aileen Buckley, Barbara Buttenfield, Nicholas Chrisman, Keith Clarke, Helen Couclelis, William Craig, Robert Cromley, Geoffrey Dutton, J. Ronald Eastman, Stephen Egbert, Johan Feddema, Scott Freundschuh, Philip Gersmehl, Patricia Gilmartin, Michael Goodchild, Stephen Guptill, Francis Harvey, Michael Hutchinson, Piotr Jankowski, Christopher Jones, Peter Keller, Karen Kemp, Fritz Kessler, Menno-Jan Kraak, John Krygier, Mark Kumler, Nina Lam, Mitchel Langford, Peter Laskowski, Jay Lee, Robert Lloyd, William Mackaness, Alan MacEachren, Scott Mackay, George McCleary, Robert McMaster, Janet Mersey, Mark Monmonier, Joel Morrison, Timothy Nyerges, Nancy Obermeyer, Michael Phoenix, Thomas Poiker, Keith Rice, Dianne Richardson, Robert Rundstrom, Robert Rugg, Alan Saalfeld, Eric Sheppard, Daniel Steinwand, David Theobald, Derek Thompson, Barbara Tversky, E. Lynn Usery, Denis White, David Woodward, Michael Worboys, May Yuan

Designing Effective Bivariate Symbols: The Influence of Perceptual Grouping Processes

Elisabeth S. Nelson

The purpose of this research was to empirically assess perceptual groupings of various combinations of symbol dimensions (e.g., graphic variables) used in designing bivariate map symbols. Perceptual grouping ability was assessed using the theory of selective attention, a construct first proposed in psychological research. Selective attention theory contends that one's ability to analyze a symbol's dimensions-such as color or size-is affected by other dimensions present in the same symbol. Symbol dimensions are described as either separable (capable of being attended to independently of other dimensions), integral (cannot be processed without interference from other dimensions), or configural (i.e., show characteristics of both integrality and separability, which may also form new, emergent properties). Without empirical evidence describing such interactions for various combinations of symbol dimensions, cartographers cannot truly evaluate the functionality of the symbols they use on maps. The symbol dimensions or graphic combinations chosen for this study were selected to incorporate a wide range of traditional cartographic symbolization, including line and lettering symbolization, areal shading, dot patterns, and point symbols. Combinations were examined in an abstract setting using a speeded classification task, which is the traditional means of studying selective attention. Subject reaction times provided an assessment of the levels of integrality, separability, and configurality. Results suggest that most symbol dimension combinations are either separable or exhibit evidence of asymmetrical dimensional interactions. Findings from this study will be integrated into subsequent experiments, the results of which will assist cartographers in the design of complex map symbols.

Developing a Geographic Visualization Tool to Support Earth Science Learning

Mark Harrower, Alan MacEachren and Amy L. Griffin

This paper reports on the development and assessment of the EarthSystemsVisualizer (ESV), a geovisualization tool designed to facilitate learning about global weather. Our goals in designing ESV were to evaluate two exploratory spatial data analysis (ESDA) techniques, temporal brushing and temporal focusing, and to determine whether interactive geovisualization tools influence problem solving strategies, approaches to learning, and students' ability to generate hypotheses about earth-science processes. Focus group sessions were conducted with both expert and novice users to assess an initial design for the ESV interface prior to conducting a task-based assessment of ESV use. Changes were implemented in response to the focus group results, including the redesign of a temporal legend and improved speed and direction controls. Our task-based assessment considered student reactions to components of ESV, especially whether they could use it to answer questions about global-scale weather processes, and whether the system (particularly its focusing and brushing tools) had an impact on the hypotheses generated about relationships among weather variables. The assessment revealed that focusing and brushing had little impact on students' ability to answer questions about weather processes, and that performance suffered for students who were confused by the focusing and brushing tools. In fact, students who understood the tools performed the best, but students without the tools performed better than those who had the tools but were unsure how to use them. We also concluded that the level of the visualization system must be well matched to the knowledge users have about the application domain: students who already possessed an advanced understanding of meteorology or climatology benefited less and were more critical of the system than students with an intermediate or a novice level of understanding.

KEYWORDS: Map animation, geographic visualization, interactive cartography, temporal legends, spatiotemporal analysis, user testing

Interrogating Land Cover Categories: Metaphor and Method in Remote Sensing

Paul Robbins and Tara Maddock

The increasing sophistication of classification techniques used in land use and land cover analysis has not been matched by attention to the origin and effects of land cover categories. While classifications appear unproblematic and self-evident, they carry with them their own histories, meanings and effects, which remain largely unexamined. In an effort at such scrutiny, we examine the origins of land cover categories deployed in remote sensing and conclude that categories are theory-laden metaphors and occur epistemologically prior to any clustering algorithm, no matter how sophisticated. We describe the problematic effects that the imposition of classification systems in place of in situ knowledge of the landscape can have, especially in a colonial or post-colonial context. As an alternative to imposed classification, we propose and demonstrate an empirical technique based upon a growing body of work in participatory GIS. The method compares image classifications based on local and expert knowledge, using a case study from Rajasthan, India, concluding that differing metaphors of landscape lead to divergent measures of land cover.

KEYWORDS: Land use and land cover, remote sensing, metaphors, participatory geographic information science, Rajasthan, India

Book Reviews

Remote Sensing of the Environment: An Earth Resource Perspective

John R. Jensen. Prentice Hall, Upper Saddle River, New Jersey. 2000. ISBN 0-13-489733-1. xvi and 544 pp, appendix, photos, maps, diagrams, tables, 32 pp of color plates. Hardcopy $91.

Reviewed by: James W. Merchant, University of Nebraska-Lincoln

John R. Jensen's new book, Remote Sensing of the Environment: An Earth Resource Perspective, is a very welcome addition to the existing array of texts dealing with remote sensing, and a most able complement to his widely used book on digital image analysis, Introductory Digital Image Processing: A Remote Sensing Perspective (2nd ed., Prentice Hall, 1996). In the preface of the new book it is noted that the "book was written for physical, natural and social scientists interested in how remote sensing ...can be used to solve real-world problems." In this task, it succeeds admirably.

Remote Sensing of the Environment is comprised of 13 chapters, the first nine of which cover the physical fundamentals of remote sensing; principles of electromagnetic radiation; aerial photography and aerial platforms; elements of visual image interpretation and photogrammetry; multispectral remote sensing systems; thermal infrared systems; active and passive microwave remote sensing techniques; and lidar remote sensing. Four concluding chapters deal with applications of remote sensing in the vegetation sciences; water resources sciences; the urban landscape; and investigations of soils, minerals and geomorphology. Each of the book's chapters includes several pages of references, many from journals published as recently as 1999. The text is, in fact, distinguished throughout by its up-to-date coverage, including, for example, thorough treatments of both the IKONOS and Terra satellite systems launched in 1999, and of sensors such as EO-1 that are scheduled to become operational in the near future. An unusually detailed six-page table of contents supplements a comprehensive index to enable the reader to easily find material of interest.

The narrative is enhanced by hundreds of well selected and well designed tables, diagrams, photographs, maps and images, including 32 pages of color plates. The reproduction of both black and white and color graphics is uniformly excellent, and enhanced by the book's relatively large (8.5 x 11 inch) format. Indeed, the artful design of the integrated text and graphics makes for a far more compelling and interesting presentation than that offered in other contemporary introductory remote sensing textbooks.

The book concludes with a thorough appendix that provides guidance to additional sources of information on remote sensing, including listings of major textbooks, on-line tutorials, professional societies, national space agencies, major journals, and sites for acquiring aerial photography and multispectral and radar data. Internet addresses are provided for virtually all listings. To augment the text and to provide for updates, the author has established two Internet sites at http://www.cla.sc.edu/geog/rsbook/links/ (still under construction as of August 2000) and http://www.cla.sc.edu/geog/rsbook/exercises/. The latter site provides 13 laboratory exercises keyed to the chapters in the book.

Remote sensing is a large, diverse, and rapidly changing area of technology with myriad applications. No single book can be expected to cover all aspects of the field in equal depth, yet Remote Sensing of the Environment is outstanding in a great many areas. Among the book's many strengths are its discussions of energy-matter interactions; physical characteristics and operation of virtually every significant sensor (past, present and near-future); analysis and applications of hyperspectral data; formulation and application of vegetation indices; and estimation of biophysical parameters via remote sensing. Complex topics such as radar interferometry and the bidirectional reflectance distribution function are treated with technical rigor, but in such a way that they will be readily comprehensible to most readers.

Of course, in order to keep the book a reasonable length, some compromises in subject matter were clearly inevitable. Procedures for collecting ground truth and methods for assessing the accuracy of products derived from remote sensing (e.g., development and analysis of error matrices) are, for example, occasionally alluded to, but never presented in detail. Moreover, there is scant discussion of digital image analysis techniques (e.g., multispectral classification), even though images and maps derived via digital image processing are scattered throughout the book (e.g., plates 10-2 and 10-5). In the preface, it is explicitly stated that this book is intended as a companion volume to Jensen's previously published text on digital image processing (in which, incidentally, accuracy assessment is also discussed). Nonetheless, Remote Sensing of the Environment would be strengthened by an addition of an elementary, conceptual overview of some important techniques in digital image analysis, and a brief introduction to accuracy assessment. Such additions would better enable the book to serve as a stand-alone text for a one-semester survey course in remote sensing of the sort that most in the book's intended audience are likely to take.

The many virtues of Remote Sensing of the Environment: An Earth Resource Perspective set it quite apart from other contemporary works covering some of the same subject matter. It is a book that would serve as a superior text for an introductory and/or intermediate level course in remote sensing. As a reference, the volume should be in the library of virtually every remote sensing specialist.

GIS and Health

Anthony Gatrell and Markku Löytönen. 1998. Taylor & Francis, Philadelphia, Pennsylvania. ISBN 0-7484-07790. 212 p.

Reviewed by: Deborah Thomas, University of Colorado at Denver

The recent burgeoning of GIS applications in public health calls for a thorough appraisal of the use of this technology within the discipline. Gatrell and Löytönen compiled a book that not only presents the complexity of using GIS for research in public health, but also suggests future directions for GIS applications in this area. This book is the sixth in the GISDATA Series edited by Masser and Salgé that highlights advances in the European GIS community. Even with a European focus of the volume, many of the topics presented are relevant to a broader set of GIS and public health researchers.

The book is divided into two sections: methodological issues and applications. The editors set the tone of the book in Chapter 1. By their own admission, the focus of the book is on environmental/spatial epidemiology rather than on more extensive public health issues, such as access to health care or equity of provided services. Although several of the chapters briefly touch on broader public health questions, a distinct bias toward epidemiology exists throughout the book.

The next six chapters focus on methodological issues that have come up with the use of GIS for public health analysis. Jacquez leads the way by cautioning against the "gee whiz" effect that GIS can have when misleading maps are produced. Instead, the author contends that GIS should be used within an appropriate study design and that the same set of data cannot be used to both generate and test a hypothesis. Haining and Kulldorff concentrate almost entirely on spatial statistical analysis in the next two chapters, barely mentioning GIS. Admittedly, the use of GIS is closely linked to spatial statistics and thus, one cannot be discussed without the other. In fact, many of the data and methodological issues presented are directly relevant to GIS. The discussion turns to air pollution modeling when Collins illustrates how environmental modeling and health data can be combined in a GIS environment. In the last chapter of this section, Löytönen deals with the issue of temporal and spatial analysis. The author notes that people move through space and experience differing environmental exposures, depending on where they have lived.

The next seven chapters highlight GIS applications in public health, most of which fall into the category of hypothesis generation. The first two chapters of this section focus on Italy. Trinca illustrates how the use of GIS is expanding rapidly in the study of environmental health via several successful applications in Italy. The author notes that many institutional and perceptual barriers to using the full capabilities of GIS still exist. Braga et al. present an interactive mortality atlas developed using GIS and statistical analysis that has expanded the access to this particular type of health data at a national scale. Van den Berg also portrays the establishment of an interactive spatial information system for Pomerania, Germany. Both of these interactive systems highlight challenges encountered with the visualization of data; integration of datasets; lack of geographic understanding among public health professionals; and difficulty in integrating statistical methods directly with a GIS. Another case study by López-Abente explores the use of Bayesian Analysis for assessing cancer clusters, utilizing GIS primarily for mapping, and again stressing the necessity for interfacing GIS and spatial statistical techniques. The cancer theme is continued by Teppo, who examines the potential for analyzing cancer data in a GIS context in Finland. Shifting to the U.K., Wilkinson et al. explore the uses of GIS in public health, primarily focusing on the potential for improving population data. Lovett et al. continue this topic by looking at the potential for using patient registers in public health analysis.

Taken together, the methodological discussions and the case studies depict the status of GIS in public health analysis. Even though the chapters cover a lot of ground, several overarching themes reappear throughout the book. The collection effectively illustrates the limitations specific to the use of GIS within the context of environmental epidemiological studies. These limitations partly arise from the nature of health data, such as spatial accuracy or aggregation, but also from a lack of geographic knowledge within the public health profession. Those who often end up interpreting the maps have little geographic or GIS training and may be unfamiliar with map interpretation or the technology's limitations. Nearly all the authors propose incorporating GIS into study design appropriately. They also maintain that one of the primary strengths of GIS is in the integration of disparate datasets. There seems to be a general consensus that the best use of GIS is for hypothesis generation, at least in its current state of development.

Perhaps arising from the focus on environmental epidemiology, several topics related to GIS and public health receive little attention. While many of the authors mention issues of confidentiality, the unique difficulty of sharing public health and medical data is not explicitly discussed. Also, there is no real discussion of other technologies that integrate with GIS, such as remote sensing, and the potential use of these in the study of public health. Further, air pollution is emphasized, but there are other pathways of exposure not presented, such as groundwater contamination, that can also be modeled in a GIS environment or in conjunction with a GIS.

The sophisticated treatment of the integration of GIS with epidemiological methods assumes a basic knowledge of study design, spatial statistics, and GIS. Consequently, this book would be extremely useful for epidemiologists, geographers, or environmental scientists interested in understanding the potential uses of GIS in this field, but may be less useful for a practitioner or someone new to this field of research. Even with the emphasis on environmental epidemiology, the collection of essays underscores that now is the time for researchers in public health to dictate the direction of the development of new GIS tools that explicitly meet the needs of rigorous study designs.