Appropriate Selection of Cartographic Symbols in a GIS Environment

Appropriate Selection of Cartographic Symbols in a GIS Environment Steve Ramroop Department of Information Science, University of Otago, Dunedin, New Zealand. Tel: +64 3 479 5608 Fax: +64 3 479 8311, sramroop@infoscience.otago.ac.nz Presented at the 10th Colloquium of the Spatial Information Research Centre, University of Otago, New Zealand, 16-19 November, 1998 Abstract Geographic Information System (GIS) is becoming prevalent in most professions forming a niche across business sectors. Consequently, users are becoming producers of maps. Map making was traditionally the work of Cartographers who are trained in the field of map layout and design. Their work is focussed around the presentation of spatial information such that the chosen data set is effectively emphasized and transferred onto the final representation. In this paper the tools available to the Cartographer is identified and a symbol selection methodology is presented. The methodology identifies distinct steps to be followed in deciding upon the most appropriate symbol to be used with the view that the spatial component of the overall information must be effectively presented. Wood s (1972) Cartographic Communication Model is used as the model which depicts the intention of Cartographers in transferring reality into maps. The symbol selection methodology is presented by first categorizing data into information types. Information types are then related to user perception properties which are linked to the symbol choice. These symbols are called visual variables. They are the tools available to the Cartographer. Keywords and phrases: users; producers; visual variables; symbol selection methodology. 1 Introduction Geographic Information System (GIS) is a fast growing technology and tool being used by many professionals and non-professionals. GIS is used in almost all business sectors where decisions are made which can range from real-time decisions to long term business operations. In this regard, the best solution is sought by many users who are involved in decision making processes. Apart from the normal textual and attribute information which is normally associated with assisting the decision making processes, GIS presents the spatial component in viewing information. Hence, GIS analyst should ensure that the benefits obtained from displaying information spatially is spontaneously realized by end users who are making use of GIS results. The author of this paper emphasizes the need to apply the principles of symbol selection in Cartography to GIS results. This is important because with the advent of desktop mapping software being made available to anyone who seeks it and the symbol libraries inherent in the software s functionality, have further compounded the problem of selecting appropriate symbols. For example, almost all Desktop GIS has an output module whereby users have the ability to display results using symbols available from the system s library. Selections are usually based upon the experience and preferences of the person preparing the map. As such, the ease in which colours, symbols and text fonts applied to maps becomes subjective. Since users are spending 80 percent of the cost in their overall GIS implementation to capture data Proceedings of the Spatial Information Research Centre s 10th Colloquium 271

(Aronoff 1993) then full benefit should be made of final GIS results. In almost all GIS applications the final result is a map, which can be a hardcopy or softcopy product. In such instances results must convey to the user, the information in a manner that they can readily interpret and as much as possible, relate to reality. The final display of results can influence the effectiveness and use of the results. In this regard, there is the need for user awareness of the principles of Cartography in choosing symbols. The intention in presenting this paper is to present the essential tools used by the Cartographer and to describe a methodology in selecting appropriate symbols. Based upon the choice of the symbol selection, much more information can be inferred from the displayed result. For example, a population density map would not only show the density but also display the spatial patterns such as the distribution and areas of high concentration. These are just some of the many additional information obtained from the maps. Cassettari (1993) commented that from the Cartographer s perspective, visualization involves the visual processes that allow users to identify patterns and to create, and manipulate mental images. In this regard, since the making of maps is becoming easy through the use of various software, the onus is now on the GIS user to learn and apply the concepts and tools used by the Cartographer. 2 Cartographic Communication Model Elzakker (1993) stated that Cartography is defined by the United Nations as: The whole science of map making. This involves the design and production of maps used in Geodesy, Photogrammetry, Remote Sensing, and so on. However, for the purpose of this paper symbol selection is presented which is a small but significant component in the design of maps. Cassettari (1993) indicated that there is an element in the creation of a map by exercising both personal judgements and preference in the way information is portrayed. This is addressed by choosing the best available symbols which is capable of enhancing the spatial component of information. To best understand the work of a Cartographer, Wood s (1972) model is of relevance. His model describes the process followed by Cartographers in transferring reality to a map by using the process of interpretation as perceived by the Cartographer. From Wood s (1972) model shown on Figure: 1, the Cartographer s Reality must intersect the Map User s Reality. The amount of intersection between the two is a measure of the effectiveness of the prepared map. The model depicts the idea for a map which is perceived by the Cartographer as being asnapshot of reality which must as close as possible Reality Cartographer's Reality Map User's Reality External Factors Idea for map New information and/or new perepestive on existing knowledge External Factors Map Cartographer's Perception Cartographer Select Symbol Map User Detection User's Perception Human Factors Human Factors Figure: 1. Cartographic Communication Model (Wood 1972) 272 Proceedings of the Spatial Information Research Centre s 10th Colloquium

approximate to users perception. Therefore, a balance is sought in terms of the user s perception and Cartographer s perception before a map is created. The final results are Maps which are detected by the map reader and his user perception should impart a new perspective on the existing knowledge presented on the map. The process of symbol selection is not fixed because the Cartographer s perception of reality is influenced by external and human factors (such as experience, user specifications, interpretation, and so on). However, there is the need for control. As such, the Cartographer considers the following in his process of selecting symbols and preparing maps: the intended use of the map; the map user requirements; available map reproduction method(s); the potential noise in communicating the information to the user (for example, symbol choice); and the map scale. The design of a map has to take into account the problems of visual search, recognition of conventional symbols, the role of memory in interpretation, and the user s own reality in comparison to that of the Cartographer s perception of reality. Details of these are out of the scope of this paper. The Cartographer s translation of real world objects into map entities can be viewed and compared to language translation (for example translating from English to Spanish). Brown A. & Stefanovic (1989) stated: Map and symbol design is the intellectual process of the design of a harmonious set of symbols that properly portray the type, character and location of the single elements in the map. At the same time the map portrays the total subject matter of the map as well as a well balanced entity with an aim of an optional perception process at the side of the map user. As much as possible the Cartographer s perception of reality must match the user s perception. Hence its the role of the Cartographer to ensure such a balance is achieved. To achieve such a goal the Cartographer makes use of some basic tools common to all map makers. 3 Cartographic Primitives Spatially referenced data existing in reality is transformed into cartographic features used repeatedly to represent the unlimited phenomena existing in reality. Such features are termed cartographic primitives which are: Point features (for example, fire hydrants, electricity poles, and so on); Line features (for example, roads, rivers, and so on); and Area features (for example, urban regions, parcel layouts, and so on). Depending upon the mapping scale, the various phenomena can be represented using any one of the primitives or a combination of them. For example, at a scale of 1:1000 the boundaries of a town is represented as an area feature while at a smaller scale (for example 1:1,000,000) the same town will be represented as a point. 4 Visual Variables - The Cartographer s Toolbox Elzakker (1993) indicated that communication by language, words are used to present information to the recipient. Each word is compose of a number of letters (or signs) that are the building stones of thelanguage. Similarly using graphics, Cartographers have seven types of variations perceivable to our eyes which are used as the primary building stones to graphically representing information. Also, GISSTT (1995) indicated that text in association with graphic symbols are effectively combined on a map in order to communicate spatial information. In Cartography, the seven symbol variations are called visual variables which are used in the construction of symbols. The description of each of these visual variables are discussed in the following subsections. Proceedings of the Spatial Information Research Centre s 10th Colloquium 273

2345678234567890 2345678234567890 2345678234567890 2345678234567890 2345678234567890 2345678234567890 2345678234567890 234567892345678901234567890 234567892345678901234567890 234567892345678901234567890 234567892345678901234567890 234567892345678901234567890 234567892345678901234567890 234567890345678901234567890 45678901234567890 234567890125678901234567890 234567890123678901234567890 234567890123478901234567890 234567890123458901234567890 234567890123458901234567890 234567890123458901234567890 234567890123458901234567890 234567890123458901234567890 234567890123458901234567890 234567890123456901234567890 234567890123456901234567890 234567890123456901234567890 234567890123456901234567890 4.1 Position POSITION refers to the x, y, (and z) location of the information being mapped which determines the phenomena s place on the map. All symbols used on a map makes use of this visual variable, therefore, POSITION always has to be used in combination with one or more of the other visual variables. POSITION visual variable is applicable to point, line and area primitives. Examples are shown in Figure: 2. e.g. Towns e.g. Roads e.g. Land use Figure: 2. Visual Variable POSITION 4.2 Form FORM refers to symbols which differ only in shape. FORM differences are easy to draw and the variations are almost unlimited. FORM is applicable to point, line, and area symbols, however, with respect to line and area symbols FORM refers to the individual elements with which the symbol is constructed and not the overall form of the line or area feature, (Elzakker 1993). Examples are shown in Figure: 3. 78902345678901 5678901345678901 5678901245678901 45678901245678901 235678901 235678901 235678901 235678901 Figure: 3. Visual Variable FORM 4.3 Orientation ORIENTATION refers to the direction in which symbols are placed. Depending upon the individual elements used, ORIENTATION has its limitation in terms of the number of angles each element can berotated. Brown A. & Stefanovic (1989) indicated that there are six angular variations possible for the variable ORIENTATION which are: 00,300,600,900,1200, and 1500. Examples are shown in Figure: 4. Figure: 4. Visual Variable ORIENTATION 4.4 Colour Brown A. & Stefanovic (1989) indicated that COLOUR is perhaps the most powerful and most frequently used visual variable in symbol design. COLOUR is applicable to the three primitives. Examples are shown in Figure 5. (red) (green) (green) (red) (blue) (green) (brown) (red) (yellow) Figure: 5. Visual Variable COLOUR 274 Proceedings of the Spatial Information Research Centre s 10th Colloquium

4.5 Texture TEXTURE refers to the variation in density of the graphic elements forming the overall symbol, Elzakker (1993). TEXTURE is applied to the three primitives but it is less effective for point and line primitives unless they are exaggerated. Examples are shown in Figure: 6.......................................................................... Figure: 6. Visual Variable TEXTURE 4.6 Value Refers to values on the grey scale ranging from white to black. Brown A. & Stefanovic (1989) commented that VALUE is measured in terms of the ability to reflect light. VALUE can also be applied to COLOUR visual variable. VALUE is applicable to the three primitives but predominantly used to represent area primitives. Examples are shown in Figure: 7. Figure: 7. Visual Variable VALUE 4.7 Size SIZE refers to the dimensions of the symbols or in the case of area symbols, to the dimensions of the individual elements with which the symbol is built up. SIZE is applicable to all three cartographic primitives however, commonly used for line and point primitives. Examples are shown in Figure: 8. e.g. Towns e.g. Roads e.g. Land use Figure: 8. Visual Variable SIZE 5 Data For GIS Data sets are the essential ingredient for the operation of any GIS. Data sets form the link between the computer model and reality by measuring the phenomena under consideration and storing data in databases. Since data sets are the primitive source upon which decisions are made, there are some intermediate processing of the raw data before the final result is presented. This result is in fact the information used to make better decisions. In this regard, there is a distinction between data and information. 5.1 Data and Information Data is defined as: value known or assumed as the basis for inference, quantities or characters operated on by computers. Information on the other hand, is defined as: knowledge, telling, what is told, news. When comparing data and information it is apparent that data is needed to generate information. Therefore, data can be referred to that which is raw or unprocessed while information is that which has undergone some processing, such as classification, Proceedings of the Spatial Information Research Centre s 10th Colloquium 275

which makes information more relevant to the problem. Cassettari (1993) stated that, data sets are the original survey, the original remote sensing image, the census statistics, and so on. 6 Classification Of Data One of the Cartographer s role is to ensure that the phenomena being mapped is shown effectively such that users can infer much more information apart from the phenomena s location. Other information conveyed to the user are: The ability to replicate phenomena in reality. For example, Cartographers use the colour green to represent vegetation because green replicates the general colour of vegetation. This analogy is important because although a map user can constantly refer to the map legend for confirmation and clarification of the symbol used, continuous reference to the map scale is not ideal. The spatial distribution of phenomena. For example, planners can visually locate the distribution of land for urban development from a land use map. The spatial movement of phenomena. For example, in an urban area map created from temporal overlays, indicates the direction of growth of the urban area. To convey such information in the final GIS result, it is important to select the cartographic symbols which would convey such information to the user. In that way users will feel comfortable with the map presentation through its ease of use and its close approximation to reality. GISSTT (1995) commented that it should be possible to infer that some objects are distinct and yet others are similar but different, simply from inspecting map symbols. Apart from the three cartographic primitives mentioned in Section 3, there are other characteristics of geographic data which demand more than just a simple transformation into map symbols. The classification of data is listed under two general categories: Qualitative data : refers to data sets which are purely descriptive and textual. For example, on a land use map there are residential areas, agricultural areas, commercial areas and so on. Quantitative data : refers to data sets with specific numerical values. For example, a population density map. Quantitative data can be further sub-divided into: Absolute quantitative data : refers to numerically observed, measured or counted quantities. For example number of persons, number of houses, and so on. Relative quantitative data : refers to numerically calculated, or derived quantities. For example density, averages, and so on. 7 Classification Of Information To classify information the categories are similar to the Classification of Data however, with additional categories. Information is classified by Elzakker (1993) as follows: Qualitative Information Nominal qualitative - refers to textual, descriptive information about the phenomena being mapped. For example, arable land built-up area forested areas. Ordinal qualitative - refers to textual, descriptive information about the phenomena with a clear sense of order, which is not quantitatively determined. For example, primary roads! secondary roads! tertiary roads! tracks. Quantitative Information Interval quantitative - refers to the numerical ranking of information with the interval between data sets quantitatively determined, using an arbitrary zero. For example temperature, year of establishment, altitude, and so on. 276 Proceedings of the Spatial Information Research Centre s 10th Colloquium

Ratio quantitative - refers to the numerical ranking of information using an absolute zero. For example number of employees, production of factories, and so on. The other two categories refers to single numerical values which can be categorized under: Absolute quantitative - refers to the observed, measured or counted quantities. For example house hold counts. Relative quantitative - refers to the calculated, derived quantities. For example area calculations from field surveys. 8 Perception Properties As mentioned in Section 2, it is important to accept that users share common reactions when using symbols on a map. Hence the Cartographer s perception must be consistent with the user s perception. Therefore, some generalizations is made with respect to the perception properties of visual variables. Elzakker (1993) commented, experience has demonstrated that visual variables may have one or more perception properties which is briefly described in the following subsections. 8.1 Associative Perception Visual variables have Associative Perception Properties if spontaneously all symbols differentiated by the variable are seen as of equal importance. There is a INFORMATION represented by a Visual Variable with PERCEPTION PROPERTY Nominal Ordinal Interval Ratio first impression of uniformity where no symbol seems to be more important than another. For example the colour green is used to represent vegetation and not red when shown on a landscape map. 8.2 Selective Perception Visual variables have Selective Perception Properties if spontaneously all symbols differentiated by a given variable can be arranged visually in distinct groups. Hence spatial patterns are readily recognized by the map reader. For example, distribution of built-up areas on a land-use map are inferred by the user. 8.3 Ordered Perception Visual variables have Ordered Perception Properties if spontaneously all symbols differentiated by a given variable can be placed in an unambiguous order. For example, the classification of roads into first class, second class and third class roads. 8.4 Quantitative Perception Visual variables have Quantitative Perception Properties if spontaneously a distinct quantity may be associated to each variable differentiated by that variable. Therefore, by visual interpretation map users can infer that one phenomenon is roughly bigger or smaller than another. For example, the comparison of population densities between towns. The link between Information Type and Perception Property is presented by Elzakker (1993) which is shown on Figure: 9. Each Information Type is associated to a defined Perception Property. Having decided upon the perception properties, research done by Elzakker (1993) presented a link between Perception Property and the recommended visual variables as shown on Figure: 10. The result indicated that each Perception Property is assigned a recommendation on a scale of: Associative ( + - Selective ) Ordered Ordered Quantitative Absolute quantitative Relative quantitative Figure: 9. Information Perception Property (Elzakker 1993) Quantitative Ordered Very Good Good Moderate Bad. Based upon these recommendations the choice of symbols are made however, this depends upon the map reproduction facility available, production cost, and Proceedings of the Spatial Information Research Centre s 10th Colloquium 277

Perception Property Associative Selective Ordered Visual Variables Quantitative Recomendations: Position Form Orientation Colour Texture Value Size : Very Good : Good 23 23 23 23 23 23 23 23 23 23 23 123456 123456 123456 123456 the needs of the user. For example, Information with an Associative Perception Property has four recommended visual variable which are rated as being Good. The situation may exist where the user may not have a colour printer, therefore in such instances POSITION, FORM, and ORIENTATION would be the final recommended visual variables. 9 Symbol Selection Methodology In summary, the method of selecting symbols to map information is one which follows a systematic approach in categorizing data and ensuring that the user s and Cartographer s perception corresponds. The method of selecting the most appropriate symbol(s) is as follows: 1. Determine the nature of the information. 2. Identify the perception property to be conveyed. 3. Identify the recommended visual variables. : Moderate : Bad Figure 10. Recommended Visual Variables (Elzakker 1993) 4. Depending upon the map reproduction facilities available (such as plotters, printers, and so on), appropriate visual variables are selected to map the phenomena under consideration. 10 Summary The author of this paper attempted to present the importance of selecting appropriate symbols using the tools available to the Cartographer to represent phenomena in reality. One such methodology is presented. Data is transferred into information types, which has a known perception property. Each perception property has recommended visual variables which are used as a guide in choosing appropriate symbols. External factors such as printing facilities influences the final choice of symbols. Using today s computer technology, users are bombarded with a mixture of fonts, colours and arbitrary symbol libraries which if not used systematically can create maps which can inefficiently and perhaps inaccurately communicate spatial information. However, the intention of the author in presenting this paper is to sensitize and draw the awareness of a methodology in selecting appropriate symbols by introducing and using the Cartographer s toolbox. The importance in selecting appropriate symbols were identified as users with multiple professional back grounds makes use of the symbol libraries to display their GIS results. References Aronoff, S. (1993), Geographic Information Systems : A Management Perspective., WDL Publications, Ottawa,Canada. pp: 103-132. Brown A., J. Drummond, E. C. & Stefanovic, P. (1989), Map design fundamentals. International Institute for Aerospace Survey and Earth Science (ITC). Internal Lecture Notes, pp: 1-40. Cassettari, S. (1993), Introduction to Integrated Geo- Information Management, Chapman and Hall. Elzakker, C. (1993), Cartographic visualization. International Institute for Aerospace Survey and Earth Science (ITC). Internal Lecture Notes, pp: 1-60. GISSTT (1995), GIS in national development: Status and potentials, in C. Rogers, ed., Proceedings of the First Technical Conference of Geographic Information Systems Society of Trinidad and Tobago (GISSTT), Trinidad, pp. 56-61. Wood, M. (1972), Human factors in cartographic communication, Cartographic Journal 9(2), pp: 123-132. 278 Proceedings of the Spatial Information Research Centre s 10th Colloquium