point symbols—varying symbol size, shape, and orientationTHE SUBTLEST DETAILS CAN DETERMINE HOW MAP DATA is read and interpreted. The shape of a point marker, the width of a line, the arrangement of a pattern—each conveys specific information. Map designers pay careful attention to the match between visual variables (such as shape and size) and the characteristics of their data when creating customized symbols.
There are two basic groups of visual variables that structure symbol design. Size, lightness, pattern spacing, and saturation are used when symbolizing ordered data. Hue, shape, orientation, and pattern arrangement categorize map features into qualitative classes.
The combination of visual variables and feature types (point, line, and area) forms a general framework that you can use to design customized symbols for representing all kinds of map data. Using multiple visual variables together within symbols, such as hue with shape, produces an almost endless variety of map symbols. There are few excuses for using a default symbol that does not match well with the data it represents because you can make almost any map symbol you can envision.
In addition to using color symbols, map designers work with the following:
point symbols—varying symbol size, shape, and orientation
line and area symbols—varying line size and pattern and by constructing area patterns
visual variables for point, line, and area symbols that represent ordered data and qualitative data
multivariate mapping using overlays and combinations of visual variables in symbol construction
many pairings of eight visual variables for bivariate mapping
Figure 9.1 The visual variable size is used to represent data values for areas. Water use by parcel is mapped with larger symbols representing parcels with higher water use.
Figure 9.2 The visual variable size is used to represent data values at point locations. Here, hypothetical illumination amounts from individual street lamps are mapped with larger symbols for greater illumination.
Data source: Esri ArcGIS Sample Maps data.
Point symbols are not limited to representing data linked to point features on a map. Whether a feature is a point or an area depends on the map scale. A city may be a point at a small scale and an area at a larger scale. Likewise, a street address may be a point on a map of a city but an area on a larger-scale parcel map. Thus, you want to think of “point symbols” as flexibly applying to both point and area locations.
In addition to hue and lightness (chapter 8), the main visual variables used to create point symbols are size, shape, and orientation (angle). Pictograms used as point symbols are a special case of shape differences.
Symbol size is used to represent data values at point locations or for areas. Larger symbols represent higher data values. The simple examples that follow show water use and illumination strength provided by street lighting. Symbols are centered in parcel areas to represent water use for whole parcels in figure 9.1. In figure 9.2, they are located at lamppost point locations.
Point symbols that vary in size may be used to represent either total amounts or derived values. For example, sizes may represent the number of deaths or the death rate from a disease. It is more common to use point symbols for totals (figure 9.3); choropleth maps are the main method of representing rates and other derived values (figure 9.4).
When you use point symbols to represent area data, they are usually located in the middle of each map polygon. The symbol may be smaller or larger than the polygon area, depending on the data value it represents. In contrast, when color alone is used, the area of the symbol on the map is determined by the size of the map polygon (figure 9.4). High data values for small polygons, such as high-population urban areas, may have a much reduced visual impact on a choropleth map, and the data may be represented better using symbol size. Bivariate mapping can assist with this problem.
Figure 9.3 Total number of deaths among Black females in the lower Mississippi River region by state economic area (SEA). This data is represented with proportioned point symbols. Symbol areas are proportioned to data values.
Figure 9.4 Age-adjusted mortality rates for Black females mapped for comparison to figure 9.3 using lightness on a choropleth map. Classing used in the choropleth map accentuates spatial pattern in this data.
Data source: US National Institutes of Health/National Cancer Institute (NIH/NCI), http://ratecalc.cancer.gov. Maps updated by P. Limpisathian, Department of Geography, The Pennsylvania State University (Penn State Geography).
Mortality data is shown in figure 9.5 with point and area symbols. The point symbols are filled with the same colors used for the area symbols to emphasize the link between the two representations.
Both proportioned and graduated symbols use differences in size to represent data. Proportioning symbol sizes to map data gives you a fairly exact understanding of differences in magnitude between data values. A city with twice the population of another will have a proportioned symbol twice as tall or with twice the area. If the symbols are graduated, the symbol for the larger city will simply be a bit larger, but it will not attempt to show how much larger the city is. Graduating (or range grading) symbols is a less exacting method that ranks values rather than representing data amounts. Compare the marked difference in the size range between the symbols representing ten and ten thousand deaths in figures 9.3 and figure 9.5A. Proportioned symbols represent data magnitudes and graduated symbols use size to represent data order.
Figure 9.5 Total number of cancer deaths in Black women by SEA mapped with graduated point symbols (A) and area symbols (B). The same graduated color set is used to emphasize the identical classification of the data in these two representations. Data source: NIH/NCI, http://ratecalc.cancer.gov. Maps updated by P. Limpisathian, Penn State Geography.
The examples shown previously in this section have used circle areas to represent data. Lengths of linear point symbols and areas of shapes other than circles are also used for constructing map symbols.
Shape is another visual variable readily used with point symbols. Shapes may be simple geometric forms (such as squares, circles, and triangles) or more complex. Shape is used to represent qualitative differences in data values. A simple map uses a star for a local information office, a circle for the post office, and a triangle for the library (figure 9.6).
Symbol shapes can also be used to build intricate codes that vary in form to represent qualitative data. Pictograms (or icons or glyphs) make more elaborate use of shape for data representation. Common sets of pictograms are offered in the styles and the fonts that are installed with ArcGIS. The same simple map is shown with familiar and readily distinguished pictograms (figure 9.7).
Figure 9.6 The visual variable shape is used to represent categories of features at point locations. Local services are represented with three geometric shapes.
Figure 9.7 Pictograms use shape to distinguish categories. This small map shows locations for an information office, post office, and library.
Data source: Esri ArcGIS Sample Maps data.
There are few guidelines for designing or selecting shapes for map symbols. If there are symbol conventions for a map topic, such as picnicking and camping pictograms (figure 9.8 and 9.9), it makes sense to use them. Do not expect a reader, however, to be clairvoyant in their ability to interpret your pictograms. Using many small and intricate diagrams all over your map will stymie even the most diligent legend reader.
Figure 9.8 This portion of the Big Bend National Park map uses a standardized set of pictograms. Notice, also, the sparing use of simply designed leader lines for symbols in congested areas of the map.
You can use multiple visual variables together. You may find that using hue with shape is a good way to distinguish qualitative differences. You can also vary size a bit to ensure symbol differences, though you want to avoid an implied order to the symbols. More compact shapes are easier to associate with particular point locations than tall or wide shapes. You want symbols to be readily identifiable and to look markedly different from each other so readers can identify them without having to look closely at each symbol when they scan the map.
Remember that sometimes the function of symbols is to be seen as groups, not to be read one by one. If all of your symbols are small blue boxes distinguished by different tiny marks within the boxes, none of them are able to stand out from the others. Your readers will not be able to look across the map to see where particular symbols cluster. They will also have a hard time finding the features they seek in the crowd of similar symbols.
Figure 9.9 A portion of the map legend for the National Park Service map that identifies the pictograms.
Source: National Park Service (NPS).
It is a good idea to test your symbol shapes and pictograms with map readers unfamiliar with the project to check that your map is easy to understand.
Symbol angle, or orientation, can be used to vary a point symbol of constant size and shape. A simple use would be to orient a small rectangular symbol vertically, horizontally, and at various angles. The first example in figure 9.10 shows a simple arrow shape at three angles to represent water use. You can see that this visual variable is not particularly effective for this data compared to the same data shown with different symbol sizes in the second example in figure 9.10.
The varying orientation of parcels in figure 9.10A complicates map interpretation. Is symbol orientation set relative to the parcel or to the map frame? Angles are set relative to the frame, but a reader may puzzle over the question, reducing the effectiveness of your map.
Orientation can also be used to vary the internal marks within a symbol while maintaining a constant shape and size. Figure 9.11A shows different orientations for the two halves within a circle symbol representing qualitative point data. You have also seen this same data represented with different shapes, and the shapes provide more noticeable contrast between features (figure 9.11B).
Figure 9.10 A simple arrow shape at different angles to represent parcel water use (A). The visual variable orientation is demonstrated in this map. The same data is represented more effectively with symbol size (B).
Figure 9.11 Orientation can be set to vary marks within symbols (A), though this may not be as effective as using varied shapes (B).
Data source: Esri ArcGIS Sample Maps data.
Figure 9.12 The cyan symbols on this example weather map show wind direction using symbol orientation. The “station model” symbols also use many other visual variables, including shape, hue, and arrangement, to report additional weather information. Source: National Weather Service, National Oceanic and Atmospheric Administration (NOAA), http://www.wpc.ncep.noaa.gov/archives/web_pages/sfc/sfc_archive_maps.php?arcdate=11/26/2014&selmap=2014112618&maptype=namussfc.
It is debatable whether the visual variable orientation is better suited to qualitative or quantitative data. Three symbol angles could represent different categories of features, though some order seems to be implied by this progression. Orientation does not exactly suggest low to high either. Symbols that vary orientation can be useful for representing some types of ordered data, such as cyclical and directional phenomena. Weather symbols, for example, make thorough use of angles for wind direction (note the cyan symbols on figure 9.12). Temporal data, such as time of day, is also an obvious application for orientation of marks within a point symbol.
In addition to hue and lightness, the following are visual variables that may be applied to line and area symbols:
size
spacing
shape
arrangement
orientation
Figure 9.13 The visual variable size is used for line features. Traffic flow is well represented with a progression of line widths. Data source: Esri ArcGIS Sample Maps data.
Size is used to construct both proportioned and graduated line symbols. Lines that are dashed and cased combine the visual variables of spacing, shape, arrangement, and orientation. Line and area symbols use size and spacing to reflect ordered differences. Shape, arrangement, and orientation of both line and area patterns are used to represent qualitative differences.
Size is used to represent data values associated with line features by adjusting line widths. Figure 9.13 represents three levels of traffic flow on neighborhood roads.
Just as with point symbol sizes, line sizes may be proportioned or graduated. Proportioned lines are shown at line widths that represent differences in data values. A road that carries ten times more traffic would be shown ten times wider. Graduated line widths rank order lines from low to high, but the line widths are not directly proportioned to data values. Proportioned and graduated lines represent data using the visual variable size.
Proportioned or graduated lines may follow precise routes on a map, such as a road or pipeline (figure 9.14). Or they may be more abstract direction indicators, such as generalized flow lines showing the overall directions of migrations (figure 9.15, next page).
Figure 9.14 Line size is graduated to represent the number of road lanes. Data source: Prince George’s County, Maryland.
Figure 9.15 Line size is proportioned to represent net migration between California and other states, with hue representing two time periods. Map title is Net Migration Between California and Other States: 1955–1960 [red] and 1995–2000 [blue]. Source: Suchan, T., et al. 2007. Census Atlas of the United States. Washington, DC: US Census Bureau, p. 111.
There are multiple line patterns available for map design. The primary options for line symbols in map design are dashing and casing.
Dashes add pattern to a line, and differences in spacing and length of the dash pattern are used to create different symbols. Wider separation between dashes (figure 9.16) and longer dashes create coarse textures.
Figure 9.16 An example use of spacing as a visual variable for dashing lines.
Data source: Esri ArcGIS Sample Maps data.
As dash patterns become more complex, they also make use of shape and arrangement. An international boundary line that combines short and long dashes is a combination of spacing, arrangement, and shape characteristics. Figure 9.17 shows the same shape with different arrangements. Lines can also be built from differently shaped marks, such as a string of dots that contrasts with a string of crosses, as shown in figure 9.18.
Figure 9.17 Arrangement of marks along a line symbolizes different road types.
Figure 9.18 A wide variety of shapes may be used to define line symbols. Circles and crosses distinguish two road types in this example map.
Figure 9.19 The visual variable orientation is used to vary line patterns in this example map. Data source: Esri ArcGIS Sample Maps data.
Figure 9.20 Example cased lines that have varied overall widths, varied casing widths, and multiple lanes. Data source: Prince George’s County, Maryland. Map updated by P. Limpisathian, Penn State Geography.
The orientation of the marks within a line symbol can be used to indicate line category. Hatching is created when a pattern is at an angle to the line. In figure 9.19, some hatches are set perpendicular to the line direction and some are at an angle. The cased alleys are included in the example because you can interpret line casing as an extreme angle setting (although you do not create line casings the same way you create hatching). A casing pattern runs parallel with the line direction.
Casing is a commonly used and versatile cartographic symbol option for lines. Casing functions just like halos do for text—it helps increase line visibility over multiple backgrounds. It can also be used to create a wide symbol that contrasts with its background but is not overly bold (that is, it does not overwhelm the visual hierarchy among the map symbols).
Casing is formed by lines on both sides of the symbol and may be thin or thick. If the two lines that build a line symbol are similar in width, then the casing is thin. The example portion of a road map (figure 9.20) shows a variety of cased lines.
The legend is labeled with the casing characteristics (rather than numbers of road lanes, the mapped data). The 3-point gray line nested in the 4-point dark red line produces a 0.5-point casing on both sides of the line (splitting the 1-point difference between 3- and 4-point lines). The same 3-point line nested in an 8-point black line produces a thicker 2.5-point casing.
More than two layers can be used to build cased lines. The widest line on the map has three lines layered to create lanes, as shown in segments of the ArcGIS symbol property editor (figure 9.21).
In figure 9.20, notice that some casings interrupt intersecting lines and others do not. You can control the interaction between cased lines in GIS (in ArcGIS, this design detail is set using join and merge toggles or advanced symbol level drawing). In design software, layering and order settings within layers help to control the interaction of lines and casings.
A cased line will indicate a different kind of feature than uncased or dashed lines. For example, a cased highway will be distinguished from a dashed county line. You can even case a dashed line (as shown in figure 9.22 for a proposed highway off-ramp). Combinations of casing, dashing, and width, along with other visual variables, will establish a hierarchy of lines.
Wide lines can carry a pattern that involves the whole suite of visual variables that may be applied to lines: spacing, shape, arrangement, and orientation. Add color variables, and you can produce an almost endless variety of line patterns. But that does not mean that you should use intricate line symbols in their fullest complexity; remember, do not stump your reader.
Figure 9.21 The layers that build three example cased lines from figure 9.20: 3 on a 4-point line, 3 on 8 point, and 1 on 7 on 10 point. Note the symbol preview above each set of layers. Figure updated by P. Limpisathian, Penn State Geography.
Figure 9.22 A line that is dashed and cased is an example of using multiple visual variables for line pattern. Data source: Prince George’s County, Maryland. Map updated by P. Limpisathian, Penn State Geography.
A wide array of area patterns may be used on maps. They can be quite literal in their design, such as the repeated pattern of small tufts of grass and reeds used to represent a swamp (figure 9.23). Or they can be completely abstract, such as a crosshatch of evenly spaced lines across an area.
Choose area patterns by paying attention to the visual variables used to build them. You want to choose and customize area patterns so they obviously represent the logical relationships within the data, just as you do with color symbols.
Use textures that are coarse and fine to represent hierarchy (figure 9.24A). Loosely spaced patterns represent low values and closely spaced patterns represent high values. Similarly, a fill of larger shapes will indicate a higher data value than a fill of small shapes with the same spacing (figure 9.24B).
Figure 9.23 A selection of pattern styles made available in ArcGIS. Source: Esri ArcGIS software interface.
Use the shapes of elements within a pattern to represent qualitative differences within the data (figure 9.25A). A fill of small circles will indicate a different kind of feature than fills of crosses or stars. More literally, a fill of small tree drawings may contrast with a pavement pattern to distinguish park from parking. Orientation (figure 9.25B) and arrangement (figure 9.26) can also be used within area patterns to indicate qualitative differences.
As with points and lines, visual variables can be used together to increase the contrast between area fills. Alter lightness to enhance hierarchy and use hue to enhance qualitative differences.
Figure 9.24 The visual variables of spacing (A) and size (B) are applied to area symbols to establish hierarchy.
Figure 9.25 The visual variables of shape (A) and orientation (B) are applied to area symbols to establish categories.
Data source: Esri ArcGIS Sample Maps data.
Figure 9.26 ArcGIS offers two arrangement settings (grid and random) that can be useful for representing qualitative differences between areas.
Can you envision patterned point symbols that vary in spacing? Do you recall the look of line symbols that vary in arrangement? With eight visual variables and three types of features, you have twenty-four basic ways to vary symbols for representing map data.
Because there are many combinations, this chapter includes a pair of tables that provide you with a summary view of point, line, and area features represented with symbols that vary in color, size, shape, and pattern. The visual variables organized in these tables are the following:
size
lightness
spacing
saturation
hue
shape
orientation
arrangement
Size, lightness, spacing, and saturation are the visual variables well suited to representing ordered data—either rank-ordered data or numerical amounts (table 9.1). These visual variables establish hierarchies among features. Additional visual variables used for symbolizing quantitative data are perspective height, transparency, and crispness (focus).
Table 9.1
Table 9.1 These twelve examples show each visual variable (rows) suited for representing ordered data applied to point, line, and area features (columns) with simple demonstration maps. Map topics are listed below each map segment. (All of the symbols on these map segments were created in ArcGIS.)
Table 9.2 These twelve examples show each visual variable suited for representing qualitative data applied to point, line, and area features with simple demonstration maps. Map topics are listed below each map segment. (All of the symbols on these map segments were created in ArcGIS.)
Hue, shape, orientation, and arrangement are the visual variables well suited to categorizing features (table 9.2). They represent qualitative differences that are not ordered. Symbol angle may also be useful for representing some types of quantitative data, such as direction or time.
Maps excel at showing relationships between data distributions. You can symbolize multiple variables by combining the perceptual dimensions of color—hue, lightness, saturation—in arrangements such as sequential-sequential, diverging-diverging, and sequential-qualitative (chapter 8). A fast look at relationships between data variables can be made by setting one transparent over the other, also described in chapter 8. This section introduces a more complete perspective on multivariate mapping. Visual variables may be overlaid or combined into bivariate symbols. Bivariate symbols each vary by two data variables using two or more visual variables. In chapter 8 you see color symbols that vary by hue and lightness, for example; in this chapter, the additional visual variables are added to this approach.
In addition to overlaying one partly transparent color scheme on another, pattern symbols are well suited to overlay on color symbols. For example, figure 9.27 shows a successful combination of thin lines over mortality data to signal that enumeration units have sparse data and thus death rates are less reliable calculations. Each line in the pattern is a thin white and thin black line offset slightly from each other to build a pair. The white-black contrast ensures a visible texture regardless of background. The white line also contrasts with dark backgrounds and the dark line contrasts with light backgrounds. This pattern provides an overlay that is visually separable from the diverging color scheme.
Figure 9.27 Multivariate map with black-white line pattern overlay, alerting to sparse data, on a diverging scheme showing age-adjusted death rates by HSA (hospital service area), 1988-1992, from pneumonia and influenza in Black females. The legend for this map is arranged well and provides extensive annotation for map interpretation. Source: US Centers for Disease Control/National Center for Health Statistics (CDC/NCHS). 1996. Atlas of United States Mortality. http://www.cdc.gov/nchs/data/gis/atmappne.pdf, p. 110.
A second overlay of pattern shows a diverging pattern of dots that contrast well with the sequential scheme for the primary data variable (figure 9.28). Both lightness and arrangement vary for this symbol—black dots in a regular grid mark high certainty data for temperature change and white dots with an irregular arrangement show moderate certainty. A white mask is used for areas of lowest certainty in the temperature change data.
Figure 9.28 Projected change in near surface air temperature from 1981-2010 to 2071-2100 with a sequential color scheme for filled isolines. The overlay of white and black dots with regular and irregular arrangement show certainty for high radiative forcing (RCP 8.5). Certainty is based on a signal-to-noise calculation among the predictions of several dozen models. Data source: CMIP5, http://cmip-pcmdi.llnl.gov/cmip5/citation.html; World Climate Research Programme’s Working Group on Coupled Modelling, via http://www.climexp.knmi.nl. Map by D. Retchless, Penn State Geography.
Combinations within symbols, rather than as overlays, among the full set of visual variables are useful for presenting pairs of data variables for bivariate mapping. Figure 9.29 combines size and lightness in each symbol. The map shows slopes of river channels with light lines for low slope and darker lines for steeper slopes. Within the same lines, it adjusts channels with small drainage areas to thin lines and those with larger drainage areas have thicker lines. Thin, dark-colored streams are steep and drain small areas. Rivers with larger catchments have lower slopes and run along valleys (wide and light lines). The hillshade base assists map interpretation. The map combines two quantitative variables using visual variables that are both suitable for ordered data.
Figure 9.29 A bivariate map showing slope and drainage area of river channels in a portion of Pennsylvania ridge and valley terrain. Slope is represented by line lightness and drainage area by line size. Data sources: US Geological Survey, NHDPlus. Map by A. Dennis, Penn State Geography.
The second example map (figure 9.30) uses essentially the same combination of visual variables as figure 9.29 but looks quite different. School district spending per student is represented by size—in this case, graduated circles. Number of students by district is represented by lightness with redundant hue transition through magenta, purple, and blue. These visual variables work together to show relationships between these data variables. Large, dark-blue circles are districts with many students and with high expenditures per student—we see the varied character of urban school districts against a surround of smaller rural districts.
Figure 9.30 A bivariate map showing number of students by size and expenditures by lightness and hue for US school districts. Data sources: US Census Bureau; made with Natural Earth. Map by A. Dennis, Penn State Geography.
The example in figure 9.31 shows population distribution using the US Census Bureau’s American Community Survey data for race. The map uses hue (blue, purple, and red). There is slight redundant variation in lightness among these saturated hues to make them easier to distinguish. The spacing of dots represents population density (one dot for 100 people) distributed across enumeration units. The segregation of large American cities, in this case New York City, is immediately apparent across the map.
Figure 9.31 Bivariate mapping with visual variables of spacing (dot numerousness) and hue to represent population distribution by race. Data source: US Census Bureau. Map by A. Dennis and P. Limpisathian, Penn State Geography.
The full set of possible combinations of eight visual variables is outlined in figure 9.32. Three general groupings within this structure stand out: combinations for bivariate maps that combine two quantitative data variables, combinations for quantitative with qualitative data, and for mapping two qualitative variables. Maps in the previous sections fall into this structure:
lightness/size in cell 2 for the rivers map (figure 9.29);
cell 11 hue/size for the migration map (figure 9.15);
both cells 2 and 11 for the school districts map (figure 9.30); and
hue/spacing in cell 13 for the dot map (figure 9.31).
Bivariate mapping introduced in chapter 8 combines hue, lightness, and saturation in varied ways:
sequential-sequential (figure 8.18) and diverging-diverging (figure 8.17 and 8.20) are quant/quant; and
qualitative/sequential (figure 8.22) is qual/quant.
Figure 9.32 Combinations of two visual variables grouped into three sections: quant/quant, qual/quant, and qual/qual. Two-by-two examples are shown in figures that follow for the darker cells.
More challenging combinations of color schemes are sequential-diverging and qualitative/qualitative.
The upper portion of figure 9.32 (purples) lists the ten quant/quant combinations for size, lightness, spacing, and saturation. These bivariate symbols are each structured as shown in figure 9.33, with examples of area fills for combinations 1, 2, and 9 from figure 9.32: size/size, lightness/size, and saturation/spacing (figure 9.34).
Figure 9.33 General structure for combinations of two quantitative visual variables.
Figure 9.34 Quant/quant examples of area fills showing two-by-two visual variable combinations for cells 1, 2, and 9 in figure 9.32: size/size, lightness/size, and saturation/spacing.
The right-side portion of figure 9.32 (greens) lists the ten qual/qual combinations for hue, shape, orientation, and arrangement. The structure for these combinations is shown in figure 9.35, with examples of area fills for combinations 30, 31, and 36: orientation/hue, orientation/shape, and arrangement/arrangement (figure 9.36).
Figure 9.35 General structure for combinations of two qualitative visual variables.
Figure 9.36 Qual/qual examples of area fills showing two-by-two visual variable combinations for cells 30, 31, and 36 in figure 9.32: orientation/hue, orientation/shape, and arrangement/arrangement.
The third portion of figure 9.32 (oranges at lower left) lists the 16 qual/quant combinations between hue, shape, orientation, arrangement and size, lightness, spacing, saturation. Figure 9.37 shows the structure for these combinations. Examples of area fills for combinations are shown in figure 9.38: hue/spacing (13), shape/size (15), shape/lightness (16), orientation/saturation (22), and arrangement/spacing (25).
Figure 9.37 General structure for combinations of qualitative with quantitative visual variables.
Figure 9.38 Qual/quant examples of area fills showing two-by-two visual variable combinations for cells 13, 15, 16, 22, and 25 in figure 9.32: hue/spacing, shape/size, shape/lightness, orientation/saturation, and arrangement/spacing.
The many combinations for points, lines, and areas would fill out many more pages, but you get the idea from this example set. There are many options for customizing bivariate symbols. Some are used often, such as hue and lightness, and some are kind of crazy looking—they are intended to inspire representations of two data variables on maps.
As you seek to design better maps, you will put the symbolization concepts into practice with the concepts from previous chapters. You will bring together data from multiple sources and credit all of them. You will arrange the elements of the display in a balanced manner, setting up good visual hierarchies. Interesting data will be underpinned by location information from well- designed base content. You will explain what readers are seeing with a judicious combination of titles, legends, annotation, and explanation. You will use color and the other visual variables in ways that match the logic in your data, so information jumps off the page, immediately understood by your readers. You will be designing better maps.
