map text, including differences between graphic map text, dynamic labeling, and annotationLABELS ARE AN INHERENT PART OF A MAP’S SYMBOLOGY. The technical challenge of making labels is complemented by the analytical role they play on a map. Thus, clear labeling helps your audience correctly interpret the mapped data.
Clear labeling follows conventions of placement that vary for point, line, and area features. Although it is faster to set type and placement characteristics for groups of features all at once, there will be times when you need to edit individual labels to improve the clarity of the map. Cartographic conventions for placing labels often conflict when numerous labels appear on a map dense with features. Making sensible tradeoffs between labeling conventions is often a matter of knowing how to follow a convention and knowing how to bend it. Knowing the rules lets you break them less often and more skillfully.
A competent map designer makes better choices for label appearance and placement by understanding the following:
map text, including differences between graphic map text, dynamic labeling, and annotation
labels as symbols, indicating feature location, category, and hierarchy, and using ambiguity and contradiction in classification with type
label placement conventions for point, line, and area features, as well as dense label placement and trade-offs between placement rules
Your decisions about handling text should be influenced by how much text will appear and the role that text plays in the map. You may be positioning individual text elements in static positions, generating labels for thousands of features at once, or creating feature-linked annotation. Your goal is to understand these three kinds of labeling and be able to use them together for efficient and skillful map design.
The descriptions in this section use some ArcGIS terms (such as data frame, attribute table, and annotation), but the principles yielding these three kinds of labels—graphic, dynamic, and annotation—are generally applicable in GIS use. In contrast, map labels created in illustration software behave mostly as graphic text. Map labels styled with online mapping tools behave mostly as dynamic labels.
When you want to add general information to a map, such as titles, subtitles, sources, or notes, you place them as graphic text elements in the layout. These text elements are not associated with particular map features or data frames in ArcGIS. They remain the same size and in the same position as where you originally placed them, unless you drag them to a new location in the layout (figure 6.1).
Technically, you can add graphic text to a layout to label map features, but this can be problematic. The labels will not be associated with the data frame, so they will not move with changes to the map scale or the extent shown in the data frame. Thus, the label may no longer be near its feature as the map layout is edited. Adding graphic feature labels to a layout may work if the scale and extent of the underlying data frame is fixed and it contains only a few labels. Custom text editing is straightforward for these kinds of labels.
Figure 6.2 shows the disadvantage of adding feature labels as graphic elements to the layout rather than into the data frame. Panning within the data frame (the tan area) reveals more of the streets to the east of the three mapped sites. Notice that the labels and symbols for two sites (A and B) stayed linked to their geographic locations. The text for these sites was added into the data frame. The Site C label is a graphic element that was added to the layout. It remained in the same place on the page, not moving with its geographic location along N. Indian Hill Road.
Figure 6.1 The title, placed as graphic text, will stay in a fixed position in the layout regardless of changes to the geographic extent shown in the data frame.
Data sources: US Census Bureau, US Geological Survey (USGS). Maps by E. Guidero, Department of Geography, The Pennsylvania State University (Penn State Geography).
Figure 6.2 The Site A and Site B labels and symbols moved when the data frame was panned. The graphic text element for Site C remained static on the layout page because it was not added to the data frame (compare this map to figure 6.1).
Dynamic labeling allows you to automatically label multiple features in a layer, such as all roads in a road dataset that is represented on a map. If your GIS attribute table does not include a field populated with road names, consider selecting a new dataset or adding a field to the table before you start the tedious process of placing individual road names. It may take you longer to create labels as individual graphic text elements for each road than it would to enter them into a new field.
Figure 6.3 When you pan across a map with dynamic labels, you see why they are called “dynamic.” The labeling software repositions, adds, and removes labels to fit the available space. Compare the first map (A) to the second one (B): Church Rd, Green Bay Rd, and Hill Rd labels have moved. Data sources: US Census Bureau, USGS. Maps by E. Guidero, Penn State Geography.
When you use ArcGIS to create dynamic labeling, there are numerous settings that allow you to define classes of labels, set weights and buffers, allow overlapping labels, control duplication, constrain placement conventions, set the scale at which labels are shown, and set type characteristics. For example, to label all streets using 10-point Segoe UI with 1-point halos, set those parameters in the dynamic label properties for the streets layer. This is much faster than selecting and setting the characteristics of each street label individually. The result will be a reasonably complete set of map labels that are automatically placed on the map. As you change the map scale or extent, the positions of map labels will dynamically change (figure 6.3). At smaller scales, when there are many features to display, the software will label fewer features in a given area than at larger scales (figure 6.4). As more data layers are included on the map, labels will be removed to make way for additional labels competing for map space. In a dynamic map-use environment, not every feature needs to be labeled because the user can interactively query feature properties using GIS tools.
Figure 6.4 At a smaller map scale, the software dynamically selects fewer representative labels. Some streets do not have labels in the view but may be queried by interacting with the GIS software. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
If fixed label positions are important for a particular project (as they often are), you can customize the position and characteristics of text beyond the settings available for dynamic labeling by converting labels to annotation in ArcGIS.
Unlike dynamic labels, the position of annotation is fixed relative to the data, and annotation can be manually repositioned to a preferred location. Annotation characters can be edited, either manually or, in the case of feature-linked annotation, by updating the corresponding field in the attribute table. Feature-linked annotation will move if its feature is relocated or as the map view changes. Annotation cannot be moved outside of the data frame, so it is not a suitable format for titles and notes that belong in the area surrounding the map.
Labeling maps is one of the most time-consuming aspects of cartography, so you will probably want to use all three kinds of labels discussed here. To efficiently produce a well-designed map in ArcGIS, first use dynamic labeling to create a complete set of labels in generally correct locations. After setting as many global characteristics as you can for entire categories of labels, convert them to annotation. Annotation can be edited, so you can now refine placement and type characteristics for individual labels. Add graphic text only for marginal elements. This sequence saves time and produces a well-designed map by editing as few labels as possible. Figures 6.5 and 6.6 show road labels converted to annotation, selected, and moved or set bold to improve the map.
Figure 6.5 The dynamic street labels were converted to annotation. Selected pieces of annotation will be repositioned: the overlapping Ridge Rd and Meadow Rd labels will be moved apart, as will the Roger Ave and Brier St labels.
Data sources: US Census Bureau, USGS. Maps by E. Guidero, Penn State Geography.
Figure 6.6 The position of Ridge Rd has been adjusted. The text has been customized for the map topic: annotation for major streets near the three sites has been emphasized with bold type (see selection boxes around the edited annotation).
By categorizing feature labels, you help the reader find features on the map and see patterns among them. Sensible groups containing fewer features are easier to look through than large collections of uniformly labeled names (figure 6.7). Characteristics of the type help map readers identify each kind of feature. Groups of features may be distinguished by differences in symbols, but varying the type used for feature labels emphasizes these differences.
Figure 6.7 The lack of type differences between categories of labels make this map hard to read. Numbers associated with congressional districts (green boundaries) are difficult to find and interpret. Differences between lake and town labels are also difficult to sort out, especially when a town has a name like Sylvan Lake (middle right). See figure 6.8 for an improved version of this map.
Variations in the type style of feature labels can be used to categorize the features themselves. This approach organizes the themes on your map and makes it easy to understand. A reader will be able to find specific features quickly if type styles thoughtfully categorize feature types. For example, consider a map with 1,220 labels. If only twenty of those labels are river names, it will be easy for the reader to find a specific river if you can direct their attention to just those twenty labels using a distinctive font, style, or color to differentiate rivers from other features. The six congressional district numbers on the map in figure 6.8 are easier to find when distinguished with green.
Figure 6.8 The same map shown in figure 6.7 is designed with different type characteristics for categories of features. For example, different hues are used for congressional district numbers (green) and lake labels (blue). An italic serif font is also used for the lake labels, which makes them easier to search through because they are visually separate from the town labels.
Data sources: NHDPlus, US Census Bureau, USGS. Maps by E. Guidero, Penn State Geography.
A labeling challenge is to not inadvertently impose a hierarchy for different kinds of features when you categorize their names by type style. The following list contains characteristics that you can use to categorize features, without suggesting a difference in their magnitude or importance:
font
posture—roman (regular) versus italic
color hue
arrangement
The two labels in figure 6.9 use different fonts. The landform name is a serif font, and the cultural feature name below is set in a sans serif font.
Figure 6.9 Different fonts are often used to indicate different categories of features.
A professional-looking map generally relies on no more than two font families, so you should use fonts to distinguish only very broad categories. It is common to use a sans serif font to label all cultural features (roads, towns, and points of interest) and a serif font to label all physical features (mountains, valleys, and hydrography). With only a couple of font families on the map, you must rely on other type characteristics to establish additional feature categories.
Roman (regular) and italic versions of the same font are shown in figure 6.10. The creek label is italic and the street label is roman. Italic postures are harder to read than roman. They are designed to add emphasis to text by slowing the reader down. Use them to label categories with fewer or less important features.
Figure 6.10 Posture (regular and italic) indicates different categories of features.
Label hue can further accentuate category differences already established by a type style difference (figure 6.11).
Figure 6.11 Augmenting a style difference by hue further differentiates feature categories.
The arrangement of characters can also categorize features. Characters arranged to follow a line indicate a linear feature. The difference in arrangement between horizontal and curved labels in figure 6.12 implies two different kinds of features (stream and town), even though the features are not shown. The curve not only indicates location, but also categorizes its feature.
Variation in hue is again used in figure 6.13 to further distinguish two categories already established by the arrangement of the characters. Type characteristics can be combined either to emphasize existing categories or to subset categories into smaller divisions.
Figure 6.12 Labels that follow a curve indicate a different category of feature than horizontal labels.
Figure 6.13 Augment arrangement with hue to emphasize feature categorization.
Figures updated by E. Guidero, Penn State Geography.
Case, leading, and character spacing may also be used to indicate categories of features. For example, uppercase letters may indicate an area feature, and lowercase letters may be used for a point feature. Spacing between lines and characters can allude to an area feature. These three characteristics do not provide as clear a categorization of features, though, because they can also suggest differences in hierarchy.
Within one category of feature, there are often differences in magnitude that you would like to highlight. For example, cities have varied populations and rivers have different volumes of flow. Use type design to describe these hierarchical relationships.
You learned about hierarchy in the first chapter as it applied to map layout. When the notion is applied to type, it is also linked to map purpose. The more important, often larger, features should be more visually prominent, meaning higher in the visual hierarchy. Likewise, push less important features into the background to place them lower in the visual hierarchy. Thoughtful type design can accomplish this.
Hierarchy does not always follow feature size; it follows from your map purpose. A small feature, such as a tiny wetland, may be the most important feature on a map, and larger features, such as surrounding towns, may be background location information. In this situation, the wetland warrants the most prominent label.
The characteristics of type that help you establish hierarchies of features are the following:
point size
weight
scaling
color lightness
case
Point size is the most obvious type characteristic that can establish hierarchy. Larger type indicates a feature more important than those labeled with smaller type. Because size is so strongly associated with importance of a feature, you should not set a label to a larger point size simply to fill a large area. As an alternative, show large areal extents using character spacing instead (see area label placement guidelines later in the chapter). The three sizes of type shown in figure 6.14 reflect three population sizes.
Figure 6.14 Differences in point size create a hierarchy of features. Figure updated by E. Guidero, Penn State Geography.
Using a bold weight for type is an obvious way to move a feature category higher in the visual hierarchy, making it stand out from labels that are the same size and font but that are not bold (figure 6.15). Using a font family with more weights (such as light, medium, demi, bold, and black) can give you even more ability to show relative feature hierarchy.
Horizontal scaling can be used to indicate hierarchy. Although you should avoid artificially stretching or condensing a font, “condensed” or “narrow” fonts within a single font family are often available and can be used to distinguish the importance of features on a map. The two labels shown in figure 6.16 are the same point size and style, but Cayuga Lake is regular and Beebe Lake is condensed.
Altering the color lightness of type can be an effective way to push a label into the background, or lower in the map’s visual hierarchy. To maintain the integrity of the characters when printed, light labels are often set as bold type. This is an example of a compromise that uses a type characteristic in a different way than a simple hierarchy would suggest (bolder but not more important).
Figure 6.15 Differences in label weight also create a hierarchy.
Figure 6.16 A condensed font can be used to indicate a lower level in the label hierarchy.
Figures updated by E. Guidero, Penn State Geography.
In figure 6.17, Tompkins (a county name) is larger and bold because it is a larger feature, but it is lighter, which pushes it back in the visual hierarchy. The names of smaller towns are more important for this map’s purpose and remain forward in the visual hierarchy. In this case, the county category is established as a larger feature, but its visual importance is reduced with lightness.
Case also suggests hierarchy. Uppercase letters (capitals) indicate a larger or more important feature than one labeled by lowercase letters. In figure 6.18, the school name is all uppercase, and a building within the campus, a smaller feature that is lower in the visual hierarchy, is labeled using mostly lowercase letters.
Size, weight, scaling, lightness, and case can be used to communicate differences in the importance of features. Importance is determined by your map’s purpose, not just feature size.
Figure 6.17 Lightness indicates hierarchy by pushing labels to the background.
Figure 6.18 Uppercase labels are visually more important than lowercase labels.
Case, leading, and character spacing can be used to establish category, hierarchy, or both on a map. Uppercase lettering is often used for area feature labels regardless of feature importance because it can be spread across an area more elegantly than lowercase lettering. Thus, a group of uppercase area labels may be shown at smaller size to contradict the implied increase in importance that their case suggests.
Increasing character spacing and increasing leading pulls a label apart and diminishes its prominence in the visual hierarchy by making it less readable (figure 6.19). This subtle push into the background happens even though these characteristics make the label larger overall.
Character spacing is also used to imply the extent of a feature: a spaced label helps show that a particular feature in a group is larger than others, though not more important. Spacing also depends on the shape of the area and on the length of the feature name. Spacing has a weak effect on hierarchy because of these competing reasons for spacing lines and characters.
Figures 6.20 through 6.25 demonstrate how multiple type characteristics of three labels (two towns in Tompkins County) can be combined in ways that challenge the simple relationships described previously.
Figure 6.19 Character spacing tends to push a label back in the hierarchy because it reduces the label’s readability.
Size and lightness can be used to opposite effect. Tompkins is simultaneously pushed both up, by its larger size, and down, by its lighter appearance, in the hierarchy (figure 6.20).
Figure 6.21 uses lightness and size in another way: the background is a medium gray and labels are white and black, contrasting approximately equally with the background. This places them approximately equal in the visual hierarchy. Lightness shifts to a categorical rather than hierarchical distinction. Size has a strong influence on the hierarchy in this example, potentially pulling Tompkins up to greater importance than the previous example.
Figure 6.20 Tompkins is larger in size but lighter than the two town labels. All labels are the same weight, however.
Figure 6.21 An alternative use of lightness for labels that differ in size.
Figures updated by E. Guidero, Penn State Geography.
In figure 6.22, size is used with character spacing to reduce the large county label’s position in the hierarchy.
Figure 6.22 In this example, the Tompkins label has increased character spacing, making it more difficult to read and thus offsetting its larger point size.
Figure 6.23 uses three type characteristics: size, lightness, and weight. Size and weight pull Tompkins up, and lightness pushes it back down in the hierarchy.
Figure 6.23 Labels differ in size, weight, and lightness.
Figure 6.24 uses four type characteristics: size, lightness, weight, and character spacing. Size and weight pull Tompkins up, and lightness and spacing push back for a fairly balanced visual importance.
Figure 6.24 Labels differ in size, weight, lightness, and character spacing.
Size, case, weight, and lightness can be combined with character spacing and leading to separate features into groups and subtly imply differences in importance.
When you reverse your color scheme—that is, when you use a dark background and light labels—keep in mind that lightness and contrast work the opposite way as well. Very light or white labels will stand out the most against a dark background, and thus will be higher in the visual hierarchy than dark labels, which will blend into the background. Compare the appearance of Dryden and Tompkins in figures 6.25A, B, and C. The labels do not change lightness, but as the background is made darker in each successive image, the labels change places in the visual hierarchy. Whichever label has the greatest contrast with the background stands out as more important.
Figure 6.25 Using a dark background requires thinking differently about lightness and contrast. Here, the labels remain the same white and black, while the background gray varies, causing a change in visual hierarchy.
Figures updated by E. Guidero, Penn State Geography.
When making cached web maps, you need to consider the effects of anti-aliasing on layers or labels that contain transparency and will be overlaid onto other layers. Because screens always render images and objects in pixels, any circle or diagonal line, including letters, will appear to have jagged or stair-step edges. Anti-aliasing is a method used by software and operating systems to lessen or eliminate the appearance of pixels by smoothing the edges of objects or letters. The anti-aliasing process works by estimating where the curve or line of the object edge would fall on the pixel, and then setting appropriate pixels to varying levels of transparency, which fills in the gaps around the stair-step (figure 6.26). Anti-aliasing is more apparent when letters are thin, as nearly the entire letter is an “edge” that gets anti-aliased.
Figure 6.26 Anti-aliasing greatly smooths the appearance of on-screen text. Thicker letters are anti-aliased only at the edges. Figure by E. Guidero, Penn State Geography.
Anti-aliased letters will appear differently against different background colors, which is especially important when making map layers that are intended to be labeled and overlaid onto other layers or backgrounds. Consider a data layer of polygons of urban areas that will be overlaid onto a plain white background. The polygons themselves are orange, and the “empty space” outside the urban areas will be made transparent so the map reader can see the urban polygons and underlying imagery or base data. If any of the urban area labels are placed outside the polygons in the transparent space, they will be anti-aliased differently, depending on the background that appears below the labels. This will make the labels inside the polygons look different from the labels that sit above the background (figure 6.27). Notice in figure 6.27 how “Hoffman Estates” has a consistent appearance from letter to letter, while the “Strea” in “Streamwood” appears darker and thicker than the “mwood” portion.
Figure 6.27 Anti-aliasing on small letters can cause changes in appearance when text spans different backgrounds. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
This effect happens when the text is so narrow or small nearly the entire letter is anti-aliased, increasing the amount of transparency in pixels around the letters relative to object size. Increasing text size, bolding letters, or both, improves anti-aliased text because there is more “object” between edges that will not be anti-aliased (figure 6.28).
Figure 6.28 Although this text is a point size smaller than that in the previous example, it is bolded, reducing the difference in appearance across backgrounds. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
Sometimes you do not need refined map labeling, particularly in dynamic displays that have multiple query options. But when sloppy labeling impairs the effectiveness of a map, how do you improve the map and make the labels easier to associate with their features? Understanding the cartographic conventions for placing the labels of point, line, and area features will help you make wise choices when arranging all the text on your map. Labeling is a time-intensive aspect of cartography, so knowledge of labeling conventions will allow you to work more effectively to produce a quality map.
While we usually think labels are used to name locations, they also help your readers find locations. For point features especially, the labels are generally larger than the symbols, making the text easier to find than small symbols or graphics. This is why mapmakers follow conventions for label placement. The relationship between a location and its label should be as predictable as possible so the map reader does not need to work hard to determine which label belongs to which feature.
Text that labels a point location is best positioned next to that point, horizontally aligned, and with default character spacing. Cartographers sometimes choose to curve or rotate point labels to run parallel to curved lines of latitude across the map or to arc away from coastal locations into the adjacent water body. The more care you take in curving and rotating type, the longer it takes to make a map. Use horizontal alignment for point labels unless you are committed to a craftsman’s approach and have a production budget that can withstand the extra time that will take.
For horizontal labels, there is a series of preferred positions relative to the point location. There is some variability in these sorts of recommendations, but the position to the right and shifted up from the point location is always rated the best. That first position puts the point closest to the base of the beginning of the label (figure 6.29). Notice that label positions to the left and right of the point (1 to 4) are each shifted up or down. In figure 6.30, the “good” examples show labels in two of these positions.
Figure 6.29 Order of preferred positions for point location labels.
The “poor” example in figure 6.30 demonstrates the disadvantage of aligning a horizontal label directly to the right (or left) of a point location. The point symbol becomes incorporated into the shape of the word. The effect is exaggerated by using an open symbol that could be mistaken for a letter. Unless a word is completely unfamiliar, we use its shape to recognize it and read it. An aligned symbol becomes ambiguously part of the word and interferes with rapid recognition of it.
Figure 6.30 It is better to shift labels up or down relative to the point rather than aligning them with the point.
Figures updated by E. Guidero, Penn State Geography.
How you align a multiline label with a feature location is guided by the need to use the label as a location indicator. A label to the left of its point, for example, is best set right justified (figure 6.31).
Figure 6.31 Right, left, and center alignments for two-line labels positioned appropriately next to each symbol.
Another way to ensure that a label is easily recognized, and to aid visual search, is to rely on lowercase letters. They have more distinctive features than capitals do, helping us recognize words by shape, rather than reading them letter by letter. Figure 6.32 emphasizes this by showing just the top half of the same word. With only half the information available, you can probably read the lowercase label but, perhaps, not the uppercase version.
Figure 6.32 The top half of the word “Public.” Lowercase letters have more variation in shape and are, therefore, easier to read.
Figures updated by E. Guidero, Penn State Geography.
Using a lot of uppercase lettering on a map makes that map difficult to read. Lowercase labels are also not as long as corresponding uppercase labels, so they fit more easily into crowded places.
The recommended positions for point-label placement assume that there is no other information around the point. Because that is rarely the case, label placement becomes a process of making trade-offs between guidelines and the realities imposed by the geographic distribution of features. In figure 6.33, the two labels fit in the preferred position, to the right and up, relative to the points they label. But this places Pittsford equally close to both symbols. A reader can figure out which point Pittsford labels by a process of elimination, but you do not want to make your reader work that hard. A general rule is to position the text closer to the point it labels than to any other point with which it could be mistakenly associated.
Figure 6.33 Unambiguous association is more important than preferred positioning for point labels.
A set of dynamically positioned labels is shown in figure 6.34. Notice that they generally follow the hierarchy of preferred positions, though some are ambiguously located near more than one point, and many labels overlap each other or other symbols. Barnerville and Bramanville are squeezed between two point symbols, and Central Bridge could label any one of three points. Dwelly Corners and Sloansville stack on top of each other to resemble a three-line single label.
Positions are improved in figure 6.35. Some labels were moved out of preferred positions to accomplish this. Notice that Dwelly Corners was moved, and its text alignment was changed, to reduce ambiguity with Sloansville, and Mineral Springs was moved to eliminate overlap with Cobleskill.
Maintain a consistent distance between labels and their associated points throughout the map to keep the map neat and easy to read. Figure 6.36 highlights in blue the variation in gaps between labels and their symbols in the initial labeling attempt shown in figure 6.34. This inconsistency makes for sloppy design. These gaps were fixed in figure 6.35. This attention to detail is akin to kerning letter pairs. Because letters have different shapes at the corners of a label, the positions are refined by shifting them to close or open these varied gaps.
Choose a consistent size for the gap based on the size of symbols, the size of the label type, and the density of labels on the map. Make this decision while viewing the map at output size, not while it is enlarged on screen.
Figure 6.34 Labels placed automatically.
Figure 6.35 Label positions improved with custom edits.
Figure 6.36 Inconsistent gaps between points and labels: large gaps highlighted in dark blue and small gaps in light blue.
Data sources: US Census Bureau, USGS. Maps by E. Guidero, Penn State Geography.
The arrangement of other map features may also cause you to reject preferred positions for labels. The two “poor” examples in figure 6.37 show the label in the preferred position, but a curved road interferes with the label. Breaking the road for the label removes key information about the shape of the road and its relationship to the town. A better solution is to move the label to a more open position that is still near the feature. Automated labeling accomplishes this by giving the roads a slight weighting to repel labels.
Figure 6.37 Reposition labels to reduce interference with nearby features. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
Figure 6.38 shows the most open position for the label with added roads. The small road that interferes with the label is masked by slim halos to ensure that type is readable. A general rule for label placement is to break lines for labels. Where possible, position a label such that it breaks a vertically trending line to minimize the loss of information.
Figure 6.38 Breaking lines for labels makes a map more readable. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
Ideally, a point label is positioned on the same side of a linear feature as its point. In figure 6.39, the two labels are in preferred positions, but on the opposite side of the river from the points they label. Figure 6.40 shows that Monroe moved left (a less preferred position) and Adams moved down and right. A line also needs to be broken to accommodate Monroe. Choosing positions is a process of balancing conflicting guidelines.
Figure 6.39 Poorly positioned labels on the opposite side of a line feature from the point location. Data source: NHDPlus. Map by E. Guidero, Penn State Geography.
Figure 6.40 Labels are better positioned on the same side of the linear feature as the point location. Monroe is placed so that it breaks the most vertical part of the line feature, minimizing the amount of data masked. Data source: NHDPlus. Map by E. Guidero, Penn State Geography.
When labeling point features near bodies of water, position labels for coastal and shoreline features wholly in the water. Do not allow labels to span both water and land. In figure 6.41, Cannon Beach and Neskowin seem to be well positioned to the right of the points they label. But these coastal places are better labeled in the water to emphasize their coastal location. Pleasant Valley should not run from water to land. These problems are corrected in figure 6.42.
These examples demonstrate the most important guidelines for positioning point labels:
Position labels next to (right or left) and shifted up from the point feature.
Maintain a consistent visual distance between labels and point symbols throughout the map.
Break lines for labels, but minimize the frequency of breaks.
Position labels on the same side of a linear feature as the point feature.
Position labels for coastal features in water.
Contain labels entirely on land or entirely in the water.
The task of placing point labels is often a compromise between conflicting guidelines. Placing labels so that they are unambiguous and easy to read should guide your decisions.
Figure 6.41 Place labels for coastal features in the water and do not allow labels to span land-water boundaries.
Figure 6.42 Improved labeling of coastal and inland features.
Data source: USGS. Maps by E. Guidero, Penn State Geography.
Text that labels a line feature should follow along that line and be separated from it by a small gap. The text should not have noticeable character spacing, though a moderate increase in word spacing can be useful. If a linear feature is long, repeat the label rather than adding character spacing; do not try to stretch a label along the length of the line. Figure 6.43 demonstrates many errors in label placement.
Several placement problems make reading this map difficult:
Chula Creek and Pinehall Road ambiguously label both the river and road lines.
South Fork Falmouth River ambiguously labels two different river lines.
Snow Creek, Chula Creek, Falmouth River, and Hawthorne River tip upside-down.
Falmouth River is smaller than other labels of similar importance.
Three Mile River does not follow the linear feature and does not need to be broken into three lines.
Pinehall Road is too far above the line.
Halibut Tributary and Hawthorne River follow curves in the line too closely.
Halibut Tributary would fit better if positioned above the line feature.
Gypsum Creek would fit better if positioned below the line feature.
Hawthorne River is touching the line.
Figure 6.43 Poorly labeled linear features.
Data source: USGS. Maps by E. Guidero, Penn State Geography.
Improvements have been made to the map (figure 6.44):
Snow Creek is flipped to read right-side up.
Three Mile River is positioned to follow the river.
Halibut Tributary has been abbreviated, smoothed, and placed on top of the line feature for easier reading.
Chula Creek has been flipped to read right-side up and is abbreviated to fit.
Pinehall Road correctly labels just the road and is abbreviated to fit.
Pinehall Road and Hawthorne River are positioned with gaps between the label and line that are consistent with spacing used over the whole map.
Hawthorne River is moved to a straighter segment of the line and fit to a smoother curve.
South Fork Falmouth River has been abbreviated, resized, changed to a three-line label, and placed next to the river. It is not positioned to follow the river because the name is too long to fit to any of the visible segments. This is a difficult trade-off with no good solution.
Gypsum Creek has been placed below the line feature. Although not ideal, it helps prevent the descenders (y and p) from touching the line when placed the same distance away from the line as other labels.
Figure 6.44 Improved placement of linear feature labels.
Label size and weight have been changed to create visual hierarchy from smaller creeks to larger rivers. ArcGIS and graphics software can place text along user-specified curves. With only two points, you can create a simple bowed curve or a more complex s-shaped curve. Rarely will you need three or more points to define the line on which your label will sit. The curve should be a smooth, simplified approximation of the line feature, not an exact duplicate of it. The curve should not cause markedly different orientations for adjacent characters. Multiple points will make a line that is cumbersome to edit and inappropriately complex. Nine-point and three-point curves are illustrated in figures 6.45, 6.46, and 6.47.
Figure 6.45 Complex and simple curves used to label a line.
Figure 6.46 Example of overly complex splined text with nine points.
Figure 6.47 Example of suitable splined text with three points.
Data source: USGS. Maps by E. Guidero, Penn State Geography.
You can edit the shape of a curve by repositioning the points themselves (green squares, in the example) and by moving the handles (purple squares) attached to each point. Rotating the handles around the point, and changing their distance from it, will produce a variety of smooth curves. Experimenting by moving handles in all directions from their points is the best way to get used to altering the shape of the curve (figure 6.48).
Figure 6.48 Simple curves created by changing only the positions of two handles (purple squares). Figure updated by E. Guidero, Penn State Geography.
Positioning a label above a linear feature is preferable to positioning it below. Text fits best above a line because there are fewer descenders than ascenders on most labels. Also, a proper name with an initial capital letter will have a gap between the line and any lowercase letters without ascenders if placed below the line. As with other guidelines, there will be situations where you will not follow this advice to avoid overlapping other map features and labels. Regardless of the compromises you must make, avoid placing a label such that any portion of it reads upside down. Be especially aware of this when positioning vertically trending labels. Because of the shapes created by different combinations of ascenders and descenders within labels, the actual space between labels and features will vary across the map. Your challenge is to create spacing between them that looks consistent overall.
River names on the following map were placed dynamically by ArcGIS. Some are well positioned; some labels for major rivers are missing, others have awkward gaps or overlaps, or labels are ambiguously near more than one feature (figure 6.49). Refining the software’s label-placement settings would resolve some of these issues.
Some labels were repositioned manually in the next map (figure 6.50). Labels were moved to smoother bends on the rivers and rotated where appropriate. Sacandaga River was added as splined text to show an example of a smooth curve fit to a detailed line. Ideally, you would set a rhythm among labels aligned and running the same direction across a group of related linear features, but geographic reality often interferes with this goal.
Figure 6.49 Line labels placed automatically by ArcGIS.
Figure 6.50 Manually repositioned line labels with improved positions.
Data sources: NHDPlus, USGS. Maps by E. Guidero, Penn State Geography.
Positioning line feature labels according to the following guidelines creates maps that are easier to read:
Position labels to follow portions of line features.
Use default character spacing—do not spread characters along the line.
Repeat labels for long line features.
Put labels at the straightest and most horizontal portion of the feature.
Break lines that interfere with labels, but minimize the frequency of breaks by choosing label positions carefully.
Position labels above lines when practical.
Do not allow any part of a label to tilt upside down.
Use simple curves for labels to fit loosely to features with complex curves.
Maintain a small, consistent gap between labels and lines throughout the map.
Area feature labels indicate the extent of the feature by the way they are positioned. Area labels differ from point and line labels in that they may include character spacing. You can stretch labels using character spacing where needed, but do not express extent by making the point size of the characters larger, which instead indicates the relative importance of the feature.
Because most area features are not simple rectangles, the extent of these features can also be partly described using curved labels. If the shapes of these areas lend themselves to a series of simple and complementary curves, curved labels can be a pleasing addition to the design. But do not get carried away looping labels in so many directions across the map that it becomes indecipherable. Curved labels for areas are often more work than is worthwhile to create a pleasing but not distractingly busy set of labels.
Figures 6.51 and 6.52 show an area label that uses character spacing to spread across the feature. The first label is curved to suggest both the horizontal and vertical extent of the area. The second label is horizontal and spaced to suggest extent at the widest part of the area. This label is not as fancy, but it is a reasonable (and less time-consuming) solution.
Figures 6.53 and 6.54 show the same area labeled less suitably. The first label is centered in the area and its length only partly suggests the extent of the region. The slight angle is not sufficient enough to look purposeful, so instead it looks sloppy. The second label is centered in the region without enough spread to suggest extent.
Character spacing is better suited to uppercase rather than lowercase lettering. Thus, area labels are the most likely text elements on a map to be all uppercase, with other labels taking advantage of the better legibility of lowercase lettering.
Figure 6.51 The curved and character-spaced label identifies the extent of the area.
Figure 6.52 Horizontal character-spaced text also makes a suitable area label.
Figure 6.53 The slight angle on the label looks like a mistake rather than an effort to label extent.
Figure 6.54 A centered label misses the goal of indicating the extent of the area.
Figures 6.51–6.54 updated by E. Guidero, Penn State Geography.
Figure 6.55 Automatically placed area labels.
Figure 6.55 was dynamically labeled using ArcGIS. Labels are centered and are all the same size. Figure 6.56 uses character spacing to spread labels across larger areas. The character spacing was edited manually after converting the dynamic labels to annotation. Area labels were not curved in this example.
In this map, not all of the labels have enough room to be character spaced and remain within their county polygons. Expressing areal extent while retaining consistent label styles is a trade-off, and preference for one or the other is a subjective decision.
Labels spread over areas with character spacing should have a margin at the beginning and end of the word approximately equal to the character spacing. Do not spread words so much that the first letter crowds right against the boundary. Often some labels, like Doddridge, Webster, and Pleasants in this example, are longer than the area they label at the selected type size. Careful positioning prevents the labels from blocking characteristic parts of the area outlines. As always, break lines for labels unless lines are much lighter than the type.
In figure 6.56, labels are shifted up and down relative to adjacent labels. Avoid having a series of area labels that align horizontally, suggesting a “sentence” of labels, as seen with the poorly placed labels in figure 6.57.
Figure 6.56 Repositioning and increasing the character spacing may improve area labels.
Figure 6.57 Horizontal alignments yield poor area labeling.
Data source: US Census Bureau. Maps by E. Guidero, Penn State Geography.
Be aware that the letter I may look like a line segment if it is poorly positioned relative to area outlines. Serifs can help somewhat, but the best remedy is to shift the label so letters straddle lines and other features (large labels may also straddle smaller labels). This is particularly problematic with character-spaced area labels because letters may lose their relationship to each other. In figure 6.58, one could easily read Erie as two labels, one ending in “er” and the other starting with “E”.
It is particularly difficult to label large areas that cover much of a map, especially when only a portion of that area is shown. The letters of a conventionally character-spaced label are difficult to relate to the few visible area boundaries. One approach is to label along the boundary lines themselves, following the placement guidelines for linear features (figure 6.59).
Figure 6.58 Choose fonts and label positions to avoid ambiguous conflicts between labels and features. In this case, the letter I becomes lost in a boundary line. Figure updated by E. Guidero, Penn State Geography.
The following guidelines help you create clearer maps when positioning area feature labels:
Suggest the extent of the area by the position of the label.
Suggest extent using character spacing and line spacing (leading).
Suggest extent with simply curved labels.
Use uppercase letters when spacing characters.
Do not adjust text point size to fit labels into small areas or to fill large ones.
Stagger horizontal alignments of separate labels.
Adjust label position so gaps fit across features.
Ensure that individual characters are not mistaken for symbols (especially sans serif I and o).
Figure 6.59 States are labeled along the state boundaries instead of across state areas on the map. This alternative is particularly useful when only portions of areas are shown. Data sources: US Census Bureau, USGS. Map by E. Guidero, Penn State Geography.
Leader lines can assist in labeling crowded map areas. They improve clarity by allowing map labels to be placed further from their features while retaining an association with them. Use as few leader lines as possible; consider them a last resort in label placement. Do not put a leader on every label, as shown in figure 6.60.
The same map is labeled again with no leader lines, but the halos obscure multiple building features and a road intersection (figure 6.61).
A third solution uses three leader lines and breaks roads and buildings with thinner halos only when necessary. It moves some labels inside their features, appropriately treating them as areas. This option offers the clearest associations between labels and features without compromising the relationships between features (figure 6.62).
Figure 6.60 Over-reliance on leader lines results in an overly complicated map.
Figure 6.61 Over-reliance on breaking lines for labels and halos obscures important map features.
Figure 6.62 Redesign of some labels and use of select leader lines clarifies the map.
Data sources: US Census Bureau, Tompkins County ITS. Maps by E. Guidero, Penn State Geography.
Figures 6.63 and 6.64 demonstrate how editing dynamically placed labels can improve a map. Parameters were set to position point labels above and to the right of point features, and to put line labels above the line features. Overlaps were allowed to ensure that the maximum number of labels was placed. Type characteristics were also set as parameters prior to dynamic labeling to further distinguish between roads and landmark buildings. A type size large enough to accommodate map readers with poor vision was used (figure 6.63).
The dynamic labels were converted to annotation and repositioned to remove overlaps and clarify ambiguities. Adjustments were made, label by label, to improve associations between labels and features. The labels of three of the major streets were changed to bold type to improve feature classification (figure 6.64).
The large type and small size of the map made fitting a complete set of labels onto this map a challenge. After experimenting with various positions, just a single leader line to the library was necessary to label all roads and landmarks, and to resolve overlaps (figure 6.64). Notice that few of the point labels remained in the preferred position—up and to the right—after all the constraints on label positioning were considered.
Refining the positions of labels requires experimentation. It depends on type sizes and styles, so set characteristics carefully as you begin the process of dynamic labeling. Custom editing is an important, yet time-consuming, part of the design process. Use leader lines sparingly to make key links only in the densest areas of a map.
Figure 6.63 Automatically placed and symbolized labels in ArcGIS.
Figure 6.64 Annotation labels were repositioned to clarify links to features. A single leader line was used for the tightest space, associating the public library label with its symbol.
Data sources: US Census Bureau, USGS. Maps by P. Limisathian, Penn State Geography.
