Lesson 4: Color in Symbol Design

• Overview

Useful resources for this lesson include:

• I.   Color as a Map Symbol

Examples:
Point, Line and Area Symbols
Topographic Map - Associative Color
Point Symbols - Qualitative Color Contrasts
Qualitative Hue
Quantitative Graded Series
Color Combinations for Figure-Ground (not a map symbol, but to accentuate the color symbols on the map)
 

• II. Qualitative Color Symbolization

Hue with pattern and/or letter overprint are methods to create a greater variety of qualitative hues on area maps that contain many areas. In this way the same hue may be used two or three times, for example, once without a pattern and twice with separate pattern overprints. Patterns can be in for form of lines, dots, pictographs or in reverse (i.e., white on dark colors). Likewise, the same color may be applied to a group of several features (e.g., soil subgroups within a major soil group) with each individual feature identified by a unique letter overprint. Such maps are somewhat more time consuming to interpret as the map and legend must be consulted carefully in order to distinguish and associate map features. Patterns can also be disturbing to the eye if used in large map areas.

Hue conventions are a common qualitative use of color, particularly in topographic and other qualitative physical maps. Some typical hue conventions for color include: blue for water, green for vegetated areas, brown for contours or other land surface symbols, yellow for dry or sparsely vegetated areas, and red for roads (from Robinson, et.al., 1995, pg. 382). Several government mapping agencies have developed color specifications or standards for point, line and area symbols to be used on all maps for a given scale. Users of these maps learn to associate certain colors with associated features and come to expect such color standards on all maps from that agency.

There are some perceptual problems associated with qualitative color mapping. First of all, it is difficult to create many different hues that are both equal in value/chroma and clearly distinct as was shown in the Bosnia-Croatia maps above. Robinson, et. al., (1995, pg. 386) suggest that between 8 and 15 different hues can be easily distinguished as long as they are very different from one another. While the human eye can differentiate similar hues when placed in succession, however, the same hues may not be easily differentiated when placed among other hues, because the appearance of a hue may be altered by the appearance of different surrounding hues as was discussed in Lesson 2: Color Perception. A given hue will look darker, for example, when surrounded by lighter hues and lighter when surrounded by darker hues (simultaneous contrast). It may also be difficult to match a map hue with it's associated legend box hue (successive contrast). The same hue will also look darker when covering larger areas, and lighter when covering smaller areas. In order to maintain similar visual strength, therefore, when using too many hues for maintaining fairly equal visual strength, it is best to reserve the somewhat darker hues for smaller areas and somewhat lighter hues for larger areas.

Cynthia Brewer describes and illustrates some interesting uses of qualitative hue in conjunction with a second map variable whereby the qualitative variable is represented with different hues and the quantitative variable comprises "lightness steps" within each hue. In a qualitative binary example, hues represent different employment sectors, while one lightness step per hue indicates membership status within the OECD. A qualitative sequential example differentiates employment sectors by hue (qualitative) and percent employed in the dominant sector (sequential) by two lightness steps within each hue. (See below for a more detailed description of Brewers color schemes).

• III. Quantitative Color Symbolization

There are several ways in which to achieve a visual hierarchy of hue to represent differences in quantity or magnitude. These include single hues or grey scales, hue progressions, double ended hue progressions, and full spectral progressions. The various hue scales are covered in detail within both textbooks (Dent, 1999 and Robinson, et. al., 1995) but with somewhat different terminology. Dent (1999, pgs. 300-305) presents quantitative color schemes by Cuff, Mersey, and Brewer, while Robinson, et. al., (1995, pgs. 387-396) discuss their own works. Some of each will be covered briefly below with reference to relevant authors and with links to appropriate graphic examples. Apart from Cindy Brewer's graphics, the map and graphic examples linked to this lesson were developed by Bo King and Laurie Garo, Cartography Laboratory, UNC-Charlotte, 1998-1999.

Single Hue Series

This method of representing  ranked or quantitative classes of data uses one hue and varies the chroma, and possibly the value, to create a range or graded series from light to dark. This method is referred to in Dent (1999) as the Gray and Simple Hue Plan by Cuff and the Hue-value series by Mersey. It is called the Single Hue Progression in Robinson, et.al. (1995).  In a single hue graded series one hue is selected and its chroma is varied by increasing the percentage of ink used to create the hue, e.g., equally varying the percentages of both cyan and yellow to create light to dark green as shown in the legend example linked here. This plan works well for representing  5 or fewer classes of data, although in some cases it may work for 6 classes as is demonstrated again in this legend example. Depending on the hue, it may or may not be necessary  to vary the value of the hue by adding black. With more than 5 classes, however, it is difficult to create enough visible difference between chroma steps of a single hue. It is therefore necessary to either add black to the top color(s), as with 6 classes, or to go to a hue progression for 7 or more classes. Hue progressions are covered in the next section. Note that for black and white mapping a  Gray Scale Progressionis possible with 5 or fewer classes. Note also that yellow is not a suitable color for a hue progression, especially when there are 4 or more classes.

The following are some examples of  5-class, single hue choropleth maps on Per Pupil Expenditure for Public Education in North Carolina, 1994-1995:

Poor Single Hue Choropleth Map:  In this map, a gradation of Cyan is used to represent 5 classes of data. The graded series is ineffective because there is not a clear visual difference between each hue in the series. All hues are fairly dark, and the highest and lowest classes do not stand out as visually distinct. In addition, it is difficult to match each hue on the map with it's corresponding hue in the legend. Note: the percentages of Cyan used here (varied chroma) are 100, 85, 70, 55, and 40%. The percentage range varies depending on the hue used, but in general, the lightest hue should be quite light, with 15-20% difference at the lower end of the series and 20-25% difference at the high end. Depending on the hue, it may be necessary to alter the value for the highest class by adding a small amount of black to the full chroma hue (e.g., 100% Cyan and 10% black will result in a cyan that is just a bit darker, thus more distinct than the others in the series).

Good Single Hue Choropleth Map:The map linked here represents a good example of chroma/value gradation for a single hue. Cyan, in this case, is presented with much greater variation in chroma from hue to hue, and the value is lowered just slightly for the highest class (darkest, visually strongest hue in the series) by adding a touch of black. Percentages of Cyan used here are 100% Cyan, plus 10% black, then 80, 48, 24, and 8% Cyan, each with 0% black. Notice that apart from the 2 lightest classes, there is at least 20% difference in chroma between each hue.

Good Gray Scale:  Shown at this link is a black and white version of the education expenditure map. It is possible to create an effective graded series of gray tones for 5 or fewer classes. This map uses 100, 70, 50, 24, and 8% black. The result is that  each class is visually distinct and can be matched from map to legend. Notice that at the upper end of the scale, a difference of 30% is used to help distinguish a very dark gray tone from black, while a difference of 16% is sufficient to differentiate the lighter gray tones for the lower classes. Note also that when using black, the black polygon outlines dissappear and must, therefore, be printed in light gray (shown here) or white in order for area boundaries to be visible.

Hue Progression

A Hue Progression is a graded series of 2 or more hues where the chroma/value ranges from light to dark. These are referred to in Dent (1999) as the Part-Spectral Plan by Cuff, and the Spectral Series by Mersey. Robinson, et. al., (1995), call them Complementary Hue Progressions. While a single hue gradation will work for a 5-class map, sometimes a hue progression will enable more clear differentiation between classes. A hue progression is normally used for more than 5 classes, always for more than 6. An effective hue progression should combine 2 or more hues that are complementary on the color wheel or go well together visually. Some examples include yellow-to-green, yellow-to-blue, yellow-to-orange-to-red-to-brown, and blue-to-purple. Blue and green do not work well together, and green to red does not form an effective color combination for hue progressions. In addition to hues, the chroma and value of each hue in the progression is varied in order to achieve the light to dark progression.

The following are some examples of  Quantitative Choropleth Maps on Per Pupil Expenditure for Public Education in North Carolina, 1994-1995 that use a hue progression for 5 classes.

Poor Examples of 5-Class Choropleth Maps

Blue-to-Green    This first example demonstrates the ineffective combination of blue and green as a hue progression and also illustrates how the lack of gradation of chroma and value among hues results in a more qualitative appearance to the map. There is no distinct variation in lightness to darkness (chroma,value) within the hue progression, thus the differences in data quantities from county to county is not evident.

Pink-to-Purple   The map shown here has a more pleasing combination of hues and somewhat more clear differentiation from hue to hue within the progression. But there is still not an effective gradation of chroma and value, thus the very high quantities do not stand out significantly from the very lowest quantities on the map.

Good Example of a 5-Class Choropleth Map

Yellow-to-Green:   Shown here is a nice example of an effective hue progression. The yellow-to-green progression of hues works well together visually. In addition, the chroma and value are varied to give the light to dark effect that highlights well the highest and lowest data quantities, and provides clear differentiation of each class. Map reading is easy as well because each hue can be easily matched from the map to the legend.

The following are some examples of  7-Class Ordinal (Ranked) Choropleth Maps on North Carolina Architectural Survey Status, 1994:

Poor Example of a 7-Class Choropleth Map

Yellow-to-Green:  Shown here is another yellow to green hue progression. This hue progression is ineffective, however, because the chroma and value are not varied appropriately to reflect the "Most" to "Least" classification of the data. The yellow representing the "Least" category is visually stronger than most other hues in the progression, thus, the least important category stands out visually as strong as the most important category, and is visually stronger than several other categories on the map. When using yellow in a progression it is important to remember that a full or near full chroma yellow is very bright and will stand out strongly on the map. Yellow, when used at the lower end of the scale, should be quite light, e.g., 20 or 30%. Likewise, when creating the medium to light yellow-greens at the lower end of the scale, be sparing with percentages of yellow used.

Good Example of a 7-Class Choropleth Map

Yellow-through to-Brown:  Here, a  yellow-orange-red-brown hue progression is used to represent this ordinal data set. Like yellow to green, this hue combination works well. In this map, the gradation of chroma/value from light to dark effectively portrays the low to high ranking of the data. In terms of map interpretation, each unique hue can easily be matched from map to legend.

Double-Ended Hue Progression

Double ended hue series' are used for representing two opposing characteristics of a data set on one map. Typical double-ended data sets include temperatures above and below zero, economic gain and loss, and positive and negative population change. The progression is called the Double-Ended Plan in Dent (1999), and the Bi-Polar Hue Progression in Robinson, et., al (1995). Shown at the link above is a legend for a double-ended hue progression for representing temperatures above and below zero using the color associations of reds for warmer temperatures (above zero) and blues for cooler temperatures (below zero). An effective double-ended progression uses 2 distinctly contrasting yet visually complimentary (pleasing) hues that clearly differentiate the opposing sides of the data. In addition, when each characteristic (e.g, positive vs negative quantities) has the same number of classes,  the chroma/value gradation per hue should be as similar as possible. In all cases, the highest and lowest classes within each side of the data should stand out as visually distinct.

Below are examples of a double-ended progression to represent Change in Manufacturing Employment in North Carolina, 1973-1995:

Double-Ended - Poor Example:In this map the gains in manufacturing employment are represented by red-to-orange, while the losses are shown in brown-to-gray. Red and brown are not distinctly different enough for there to be clear visual differentiation of gain vs loss. Also, the chroma/value gradation of the browns is softer than that for the reds, thus the "gains" categories stand out much stronger than do the losses. The greatest gain is much more evident than the greatest loss. Also, the brown gradation to gray is not an effective hue gradation for brown.

Double-Ended - Better Example:  Here is another example of the red and brown combination but this time the chroma/value gradation is much more similar  for each side. In this map the greatest gains and losses stand out clearly and distinctly from one another and from other classes. The result is better than the first example, but still not ideal. Note that the middle orange is brighter than the middle brown, thus stands out more strongly. And at the lighter end of each hue scale, the light orange and light brown are not visually distinct, thus, the lower gains and lower losses are not clearly and easily separable.

Double-Ended Progression - Good Example:  This double-ended plan uses color associations to separate the 2 components of the data set. Green represents gains in employment, symbolizing monetary gain, while red illustrates economic loss (negative quantity, or being "in the red"). The hue differences are clearly distinct yet visually pleasing together. The chroma/value gradation is nearly equal for each hue, thus the high, middle, and low categories are clear for both the gains and the losses.

Full-Spectral Progression
 

Full Spectral Plan in Dent (1999), and Full Spectral Progression in Robinson, et. al (1995)

Two Variable Hue Progression
 
 
 

Exercise 4

References

Go to Overview  above

Return to: Module Overview OR Go Back to: Table of Contents

Created 2/20/98 by Laurie A. B. Garo. Last updated 6/10/99 by lg.
The URL for this page is http://www.uncc.edu/lagaro/cwg/color/color_symbol.html