Developing a cartographic language by Alastair Pearson

1
Developing a cartographic languagebyAlastair
Pearson
Geographical Data Presentation
2
Developing a cartographic language

• What is the image or map supposed to show?
• What impressions does the analyst wish to
  create in the mind?
• an impression of reality or database?
• can a cartographic language help?

3
Cartographic Language

• useful analogy
• make sense of marks on sheet of paper if we
  understand the graphic language
• a sort of written language

4
However

• definitions are not fixed - provided with the map
  in the form of a legend
• cartographer can vary meaning of symbols
• try to match variation of symbols to the
  variation of the things they represent.

5
Interrelated issues

• Spatial dimensions of the features to be mapped
• positional, linear, areal and volumetric
• level of measurement at which data are collected
  (categorical / numerical)
• graphic primitives used to represent the features
• point, line, area and volume

6
A simple cartographic language?

• point point line line etc
• create grammatical rules
• Scale dependency?

7
Scale dependency

• Categories of phenomena are scale dependent and
  subjectively interpreted
• e.g. city can be point at small scale, area at
  large scale
• subjective decision by cartographer
• simple languages are not possible

8
Matching graphic variables to the level of
measurement
9
Levels of measurement

• 1.Categorical or qualitative (lower level of
  measurement)
• i) Nominal (no natural or implied order)
• e.g. wood, farmland, urban etc, gender
• ii) Ordinal (implied order) e.g. high, medium,
  low risks, socio-economic status

10

• 2. Numerical or quantitative (higher level of
  measurement)
• Ordered with quantities assigned to that order
• i) Continuous
• (take any value within range of variation) e.g.
  rainfall, pressure, altitude, age, bedrooms
  (phenomenon exists between obervations)
• ii) Discrete phenomena (individually
  distinguishable from place to place)
• e.g. factory output figures (phenomenon does not
  exist between obs.)

11
Typology of graphic variables

• Need to match variation of phenomena to variation
  in graphic symbols
• matching must be intuitive - easily translated by
  variety of users

12
Key works

• Bertin J. (1983) 'Semiology of graphics
  Diagrams, networks, maps. Madison University of
  Wisconsin Press
• Modified by McCleary G.F. (1983) 'An effective
  graphic 'vocabulary''. IEEE Computer Graphics and
  Applications, March p.46-53
• Discussed by MacEachren A.M. (1994) 'Some truth
  with maps A primer on symbolization and design.'
  Washington AAG.

13
Bertin's Graphic Variables

• location in space
• size
• shape
• colour value or brightness
• colour hue
• colour saturation
• orientation
• texture

14
1. Location in space

• spatial location is numerical level of
  measurement
• numerical information by measurement of distances
  possible
• ordinal measurement also possible
• beware differences between projections
• location largely dictated by phenomena

15
2.Size
16
2. Size

• intuitive ordinal representation
• size easily differentiated by human perception
• larger symbols more visible
• useful for numerical information
• estimate ratio of one symbol to another
• avoid use for nominal distinctions

17
3. Shape

• indicate no inherent order
• represent nominal differences
• human perception less sensitive to variations in
  shape
• maps must be 'read' or studied carefully
• shapes of same size tend to look alike

18
4. Colour value or brightness

• High values - light low values - dark
• pure white to pure black
• 0 - 100 black
• human perception of brightness does not
  correspond linearly to measured reflectance

19
Colour value
20
5. Colour hue

• red, green, blue etc
• hue is measure of wavelength reflected or emitted
• long (red) to short (blue)
• used for nominal distinctions
• humans have high visual acuity for hue
• greatly reduces if symbol is small

21
Colour hue
22
Colour hue (cont.)

• temptation to use hue variation on all maps
• applying hue variation to ordered or numerical
  data is risky
• hues have a logical order (electromagnetic
  spectrum) but difficult to remember
• a range of hues ordered by value colour
  perception

23
Hue and contrast

• spectrum divides into two value ranges
• Highest value hue (yellow) occupies centre of
  spectrum
• i) yellow - green - blue - violet
• ii) yellow - orange - red

24
Yellow - orange - red
25
6. Colour saturation

• relates to the purity of the hue
• used for providing a visual order
• narrow range of wavelengths purer hue
• wide range muddy or impure
• rarely used in isolation
• used with variations in hue and brightness to
  enhance perception

26
7. Texture

• spatial frequency of components of a pattern
• fine - coarse range
• large areas are required to distinguish pattern
• suited to area differences - though can be
  applied to point or line symbols

27
Texture
28
Texture (cont.)

• can represent ordinal but best for nominal
  categories
• also used to assist in depth cueing in creating
  visula levels
• coarser - closer finer - further away

29
8. Orientation

• differences in orientation are very noticeable
• often misused
• useful for exploratory analysis when meaning is
  unimportant

30
Orientation (cont.)

• varying orientation differentiates without
  necessarily implying order
• can imply order if used to show wind direction or
  slope aspect

31
Now fit the variables to the level of measurement!
32
FIN