Upload
clement-gilmore
View
244
Download
0
Embed Size (px)
Citation preview
Price Ch. 2 Mapping GIS Data
‣ GIS Concepts
• Ways to map data
• Displaying rasters
• Classifying numeric data
Map Types and Data Types
‣ Single symbol maps
‣ Unique values maps
‣ Quantities maps
• Graduated color
• Graduated symbol
• Dot density
‣ Nominal data
‣ Categorical data
‣ Ordinal data
‣ Interval and Ratio data
Nominal data
‣ Names or uniquely identifies objects
• State names
• Owner of parcel
• Tax ID number
• Parcel ID Number
‣ Each feature likely to have its own value
‣ Usually portrayed on a map as labels
Single symbol maps
‣ Display all features with the same symbol
‣ Combine with labels to portray nominal data
Categorical data
‣ Places features into defined number of distinct categories
‣ Category names may be text or numeric
‣ Portrayed by different symbol for each category
Unique values maps
‣ Different symbol for each category or value
Geologic unitsVolcano types Road types
Ordinal data
‣ Type of categorical data
‣ Ranks categories along an arbitrary scale
• Low, Medium, High slope
• Village, Town, City
• Grades: A, B, C, D, F
• Rank of Best City to Live In: 1, 2, 3…
A 0-40%B 40-70%C 70-100%
Portrayed as categories but choosing variations in symbol size or color to indicate increase
Interval or Ratio data
‣ Interval data places values along a regular numeric scale
• Supports addition/subtraction
• Temperature, pH, elevation
‣ Ratio data places values along a regular scale with a meaningful zero point
• Supports addition, subtraction, multiplication, division
• Population, rainfall, median rent
Mapping numeric data
‣ Interval and ratio data must be divided into classes before mapping
‣ Mapped using variations in symbol size, thickness, or hue
Classed mapsGraduated color map(choropleth map) Graduated symbol map
Colors for choropleth maps
‣ Generally use change in saturation or close hues to indicate increase
‣ Avoid using too many colors which tend to mask increase
Normalizing classed maps‣ If the size of the sample
impacts the measured value, data should be normalized
• By percent of total
- Percent of farms in each state
- Percent of mobile homes in each state
• By another field
- Farms divided by area
- Mobile homes divided by total housing units
2-12
Number of farms
Number of farms per sq. mile
Unclassed maps
Proportional symbol map Dot density map
Chart Maps
Proportional chart map
Symbol psychology
Where is the water?
Where is there less rain?Which towns have more people? What’s there?
Where’s the danger?
Displaying rasters
Raster types‣ Discrete data
• Represents discrete objects such as lines or polygons
• Takes on relatively few values
• Adjacent cells often have same values
• Values may change abruptly at boundaries
‣ Continuous data
• Thousands or millions of potential values
• Few adjacent cells have same values
• Values may change rapidly from cell to cell
Raster types
‣ Thematic rasters
• Contain quantities that represent map data such as land use or rainfall
• May be continuous or discrete
‣ Image rasters
• Contain satellite or air photo data
• Generally represent brightness
• Usually continous
Displaying thematic rasters
Unique valuesDiscrete color
Interval/Ratio data
Classified
Categorical/Ordinal data
Stretched
Slicing
256 colorsBins raster values from 0-255, to match color ramp values
Stretching
256 colorsAfter slicing, stretching enhances display by removing less common values at the tails
Original slice Standard deviation stretch
Image display
Single band image Three band composite image
Image values represent brightness as 0-255 digital numbers (DN)
Stretching images
Images usually contain 0-255 range values already, but may not utilize full range.
Stretching maximizes brightness and contrast
Different stretches: Min-Max, Standard Deviation, Equalize…
No stretch Standard deviation stretch
0 255
Effects of stretching
No stretch Min-Max Standard deviation
RGB Color composites
Image bands
Composite color image
Landsat Band Combinations
True color3-2-1R-G-B
False color4-3-1
Other7-4-1
Bands 1-7 represent different wavelengths of light
Indexed color rasters
Common for scanned maps
More efficient way to store colors for scanned rasters than RGB bands
Each color on the map is indexed to a special unique values palette
Can modify the color choices individually
Colormap
Nodata
Nodata0 is another common nodata value
Transparency
Geology
Hillshade
Layer Properties: Display tab
Classifying numeric data
Classifying Data
• Applies to both vector and raster maps
• Different classification methods available
• Choice impacts map appearance and validity
• Best method depends on data distribution and objective of map
Same data, different classifications
Common data distributions
Value
Num
ber o
f sam
ples
Normal
Uniform
Skewed
Bimodal
Jenks Natural Breaks
• Exploits natural gaps in the data• Good for unevenly distributed or skewed data• Default method, works well for most data sets
Class breaks
Equal Interval
• Specify number of classes• Divides into equally spaced classes• Works best for uniformly distributed data
Defined interval
• User chooses the class size• Data determines number of classes• Works best for uniformly distributed data
Quantile
• Same number of features in each class• May get very unevenly spaced class ranges• Results depend on data distribution
Geometrical Interval
• Multiplies each succeeding class boundary by a constant• Works well for normal and skewed distributions
Standard Deviation
• Shows deviation from mean• User chooses units e.g. 0.5 standard deviations• Assumes data are normally distributed