Geospatial Data Types

1
Geospatial Data Types
2
Data Types

• Two general views to organizing spatial data
• Objects
• Monitoring measurement points, rivers, structures
• Have attributes or features attached to them
• Point, line or area format
• Values exist at entity locations
• Commonly stored and rendered in raster format
  (grids)
• Fields
• Continuous data such as temperature gradient
  fields and satellite imagery
• Values exist over an area
• Every location has a value
• Commonly stored and rendered in raster format
  (grids)

3
Haining, 2003
4
Raster and Vector Data Models
Real World
600
1
2
3
4
5
6
7
8
9
10
B
G
1
Trees
500
B
G
G
2
3
B
400
G
B
G
4
Trees
Y-AXIS
5
B
G
G
300
B
BK
6
G G
House
B
7
200
B
B
8
River
100
B
9
500
200
600
400
300
100
B
10
X-AXIS
Vector Representation
Raster Representation
adapted from Lembo, 2003
5
Vector Advantages and Disadvantages

• Advantages
• Good representation of reality
• Relatively compact data structure
• Accurate graphics
• Disadvantages
• Complex data structures
• Some spatial analysis is difficult or impossible
  to perform

6
Raster Advantages and Disadvantages

• Advantages
• Simple data structure
• Uniform size and shape
• Computationally cheaper to process
• Disadvantages
• Large amount of data
• Less visually pleasing (blocky)
• May lose information due to generalization
• Projection transformation is difficult
• Different scales between grids can make
  comparison difficult

7
Objects and Fields
Objects and fields can be transformed to the
other type
Objects Vectors
Fields Raster
adapted from Bolstad, 2002
8
Vector vs. Raster
Burroughs, 1996
9
Landcover Raster Grid
Legend
Mixed conifer
Douglas fir
Oak savannah
Grassland
10
Raster Grid
Matrix of Equal-Area Cells
Pixel
columns
Abbreviation for PICTURE ELEMENT, which is the
smallest unit in an image. In raster based GIS
systems, attribute information can be assigned to
each pixel.
rows
The bounding box defines the geographic extent of
the grid in terms of its coordinates min_x,
max_x, min_y, max_y
11
Grid File Format (ASCII)
ncols 6 nrows 6 xllcorner 210 yllcorner
370 cellsize 20 nodata_value 0 5, 6, 7, 8, 10,
13 5, 7, 8, 10, 12, 13 4, 5, 8, 12, 15, 15 3, 4,
5, 13, 15, 16 3, 5, 11, 14, 15, 17 2, 4, 5, 16,
16, 17
12
Table Format
X Y Value
220 380 2
220 400 3
220 420 3
220 440 4
220 460 5
220 480 5
240 380 4
240 400 5
240 420 4
240 440 5
240 460 7
240 480 6
13
Contoured Plots
Also known as an Isopleth Plot
14
Map Scale

• Map scale is based on the representative
  fraction, the ratio of a distance on the map to
  the same distance on the ground.
• Most maps used in GIS fall between 11 million
  and 11000.
• A GIS is scaleless because maps can be enlarged
  and reduced and plotted at many scales other than
  that of the original data.
• To meaningfully compare maps in a GIS, both maps
  MUST be at the same scale

15
Scale of a baseball earth

• Baseball circumference 226 mm
• Earth circumference approx 40 million meters
• Scale is 1177 million

16
Scale Dependent Measurements
How long is Maines coastline?
length340 km
length355 km
length415 km
From Longley et al., 2001
17
Resolution
25 meter
5 meter
Same number of pixels (rows and columns)
1 meter
18
Resolution
1 meter
5 meter
25 meter
Same geographic area (m X m)
19
(No Transcript)
20
(No Transcript)
21
Spatial Dimensionality
Another way to classify spatial object types is
by their dimensionality
0-dimensional, L0 points and nodes 1-dimensional,
L1 lines 2-dimensional, L2 (x,y) areas,
polygons 3-dimensional, L3 (x, y,
z) volumes 4-dimensional, L4 (x, y, z, t) 3-D
plus time
22
2.5 Dimensions
23
Attributes
Attributes are the values and properties of an
object or entity
24
Types of Attributes
A common approach to classifying attributes is
based on their level of measurement

• Nominal Simply identifies or classifies an
  entity so that it can be distinguished from
  another. e.g. sensor ID, building name
• Cannot be manipulated using mathematical
  operations. However, frequency distributions are
  meaningful.
• Ordinal Values based on an order or ranking,
  e.g. agricultural potential classes
• Cannot be manipulated using mathematical
  operations. However, frequency distributions are
  meaningful.
• Interval Differences between entities are
  defined using fixed equal units, e.g. Celsius
  temperature
• Can be manipulated using addition and subtraction
• Ratio - Differences between entities can be
  defined using ratios, e.g. distance
• Can be manipulated using multiplication and
  division
• Cyclic - differences between entities depending
  on repeating sequence, e.g. wind direction

25
Structured Query Language (SQL)
SQL is a formal search language that allows you
to work with, access and filter data stored in a
relational database format
The most common use for SQL is to retrieve
subsets of data based on specified conditions
SELECT column name
FROM data table name
WHERE data condition
26
ArcGIS Select by Attribute
SELECT FROM MO_STN WHERE O3 gt 80 AND PM25 gt 15
27
Defining Reclassification Categories
28
Classification Schemas
Natural breaks classes are defined according to
apparently natural groupings of data values. (GIS
programs that automatically determine classes
usually base them on relatively large jumps in
data values.)
Quantile breaks classes are defined by having an
equal number of observations
Equal interval breaks classes are defined by
uniform intervals
Standard deviation breaks classes are defined by
differences from the mean value.
29
Color Brewer
http//www.personal.psu.edu/faculty/c/a/cab38/Colo
rBrewerBeta.html
30
Graphic Visualization Components
31
Summary

• Two general data types object field
• Generally, handled as either vector or raster
• Data can have multiple attributes (properties)
  associated with features or grid cells
• Levels of measurement helps formalize the
  arithmetic and statistics that are appropriate
  for a particular dataset

32
Date Topic Reading Problem Set Tutorial
31-Aug GIS Overview Bolstad Chp 1 Gorr, Chp1
7-Sep Geospatial Data Longley Chp 3 Gorr Chp2-3
14-Sep Projections and Coordinate Systems Bolstad Chp 3 Problem Set 1 distributed Gorr Chp4, Chp 5 (p. 172-180)
21-Sep Feature Analysis Bolstad Chp 9 Gorr Chp 8 (p. 272-290), Chp 9
28-Sep Surface Analysis Bolstad Chp 10/11 PS1 due PS2 distr. Handout Suitability Analysis
5-Oct Spatial Data Analysis Bolstad Chp 12 Handout California Air Pollution
12-Oct Spatial Modeling / Web GIS Bolstad Chp 13 PS2 due Gorr Chp 8 (p. 291-299), Handout Groundwater Modeling
19-Oct Exam / Project Presentations
33
Gistutorial\UnitedStates States Counties Cities
Capitals Utah Nevada Pennsylvania Gistutoria
l\Layers Tutorial3-1.mxd Tutorial3-NativeAmerica
ns.mxd