2: Chapter Data Models
GIS Data Models
I
I
Coordinate Data
Common Spatial Data Models
I
Raster Data Models
Vector Data Models
I
Comparison Between Raster & Vector
I
I
Maps as Numbers
I
The Spatial Data Model
I
I
I
I
A logical data model is how data are organized for use
by the GIS.
Function of spatial data model is facilitate answering
questions like
?,
?,
?,
?,
?,
?
Remember, models are
,
based on
, and therefore, not necessarily
a good reflection of what true
GISs have traditionally used either raster or vector for
maps.
Why Topology Matters
Attribute Data and Types
Maps as Numbers
I
I
I
GIS requires that both feature characteristics
(attribute data) and maps (spatial data) be
represented as
.
Files can be written in binary or as ASCII text.
Binary is faster to
and
,
ASCII can be
and edited by humans
but uses
.
1
Maps as Numbers
I
Geographic data model may be defined as
I
I
the objects in a spatial database
plus the spatial relationship among them
2
Objects and Maps
A raster data model uses a grid.
Like a map, a GIS data model is a
digital representation of
features as either a
,
,
, and
.
I While most GIS systems can handle
raster and vector, generally only one is
used for the internal organization of
spatial data.
I
I
I
I
I
One grid cell is one unit or holds
.
Every cell has a value, even if it is
.
A cell can hold a number or an
standing for an attribute.
A cell has a resolution, given as the cell size in
.
Generic structure for a grid
3
Data Structure
(Hard Classifier)
Data Structure
(Soft Classifier)
000000000000
?
?
Water
Forest
Agriculture
Road
?
?
?
?
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
2
0
0
0
0
0
0
0
0
0
4
10
2
0
0
1
0
0
0
0
0
0
8
10
5
0
0
2
0
0
0
0
0
0
3
9
9
0
0
0
0
0
0
0
0
0
0
2
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
?
?
?
?
?
?
?
?
?
0
0
1
?
?
?
0
0
0
The mixed pixel problem
?
?
0
0
0
?
?
?
0
0
0
000000000000
?
0
0
0
000000001100
?
0
0
0
100000011100
?
0
0
0
100000011100
?
0
0
0
000000011100
?
0
0
0
000000001100
?
0
0
000000001000
The mixed pixel problem
0
000000000000
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Water
Forest
Agriculture
Road
Rule:
Winner takes all
4
The mixed pixel problem
Reduce chance of mixed
pixels by decreasing cell size?
Rule:
Water
Forest
Agriculture
Road
Road, Water and Ag dominate
1. If road present -> road,
2. else if water present -> water,
3. else if Ag present -> Ag
4. else winner
Types of data represented in
raster cells (Table 2-1, pg 45)
Class code
I Table ID
I Physical analog (e.g., elevation)
I Statistical value (e.g., population density)
I Point ID (nearest hospital)
I Line ID (nearest road)
I Polygon (area) ID (county)
I
Need for data compression techniques
File type
Uncompressed
Compressed
File name
redwoods.bmp
redwoods.jpg
Size
2305 kb
271 kb
Uncompressed
Compressed
0
500
1000
1500
2000
2500
File size (KB)
5
Need for data compression techniques
Raw data
Need for data compression techniques
Run-length codes
1111111111111111
1111111111111111
1111111111111111
1111111111111111
1111111111111111
1111111111111111
1111111111111111
1111111111111111
1111111122222222
1111111122222222
1111111111122222
1111111111122222
1111111111112222
1111111111112222
1111111111112222
1111111111112222
Need for data compression techniques
The quad-tree structure
0
1
01
11
2
136 1
82
81
82
11 1
62
11 1
62
12 1
42
12 1
42
12 1
42
12 1
42
Need for data compression techniques
UnCompressed
Quad-tree
3
2
Run-length
31
0
50
100
150
200
250
300
File (KB)
6
Need size for data compression techniques
Continuous surface data
The Vector Model
I
A vector data model uses points stored by their real
(earth) coordinates.
Lines and areas are built from sequences of points in
order.
Lines have a direction to the ordering of the points.
Polygons can be built from points or lines.
I
Vectors can store information about topology.
I
I
I
(0,100)
(40,100)
C
D
(60,100) (80,100) (100,100)
(0,100)
(40,100)
C
E
F
(0,65)
D
(40,50)
(70,45)
(40,50)
(70,45)
B
(25,0)
F
(20,55)
AB
(0,0)
E
(0,65)
(20,55)
A
(60,100)(80,100)(100,100)
H
(40,0)
(0,0)
(80,40)
G
(100,100)
(25,0)
(80,40)
HG
(100,100)
(40,0)
POLYGON file
ID
NUM
A
5
B
5
C
6
:
:
X-Ys
0,0
25,0
0,65
:
0,65
20,55
0,100
:
20,55
40,50
40,100
:
25,0
40,0
40,50
:
0,0
25,0
20,55
0,65
7
Topological errors
VECTOR
At first, GISs used vector data and cartographic
spaghetti structures.
Vector data evolved the arc/node model in the 1960s.
Hierarchical
I
I
I
Polygons are made up of Arcs (lines)
Arcs are made up nodes (endpoints) and vertices (points)
Vertices and nodes are recorded separately
I
I
I
Stored with the arcs is the topology (i.e. the
connecting arcs and left and right polygons).
I
5
6
C
9
D
3
10
E
11
F
4
15
7
16
12
AB
18
22
H
21
1
2
8
19 13
20
17
G
NODE (POINT) file
ID
X
Y
1
0
0
2
25
0
3
0
65
4
20
55
5
0
100
6
40
100
:
:
:
21
55
18
22
52
20
A7
5
3
A2
A10 9
6
4
A4
A3
AB
A14
A5 18
22
A6
14
1
11
C A8 D E
F A17
A11
A13
7
A1
10
A12
2
15
16
12
A16
19 13
17
HA15
G
20
21
14
8
POLYGON file
ID ARCS
A A1,A2,A3
B A3,A4,A5,A6
C A7,A8,A9,A2
:
:
NODE (POINT) file
ID
X
Y
1
0
0
2
25
0
3
0
65
4
20
55
5
0
100
6
40
100
:
:
:
21
55
18
22
52
20
EXTENT
0
25
20 40
0
40
:
:
0
0
50
:
65
55
100
:
ARC file
ID
BP
A1 2
A2 3
A3 4
A4 4
:
:
A16 16
A17 16
IP
1
:
11
EP
3
4
2
7
:
17
10
LP
B
C
:
F
RP
A
A
A
B
:
E
...
8
TOPOLOGY
I
I
I
I
I
I
Topology Matters
Topological data structures dominate GIS software.
Topology allows automated error detection and
elimination.
Rarely are maps topologically clean when digitized or
imported.
A GIS has to be able to build topology from
unconnected arcs.
Nodes that are close together are snapped.
Slivers due to double digitizing and overlay are
eliminated.
Comparison Raster to Vector
Characteristic
Data structure
Accessibility
Data Storage
Analysis
Coord precision
Coord conversion
Display/Output
- continuous
- discrete
Raster
simple
easy
large
easy
cell size
slow
Vector
complex
complex
small
complex
maybe greater
simple
good
poor
I
I
I
The tolerances controlling snapping,
elimination, and merging must be considered
carefully, because they can move features.
Complete topology makes map overlay
feasible.
Topology allows many GIS operations to be
done without accessing the point files.
Attribute data
poor
map-like
I
I
I
I
Attribute data are stored logically in flat files.
A flat file is a matrix of numbers and values stored in
rows and columns, like a spreadsheet.
Both logical and physical data models have evolved
over time.
DBMSs use many different methods to store and
manage flat files in physical files.
9
EXCHANGE
I
I
I
I
I
Most GISs use many formats and one data structure.
If a GIS supports many data structures, changing
structures becomes the users responsibility.
Changing vector to raster is easy; raster to vector is
hard.
Data also are often exchanged or transferred
between different GIS packages and computer
systems.
The history of GIS data exchange is chaotic and
some people feel it has been wasteful.
10