1/30/2018

Geographic Information System

Arthur Godfrey, Dr. Batte

Lecture 1
A gentle introduction to GIS

1
 1/30/2018

DATA

Definitions
“…a system of hardware, software, and procedures designed to support the capture,
management, manipulation, analysis, modeling, and display of spatially referenced data
for solving complex planning & management problems” (Rhind, 1989).
“…a computer system capable of assembling, storing, manipulating, and displaying
geographically referenced information…” (USGS, 1997)
“…a set of computer-based systems for managing geographic data and using those
data to solve spatial problems” (Lo & Yeung, 2002).

“….a computer system that allows the analysis and display of data with a spatial
component (Phillips, 2002).
4

2
 1/30/2018

We may distinguish three important stages while working with geographic

data:
- Data preparation and entry
- Data analysis
- Data presentation

Spatial data and Geo-information

Another subtle difference exists between the terms DATA and INFORMATION.
Most of the time, we use the two terms almost interchangeably, and without the risk of
being ambiguous.

By DATA, we mean representations that can be operated upon by a computer.

More specifically, by spatial data we mean data that contains positional values.
By INFORMATION, we mean data that has been interpreted by a human being.
Humans work with and act upon information, not data.
6

3
 1/30/2018

Lecture 2
Geographic information and spatial data types

Objectives
Distinguish between different types of geographic phenomena like fields and
objects and between discrete and continuous phenomena.
Understand the principle of “spatial autocorrelation”
Understand the concept of the following computer representation:
• Regular tessellation
• Irregular tessellation
• Point, Line and Polygon

Can apply the topological properties of interior and boundary to define

relationships between spatial features.
8

4
 1/30/2018

Notions to be grasped
❖ Select a suitable computer representation, based on the different types of
geographic phenomena.
❖ Difference between the real world and computer representations.
❖ Different types of geographic phenomena.
❖ Boundaries (crisp – fuzzy).
❖ Regular versus irregular tessellations (quadtrees).
❖ Vector representations (TIN, point, line, area).
❖ Properties of geometric space used for spatial data (Euclidean space, metric space,
topological space, interior and boundary).
❖ Two dimensional topology.
❖ Scale and Resolution. 9

Spatial Data Modeling

10

5
 1/30/2018

Different types of geographic phenomena

Some phenomena manifest themselves essentially everywhere in the study area, while
others only occur in certain localities.
1. A Field is a geographic phenomenon for which, for every point in the study area, a
value can be determined. (temperature, barometric pressure and elevation).
2. Objects populate the study area, and are usually well distinguishable, discrete,
bounded entities. The space between them is potentially empty.

11

Geographic Fields
Fields can be discrete or continuous, and if they are continuous, they can even be
differentiable.
In a continuous field, the underlying function is assumed to be continuous
(temperature). Continuity means that all changes in field values are gradual.
In a differentiable field, we can determine a measure of change per unit of distance
anywhere. (for elevation, this measure would be slope).

12

6
 1/30/2018

Geographic Fields
In a discrete field,
Discrete fields, cut up the study space in mutually exclusive, bounded parts,
with all locations in one part having the same field value. (land classification,
geological units)
One may note that discrete fields are a step from continuous fields towards
geographic objects:
• Discrete fields as well as objects make use of ‘bounded’ features.
• A discrete field still assigns a value to every location.
13

Different types of geographic phenomena

14

7
 1/30/2018

Continuous Field Discrete Field

Geological units
Elevation 15

Spatial Autocorrelation
• locations that are close are more likely to have similar values than
locations that are far apart.
• Principle is used when an interpolation is performed.
• Continuous field phenomena are based on the spatial autocorrelation
concepts.
• These fields although characterized by a continuous function must be
finitely represented.
• This also applies for lines.

16

8
 1/30/2018

Boundaries
Which phenomena have boundaries:
-Discrete fields
-Objects
Two different types of boundaries:
• Crisp boundaries
• Fuzzy boundaries

17

Fuzzy boundaries (dynamic) Fuzzy boundaries (natural)

18

9
 1/30/2018

Computer representations of geographic information

Regular tessellations (mosaic)
▪ Is a partition of space into mutually exclusive cells that together make
up the complete study area.
▪ In a regular tessellation, the cells are the same shape and size.
▪ The field attribute assigned to the cell is associated with the entire area
occupied by the cell.
▪ Square, regular tessellations are known under various names: raster,
grid.
▪ The size of the area that a grid cell represents is called the raster’s
resolution.
19

Examples of Regular Tessellations

20

10
 1/30/2018

Regular Tessellation Cont..

Advantages:
• We know how they partition space. This leads to fast algorithms.
Disadvantages:
• They do not adapt to the spatial phenomenon we want to represent.
• Cell boundaries are both artificial and fixed: they may or may not
coincide with the boundaries of the phenomenon of interest.

21

Irregular Tessellations

Again, these are partitions of space into mutually exclusive cells, but now,
the cells vary in size and shape, allowing them to adapt to the spatial
phenomena they represent.
One example: region quadtree.
It splits up the area into four quadrants.

22

11
 1/30/2018

Region Quadtree

This procedure stops when all cells in a quadrant have the same field value.

23

Vector representations
In vector representations, an attempt is made to associate
georeferences with the geographic phenomena explicitly.

Coordinate pair from some geographic space also known

as a vector

Note: rasters do not explicitly store georeferences of the phenomena they

represent.
24

12
 1/30/2018

Georeferences

Instead, they provide a georeference of the lower left corner of the raster, for
instance, together with an indicator of the raster’s resolution, thereby
implicitly providing georeferences for all cells in the raster.
25

Triangulated Irregular Network (TIN)

It is built from a set of locations for which we have a measurement.
The locations can be arbitrarily scattered in space (not a nice regular grid).
Observe that in 3-D space, three
points uniquely determine a plane,
as long as they are not positioned on
the same line.
If we restrict the use of a plane to the
area between its three anchor points,
we obtain a triangular tessellation of
the complete study area. 26

13
 1/30/2018

TIN
A plane fitted through the anchor points has a fixed aspect and gradient and
can be used to compute an approximation of e.g. elevation of other
locations.

27

Does TIN belong to vector?

YES, each anchor point has a stored georeference.

NO, as the chosen triangulation provides a tiling of the entire study space.
However, the cells of the tiling do not have an associated stored value as is
typical of tessellations.

28

14
 1/30/2018

Point representations
Points are defined as single coordinate pairs (x,y), when we work in 2D or
coordinate triplets (x,y,z) when we work in 3D

Line representations
• Used to represent one-dimensional objects
(roads, railroads, canals, rivers…).

• Line is defined by 2 end nodes and 0-n internal nodes.

• An internal node or vertex is like a point that only serves to define the
line. 29

Line representation
A network is a collection of connected lines.

Area representations
When area objects are stored using a vector approach, the
usual technique is to apply a boundary model. This means
that each area feature is represented by some arc/node
structure that determines a polygon as the area’s
boundary.

30

15
 1/30/2018

Area representation.

A simple but naive representation of area features would be to list for each
polygon simply the list of lines that describe its boundary.
Why is this not working?
-Redundancy
-If we want to find neighboring polygons, we have to do time-consuming analysis
(topology)

31

Area representation

32

16
 1/30/2018

Topology and Spatial Relationships

Topology deals with spatial properties that do not change under a

transformation

E (polygon, area) in figure above is still inside D (another polygon or area)

33

Topology and Spatial Relationships

The neighborhood relationships between the areas remain, and the boundary
lines have the same start and end points.
The areas are still bounded by the same boundary lines, only the shapes and
lengths of their perimetry have changed.

34

17
 1/30/2018

Topology and Spatial Relationships

We can define within the topological space, features that are easy to handle
and that can be used as representations of geographic objects.
These features are called simplices as they are the simplest geometric
shapes of some dimension:
▪ point (0-simplex),
▪ line segment (1-simplex),
▪ triangle (2-simplex),
▪ and tetrahedron (3-simplex).
35

Topology of 2D
We use the topological properties of interior and boundary to define
relationships between spatial features.
Region A →Boundary A and Interior A
Region B →Boundary B and Interior B
When the two regions meet this is defined as the Boundary of A intersects
the boundary of B.
The interiors do not intersect and the boundary of one does not intersect the
interior of the other.
36

18
 1/30/2018

Topology of 2D
The mathematical definition of meets.
A meets B = interior (A) ∩interior (B) = Ø∧
interior (A)∩boundary (B)= Ø∧
boundary (A)∩interior (B)= Ø∧
boundary (A)∩boundary(B)≠Ø

There are 7 other spatial relationships that can be defined in a

similar way.
37

Spatial Relationships

38

19
 1/30/2018

Rules of topological consistency

39

2½-D
• Rules of 2D topology apply.
• Nodes have a z-value.
• A node with an x-and y-coordinate can have only one z-value.
• Two different nodes, with identical x, y values, cannot have different z-
values.
• Consequently, true solids cannot be represented in a 2½D GIS.

40

20
 1/30/2018

2½-D

41

Scale and Resolution

Map scale can be defined as the ratio between distance on a paper map and
distance of the same stretch in the terrain.

1:50,000 →1 cm on map = 50,000 cm in reality

1 cm on map = 500 meters in reality

Large-scale →much detail of a small area

Small-scale →few detail (world map)
42

21
 1/30/2018

Scale and Resolution

The size of the area a pixel cell represents is called resolution.

43

Representation of geographic fields

44

22
 1/30/2018

Representation of geographic objects

45

Spatial Data Modeling

46

23
 1/30/2018

Raster and vector compared

47

Lecture 3
Data Processing Systems

48

24
 1/30/2018

Contents
• What are data processing systems?
• Functional GIS components
– input
– output
– data storage
– spatial queries & analysis
• GIS & databases
-databases
• DBMS
• Example
- using GIS & DBMS together 49

Data - Processing - Systems?

50

25
 1/30/2018

Data Processing Systems

Data processing systems are computer systems with:

1. appropriate hardware components for processing, storage and

transfer of data.
2. software components for management of the hardware and the
data.

51

Hardware and Software trends

• From minicomputers to PC’s
• Increasing performance
– Speed
– Storage
• Software boom
• Computer networks
• Mobile phones
• Open systems
52

26
 1/30/2018

Computers CPU: 3 GHz

Mem: 1 Gb
Disc: 160 Gb
More than 30 years of development!

CPU: 1500 times faster

CPU: 2 mHz
Mem: 8000 times larger
Mem: 128 Kb
Disc: 16000 times larger
Disc: 10 Mb

53

Disciplines in GIS

Geography
Remote Sensing
Surveying
Cartography
Photogrammetry
Computer Science Geo-informatics
Mathematics
Archeology
Forestry
Earth Sciences
Urban Planning
Legal Sciences
54
Economy

27
 1/30/2018

GIS Components
A GIS is a computer-based system that provides the following four sets of
capabilities to handle geo-referenced data :
1. data capture and preparation
2. data management
3. data manipulation and analysis
4. data presentation

55

GIS Functional Components

Data Capture
And Preparation Storage and Maintenance

Manipulation and
Data Presentation
Analysis

56

28
 1/30/2018

Data Input
• Available data
– Tape, CD, Net
• Manual input
– Keyboard, digitizer
– On-screen, photogrammetry
• Semi-automatic digitizing
– Line-following
• Automatic digitizing
– Scanner
57

Data Preparation Operators

• Format transformation functions
• Geometric transformations
• Map projections
• Edge matching
• Graphic element editing
• Coordinate thinning

58

29
 1/30/2018

Spatial Proximity
Objects that are near in geographic space should be near in
storage space as well.

59

Raster data storage

Values of neighboring cells are close together.

60

30
 1/30/2018

Spatial analysis
GIS uses spatial analysis functions:
Operators that use spatial data to derive new geo-information.

61

Analysis of Spatial Data

• Classification
• e.g. determine land use
• Overlay functions
• e.g. spatial logic
• Neighborhood functions
• e.g. buffering
• Connectivity functions
• e.g. network analysis
62

31
 1/30/2018

Data Output
• Hardcopy
– printer
– plotter
• Softcopy
– on-screen
• Data sets
– CD, DVD, Internet
63

Database Management Systems (DBMS)

64

32
 1/30/2018

Reasons for using a DBMS

A DBMS:
• Supports very large datasets
• Maintains data correctness
• Supports concurrent use
• Provides a high-level, query language
• Uses a data model
• Provides backup and recovery functions
• Reduces data redundancy
65

The Relational Model

A Data Model is a language for the definition of:
• the data structures for storing the base data.
• the integrity constraints.
• the computer programs used to manipulate the data.

66

33
 1/30/2018

The relational data model

67

The relational data model Foreign Keys

68

34
 1/30/2018

69

70

35
 1/30/2018

Other DBMS’s
Relational databases are well suited for:
• Large amount data
• Simple data types
Not setup well to deal with spatial data
• Object-oriented / object-relational models
• GIS applications will rely on DBMS add-on application packages
71

Vector Data Storage

Vector data and relationships are stored in tables.

72

36
 1/30/2018

Geometric Features in GIS

The user can examine information about:
where things are and what they are like.
• From spatial data to attribute data and vice versa

Raster • geometric attributes

• non-geometric attributes
• relationships

Vector
73

Topology
Topological data structure provides:
• Automated way to handle digitizing errors
• Reduced storage space (Poly boundaries stored only once)
• No overlapping polygons
• Advanced spatial analysis (adjacency, networks)

74

37
 1/30/2018

Paper maps versus GIS

Spatial information -
• Acquisition
• Storage
• Maintenance
• Analysis
• Output

75

GIS and DBMS support

GIS software provides support for
– spatial data
– thematic or attribute data
DBMSs are much better in table functionality.
GIS applications now make use of external DBMS for attribute data
support.
76

38
 1/30/2018

Lecture 4

• Data quality
• Spatial referencing

77

Objectives
Data quality
• Explain the relationship between accuracy and precision.
• List main components of spatial data quality.
• Understand and apply the RMSE for measuring location accuracy (section 4.3.1).
Spatial Referencing
• Understand basic concepts of spatial referencing pertinent to published maps and spatial data.
• Know the meaning of items in a map legend (vertical and horizontal datum, ellipsoid, coordinate
types, etc).
• Distinguish between local and global reference systems.
• Understand the concepts of height measurements (in relation to the geoid or an ellipsoid).
• Comprehend datum transformations.
• Know the classification of map projections.
• Know the distortion properties of map projections.
• Describe the process of selecting a suitable map projection.
• Grasp the logic of map projection equations and the principles of transforming
78
maps from one
projection system to another.

39
 1/30/2018

Components of Spatial Data Quality

• Positional accuracy
• Attribute accuracy
• Temporal accuracy
• Lineage
• Logical consistency
• Completeness
These components are considered to be an adequate description of
data quality.
79

Inaccurate data

80

40
 1/30/2018

Inconsistent data (I)

The overlapping parcel boundaries is the result of digitizing from

existing parcel field maps !
Case: Ghana 81

Inconsistent data (II)

Corrupted registration system !

Case: Ghana 82

41
 1/30/2018

Incomplete data

83

Lineage

Refers to the history of the data set

• Source of data.
• Methods of acquisition or derivation.
• Date and scale of source material.
• Structure of source material into map sheets or administrative
units.
• Accuracy and precision of instrumentation.
84

42
 1/30/2018

Lineage
Provide information about;
• transformations,
• interpolations,
• cartographic displacements,
• exaggerations etc.
Any data set created from distinct sources should be provided
with lineage information to identify the quality of the data set
in terms of ‘fitness for use’. 85

Attribute accuracy in Remote Sensing: Example of an error matrix

The overall accuracy is (62+18+12) / 100 = 92 % 86

43
 1/30/2018

How to test positional accuracy

For discrete objects, e.g. cadastral boundaries, use an independent source
of higher accuracy.
• use a larger scale map.
• use check points determined by field survey.
For fuzzy objects, e.g. vegetation and soil boundaries, use fuzzy set
techniques.
These have the ability:
• to handle logical modeling (map overlay) operations on inexact data.
• of using a variety of natural language expressions to qualify uncertainty.

87

How to estimate positional accuracy

Compute accuracy from knowledge of the errors introduced by different sources

Example
1.0 mm in source document
0.5 mm in map registration for digitizing
0.2 mm in digitizing

If sources are combined independently, an estimate for the overall accuracy, will be:

( 1.02 + 0.52 + 0.22)1/2 = 1.1 mm

88

44
 1/30/2018

How to measure positional accuracy

Position accuracy is normally measured as a Root Mean

Square Error (RMSE)
Average deviation of all measured positions from their
true positions.

89

Computation of RMSE

90

45
 1/30/2018

Computation of Systematic Error

The observed errors should be checked for a systematic error
component, which may indicate a, possibly repairable, lapse in the
method of measuring.

91

Accuracy and Precision

Measurement errors are generally

described in terms of accuracy. The
accuracy of a single measurement is; “the
closeness of observations, computations
or estimates to the true values”

Accuracy should not be confused with precision, which is “a

statement of the smallest unit of measurement to which data can be
92
recorded”.

46
 1/30/2018

Spatial Referencing

Contents
➢ Spatial reference surfaces and datums
• The Geoid – vertical datum
• The Ellipsoid – horizontal (geodetic) datum
➢ Local and global datums
➢ From Earth to map
➢ Coordinate transformations
➢ Map projections
• Classification of map projections
• Universal Transverse Mercator (UTM)
• Map projection selection 93

The Geoid – Vertical datum for height referencing

94

47
 1/30/2018

The Geoid, the height reference surface

• To describe heights, we need an imaginary surface of zero
height.
• This surface must also have a physical meaning.
• A surface where water does not flow, a level surface, is a
good candidate.
• The most obvious choice is the level surface that most
closely approximates all the Earth’s oceans.
• This surface is known as a Geoid.

• The Shape of the Geoid is influenced by Gravity.

• The Geoid represents an Imagery Sea Level.
95
• Every point on a Geoid has the same height all over the world.

Vertical datums
Starting from the mean sea level points, the heights of points on the
Earth can be measured using a technique known as geodetic
leveling.

A leveling network implements a local vertical datum:

• network of leveling lines starting from the Amsterdam tide-gauge, showing some of
the benchmarks;
• how height is determined for some point, working from the nearest benchmark.
96

48
 1/30/2018

The Ellipsoid – horizontal datum for horizontal referencing

97

The ellipsoid and the horizontal datum

We also need a reference surface for the description of the
horizontal coordinates of points of interest.
Since the geoid is defined physically and not mathematically, it
cannot be used as a reference surface for horizontal coordinates.
The mathematical surface that is simple enough and most closely
approximates the local mean sea level is the surface of an oblate
ellipsoid.
98

49
 1/30/2018

Local and global ellipsoids

The ellipsoid is chosen in such way that it best fits the

surface of the area of interest (the country)

99

The Ellipsoid Reference surface for location

measurements
An ellipsoid with specific dimensions ‘a’ and ‘b’ as half the
length of the major and minor axis respectively is chosen
which best fits the local mean sea level.
Then the ellipsoid is positioned and oriented with respect
to the local mean sea level by adopting a latitude (Φ) and
longitude (λ) and height (h) of a so-called fundamental
point and an azimuth to an additional point.

We say that a local horizontal datum is defined by

(a) dimensions (a, b) of the ellipsoid,
(b) the adopted geographic coordinates Φ, λ and h of the
fundamental point, and
(c) azimuth from this point to another.

The local horizontal datum is implemented through a so-called triangulation network.

100

50
 1/30/2018

Commonly used ellipsoids and datums

101

Commonly used ellipsoids and datums

102

51
 1/30/2018

Global and Local Datums

Global ellipsoids and datum’s to approximate
the earth-as-a-whole - with the aid of
satellites- are becoming more in use (e.g.
WGS84, ITRS, ETRS89).
Changing or re-adjustment of the local
ellipsoids and datums is taking place in many
countries. 103

Datum change (example)

Mapping organizations do not only coach the user community about the
implications of the geocentric datum. They also develop tools to enable
users to transform coordinates of spatial objects from the new datum to the
old one. This process is known as datum transformation.
The tools are called datum transformation parameters.
104

52
 1/30/2018

Datum re-adjustment (example)

Why do the users need these transformation parameters?

• Because, they are typically collecting spatial data in the field using satellite
navigation technology.
• They also typically need to represent this data on a published map based
on a local horizontal datum.
The good news is that a transformation from datum A to datum B is a
mathematically straightforward process.
The bad news is that the estimated parameters may be inaccurate if the 105coordinates
of the common points are wrong.

Datum Transformations

This is often the case when we transform coordinates

from a local horizontal datum to a geocentric datum.

The coordinates in the local horizontal datum may be

distorted by several tens of metres because of the
inherent inaccuracies of the measurements used in the
triangulation network.
106

53
 1/30/2018

From Earth to Map

107

Overview

108

54
 1/30/2018

109

Coordinate Transformations

110

55
 1/30/2018

Projection change using projection equations

111

Change a map projection

Including a datum transformation

112

56
 1/30/2018

Datum transformation Molodensky 2D-transformation

113

Datum Transformations

114

57
 1/30/2018

Datum transformation from ITRF to Potsdam

115

Overview of Coordinate Transformations

116

58
 1/30/2018

2D (direct) Cartesian transformations are used for the

rectification of images using 2D Ground Control points.

117

Map Projections

118

59
 1/30/2018

Map projection principle

119

Map projection equations

A map projection is a mathematical function by which 2D

Geographic coordinates are transformed into 2D
Cartesian map coordinates.

Forward equation
Inverse equation

120

60
 1/30/2018

Map projection equations (example)

The map projection equations for the Mercator

projection (spherical assumption)

Forward equations

Inverse equations
121

Map projection class

122

61
 1/30/2018

Map projection property

Conformal
• Shapes/angles are correctly represented (locally).
Equivalent ( or equal-area )
• Areas are correctly represented.
Equidistant
• Distances from 1 or 2 points or along certain lines are
correctly represented. 123

Conformal map projection

Cylindrical

124

62
 1/30/2018

Equivalent map projection

Cylindrical

125

Equidistant map projection

Cylindrical

126
* also called Plate Carrée projection

63
 1/30/2018

Robinson projection (compromise)

Pseudo-Cylindrical

127

Classification of map projections

128

64
 1/30/2018

Universal Transverse Mercator

129

Transverse Mercator projection

130

65
 1/30/2018

UTM-projection

131

UTM-Zones

132

66
 1/30/2018

UTM Zone numbering system

133

A standard map projection system

The use of a standard map projection system for all countries
is becoming more in use ( U.T.M. )
Re-adjustment of the local map projection systems is taking
place in many countries.

134

67
 1/30/2018

Map projection selection

135

Criteria for the selection of a Map projection

Shape of the area (determines the projection class).
Purpose of the map (determines the property of the
projection).
Position of the area (determines the aspect of the
projection).
136

68
 1/30/2018

Shape of the area

137

Purpose of the map

Conformal
• maps which require measuring angles (aeronautical charts,
topographic maps).
Equivalent
• maps which require measuring areas (distribution maps).
Equidistance
• maps which require reasonable area and angle distortions
(several thematic maps).
138

69
 1/30/2018

Position of the area

139

Choosing a Map Projection

The ideal map projection for any country would be either
azimuthal, cylindrical, or conical projection, depending on the
country’s shape, with a secant projection plane located along
the country’s main axis.
The property of the map projection depends on the map
purpose.
140

70
 1/30/2018

CHAPTER 5
Data entry and Data preparation

141

From Alpha to Omega

• Spatial data input
– Direct spatial acquisition
– Digitizing paper maps
– Obtaining spatial data elsewhere
• Data preparation
• Point data transformation
• Advanced operations on rasters !!!!
142

71
 1/30/2018

Direct Spatial Data Acquisition

143

Digitizing Paper Maps

A cost-effective method of data capture is the DIGITIZING of
existing maps.
This requires the conversion of an analogue map into a
digital map.
A number of digitizing techniques exist.

144

72
 1/30/2018

Digitizing Process

145

On-tablet Digitizing (manual)

146

73
 1/30/2018

On-screen Digitizing

147

On-screen vs. Manual Digitizing

• It is more comfortable for the operator.
• It is more accurate.
– zooming facilities
• It is faster.
– semi-automatic, digitizing and editing at the same time
• There are updating procedures.
– geometrically corrected satellite imagery and scanned aerial photo’s
can be overlaid with the old vector data.
148

74
 1/30/2018

Map Registration

149

Map Registration

150

75
 1/30/2018

What to Digitize ?

Point, Line, Polygon, or...?

151

Digitizing Errors

152

76
 1/30/2018

Point & Stream Mode Digitizing

153

The Scanning Process

154

77
 1/30/2018

Working Principle
A digital scanner illuminates the document and measures
with a sensor the intensity of the reflected or transmitted light.

155

156

78
 1/30/2018

Scanning Resolution
Scanning Resolution can be expressed in:
• in millimeters
• in microns
• in dots per inch
Example:
pixel size = 0.05 mm = 50 μm = 508 dpi
157

Scanner Resolution (pixel size)

Minimum required resolution depends on the details in the map and
the digitizing technique:

• Manual on-screen map digitizing: around 200-300 dpi

• (Semi-) automatic digitizing: thickness >1/3 of the thinnest line
– lines should be at least 3 pixels wide
– a resolution of 300-600 dpi is mostly sufficient
• Hardcopy output: 2x the resolution of the printer:
– nowadays 1200 dpi
• Photogrammetric applications: 800-2400 dpi 158

79
 1/30/2018

Scanning Modes
Line Art, 1 bit

159

Grey Scale, 8 bit

160

80
 1/30/2018

Index Color, 8 bit

161

Full Color, 24 bit

162

81
 1/30/2018

Scanner Output
The scanner output is only a digital copy of the source
document in raster cell values.

• Data are NOT structured into classified and coded objects

• To obtain this, the data have to be vectorized and further
structured

163

The Vectorization Process

The conversion from raster to vector
Step 1. Skeletonizing
Step 2. Feature Forming

164

82
 1/30/2018

165

Original Line Work

166

83
 1/30/2018

Scanned Line Work

167

Skeletonized Line Work and Line Forming

168

84
 1/30/2018

Original and Vectorized Line Work

169

Vector Editing

170

85
 1/30/2018

Vectorized Lines (after vector editing)

171

Feature Forming
• Splitting lines to form line segments and nodes.
• Joining line segments to form polygons and features.
• Feature coding

172

86
 1/30/2018

Vectorized Lines
(after vector editing and data structuring)

173

Vector / Raster

174

87
 1/30/2018

Selecting a Digitizing Technique

175

Spatial Data elsewhere

176

88
 1/30/2018

Spatial Data Clearinghouses

177

Spatial Data elsewhere

Means:
Data in different formats and standards, and with different
quality !

178

89
 1/30/2018

Data Preparation

179

Spatial Data Preparation

Spatial data preparation consists of editing data that is

to be entered into the GIS database.

180

90
 1/30/2018

Data Preparation cont…

Images may need enhancements and (re)classification
Vector data may require editing, such as the removal of errors
and generating polygons.
Data may need to be vectorized or rasterized to match
existing data sets.

181

Data Preparation cont…

Additionally, processing includes associating attribute data

with objects, either through manual input or reading digital
attribute files into the GIS.

182

91
 1/30/2018

Data Checking and Clean-up

Data sets must be checked for consistency and

completeness.
This requirement applies to the geometric, topological and
semantic quality of the data.

183

Data Quality
• Geometric
Are the coordinates correct?
• Topological
Are the relationships correct?
• Semantic ≈ "meaning"
Does the data make sense?
184

92
 1/30/2018

Clean-up operations for vector data

185

Cleaning & structuring of vector data

186

93
 1/30/2018

Topological and Semantic Problems

187

Associating Attributes

188

94
 1/30/2018

Data Structure Conversion

Rasterization: assigning point, line and polygon attribute
values to raster cells.
Vectorization: producing vector data from a scanned
image; used to identify features in satellite images or aerial
photographs .

189

Topology Generation
• Topology deals with spatial relationships between features.
• Topological relations are used in a GIS to perform spatial
operations like:
• overlaying
• buffering
• network tracing
• shortest-path routing
190

95
 1/30/2018

Topological Relationships
“Green area is … covered by … blue area” (by 2D geographic extent)

191

Combining Multiple Data Sets

Three fundamental cases:
Data sets of:
• same area, different accuracy
• same area, different representation
• adjacent areas, which must be merged into a single data
set
192

96
 1/30/2018

Differences in Accuracy
Combining two data sets derived from different maps may
lead to sliver polygons.

193

Differences in Representation

194

97
 1/30/2018

Merging Data in a Database

195

Merging data in a database

Some GIS systems have merge or functions to solve
problems arising from merging adjacent data sets.

196

98
 1/30/2018

Point Data Transformation

197

Interpolation
Very often, we only have a limited number of
observation points, but we're interested in every
location in the area.
How do we find the 'missing' points?
Answer: using interpolation techniques.

198

99
 1/30/2018

Two Fundamental Cases

Discrete Field
• Qualitative
• What is it?

Continuous Field
• Quantitative
• How much is it?

199

Field Representation from Point Measurements

Qualitative point Quantitative point

measurements measurements
(e.g. geological, soil units) (e.g. elevation, temperature)
200

100
 1/30/2018

Discrete & Continuous Field

Land use Elevation

Qualitative Quantitative
P = Forest P = 112 meter
Q = Grassland Q = 168 meter
201

Discrete Field from Point Data

Thiessen polygons

Any location is assigned the

value of the closest measured
point.

202

101
 1/30/2018

Continuous Field from Point Data

Interpolation techniques:
• Trend surface fitting
• Moving window averaging
• Triangulation

Many other techniques exist!!

203

Trend Surface Fitting

The key in trend surface fitting is to find a

mathematical surface f(x, y).
Uses regression techniques to best fit the original
measurements.

204

102
 1/30/2018

Example, 1-D

205

Examples of trend surfaces obtained from regression

techniques, 2-D

206

103
 1/30/2018

Trend Surface Fitting

Simple tilted surface:

• z = f(x,y) = c1.x + c2.y + c3
• f(x,y) = -1.83934.x + 1.61645.y + 708782

This formula can be used to compute a value for any

location (x, y).

207

Global and Local Trends

Global trend surface fitting assumes that the entire

study area can represented by the same
mathematical surface.
In many cases, it is more realistic to partition the
study area (e.g. at mountain ridges for elevation
data) , and apply trend surface fitting techniques
for each part.
208

104
 1/30/2018

Moving Window Averaging

Computes new raster cell values from a

number of quantitative point measurements
within a square window.

209

Moving Window Averaging

Possible parameters settings:
• Resolution of the raster cells
• Shape/size of the moving window
• Selection criteria
which measurements do participate
• Averaging function
equal weight or weight factor

210

105
 1/30/2018

Inverse Distance Weighting

Squared
Computed field value = ?

211

Inverse Distance Weighting

10.66

Squared
Computed field value = 10.66

212

106
 1/30/2018

Triangulation

Another way of interpolating quantitative point

measurements is by triangulation.

Delaunay
213

TIN construction

A TIN is typically based on a Delaunay triangulation

of a set of input data points.

214

Bad triangulation Good triangulation

107
 1/30/2018

Delaunay & Thiessen

215

Advanced Operations on Continuous Fields

• Slope angle calculation

• Slope aspect calculation
• Slope convexity / concavity calculation
• Slope length calculation
• Hill shading

216

108
 1/30/2018

Advanced ops., cont….

• Three-dimensional map display
• Change in elevation through time
• Catchment delineation
• Dynamic modeling
– (e.g. surface run-off, erosion)
• Visibility analysis

All these operations apply a technique known as Spatial Filtering

217

Spatial Filtering
Linear Filtering or Convolution

The difference with moving window averaging is that the moving window in
filtering is itself a little raster, which contains cell values that are used in the
computation for the output cell value. This little raster is known as the filter
218

109
 1/30/2018

Computation of Slope Angle

gradient

219

Slope & Aspect

220

110
 1/30/2018

Spatial Filtering

Figure (b) and (c) illustrate an x-gradient filter, and y-gradient

filter, respectively.

221

The x-gradient filter determines the slope increase ratio

from west to east.
If the elevation to the west of the centre cell is for
example 1540m and that to the east of the centre cell is
1552m, then apparently along this transect the elevation
increases 12m per two cell widths, i.e. the x-gradient is
6m per cell width.
The y-gradient filter operates entirely analogously,
though in south-north direction.

222

111
 1/30/2018

X- and Y-gradient Calculation

223

Slope Angle & Aspect Angle Ψ

Observe that both filters express elevation gain

per cell width. 224

112
 1/30/2018

Here, f stands for the elevation field as a

function of x and y, and δf/δx, for instance, is
the elevation gain per unit of length in the x-
direction.

225

To obtain the real slope angle α along path p,

observe that both the x- and y-gradient contribute
to it.
This is illustrated in Figure above.
A, not-so-simple, geometric derivation can show
that always

226

113
 1/30/2018

Slope Angle Calculation

227

Slope Angles (example)

228

114
 1/30/2018

Slope Aspect Calculation

229

Summary
From Points to Fields
1. Discrete Fields (Vector!)
2. Continuous Fields (Raster!)

Interpolation
1. Thiessen Polygons
2. Surface fitting, MWA, IDW, Delaunay

Operations on Rasters (compulsory!!)

• Filtering
• Slope & Aspect Computation

230

115
 1/30/2018

CHAPTER 6
Spatial data analysis

231

Spatial Data Analysis Part I

The solution of a spatial problem

depends on a large number of
parameters.
Parameters are often interrelated.
Interaction of parameters is more
precisely described in an
application model.
232

116
 1/30/2018

Introduction

Application model: description of the behavior of relevant

geographic phenomena, it tries to simulate the geographic
phenomena that are present.

Models for planning and site selection are prescriptive

They quantify environmental, economic and social factors

233

Introduction

In predictive models, a forecast is

made of the likelihood of future
events, for example, of a future
landslide.
This lesson however does not
discuss analytical modeling, it
discusses analytical functions that
form the building blocks for analytical
models. 234

117
 1/30/2018

Introduction

• There are many ways to classify analytical functions.

• The one suggested by Aronoff, (1989) is used here:
• Measurements, retrieval and classification
• Overlay functions
• Neighborhood functions
• Connectivity functions
• These functions are split over 2 lectures.
• Every function is discussed for vector and for raster data.

235

Overview of this lesson:

1. Measurement operations

1.1. Vector measurements

1.2. Raster measurements
2. Spatial selection queries

2.1 Spatial Selection by Attribute conditions

2.2 Spatial Selection using topological relationships
3. Reclassification

3.1 User controlled classification

3.2 Automatic classification
4. Overlay techniques
236

118
 1/30/2018

Measurements

Only Geometric Measurements are discussed, no

measurements on attribute values.
All measurement functions are discussed both in
raster and vector (on points, lines and polygons).
Vector measurements include: location, length,
distance and area size.
237

Measurements -Vector

Vector measurements include: location,

length, distance and area size

Location → always stored by GIS

• One coordinate pair for points
• List of pairs for lines and polygons

238

119
 1/30/2018

Measurements -Vector

• Polygons have a centroid.

• The centroid is the center of mass of the
polygon, the point on which it would balance
under a homogeneously applied force like
gravity.
• Note that the centroid can be outside the
polygon
• The location of the centroid of a polygon can
be either stored or computed on the fly

239

Measurements -vector

Length is associated with

polylines, and with polygon
boundaries.
It can be stored by the GIS or
computed on the fly.

240

120
 1/30/2018

Measurement -Vector

241

Measurement -Vector

If one or both features are not

a point, we will measure the
minimal distance between the
two features.

242

121
 1/30/2018

Measurement -Vector

A bounding box is the minimal

rectangle that covers the feature.
Bounding box computation is a
Support function.

Example: when the bounding boxes do

not intersect, the features do not
intersect.

243

Measurement -Raster

Raster measurements include: location,

distance and area size
Location of an individual cell →derived
from anchor point and resolution
The cell’s location can be its lower left
corner or midpoint

244

122
 1/30/2018

Measurement -Raster

Area size →number of cells * cell size

When you know the resolution, you can calculate
the area of a single cell. In this example 30 x 30
meters = 900 m2.
The number of cells is also called the frequency
or count.

245

Measurement -Raster

Distance →standard distance function

applied to the locations of their mid-
points.
When a raster is used to represent line
features as strings of cells, the length
of a line is computed as the sum of the
distances between the cells.

246

123
 1/30/2018

Measurement -Summary
Vector:
• Vector measurements include: location (coordinates and
centroids), length, distance and area size.
• Another geometric measurement is the minimal bounding box
of lines and polygons.
• Pythagorean distance function is used.

Raster:
• Raster measurements include: location, distance and area
size.
• Standard Pythagorean distance function applied to the
locations of their mid-points.
247

Spatial selection queries

Interactive
Spatial Selection by Attribute conditions
• Relational operators
• Logical operators
• Combining attribute conditions

Spatial selection using topological relationships

• Selecting features that are inside selection objects
• Selecting features that intersect
• Selecting features adjacent to selection objects
• Selecting features based on their distance 248

124
 1/30/2018

Spatial selection queries

Interactive spatial selection is a

selection in which you select
features by clicking on the screen
(on the feature to select) or
drawing a graphic, to select all
objects within this graphic.

249

Spatial Selection by Attribute conditions

Attribute queries are selections in which

we use a selection condition on the
features attributes.
The condition is specified in a query
language.
This query language can be SQL when the
data is stored in a relational database or in
a software specific language.

250

125
 1/30/2018

Spatial Selection by Attribute conditions

A condition that tests a single criterion is

called an atomic condition.
Atomic conditions use a predictive symbol
such as < (less than).
Any of these symbols is combined with an
expression on the left and on the right.

251

Spatial Selection by Attribute conditions

The expression can be an attribute

name (Area), a constant (400000), or
an arithmetic expression:
Ranking > MinRanking + 5
Area / Length < 1

252

126
 1/30/2018

Spatial Selection by Attribute conditions

Logical connectives are used to

combine two atomic conditions to one
composite condition.
Examples of logical connectives are:
AND, OR and NOT.

253

Spatial Selection by Attribute conditions

OR returns true if one or both of

the expressions a and b is true.
NOT returns true if the expression
is false.

254

127
 1/30/2018

Spatial Selection using topological relationships

Spatial selections use topological

relationships

The steps carried out are:

1. Select one or more selection objects.
2. Apply a chosen spatial relationship to
determine the features that have that
relationship with the selection objects.

255

Spatial Selection using topological relationships

Selecting features that are inside

selection objects.
Polygons can contain polygons, lines or
points, and lines can contain lines or
points.

256

128
 1/30/2018

Spatial Selection using topological relationships

An example of a selection using the inside

relationship is shown on the left.
In picture 1 the state of Georgia is
selected, this is the selection object.
In picture 2 all the cities that are inside this
state are selected.

257

Spatial Selection using topological relationships

Selecting features that intersect.

In picture 1, the state of Georgia is
selected.
In picture 2, the interstates that run
through Georgia (intersect).

258

129
 1/30/2018

Spatial Selection using topological relationships

Selecting features adjacent to selection

objects.
Adjacency is the same as the meets
relationship.

259

Spatial Selection using topological relationships

Selecting features based on their distance.

Will search within a given distance from
the selection object, at a given distance, or
beyond a given distance.
Example (left) select all the cities within
100 kilometers from Atlanta.

260

130
 1/30/2018

Spatial Selection -Summary

Two types of selections, attribute selections and

selections using topological (spatial) relationships
In attribute queries, predictive symbols and Logical
connectives are used.
We distinguish atomic conditions and composite
condition.
261

(Re) Classification

Remove detail from an input dataset to reveal important spatial

patterns.
Reduce the number of classes and eliminate details.
If the input dataset itself is the result of a classification we call it a
reclassification.
Spatial selection is a two step process:
1. Select one or more selection objects.
2. Apply a chosen spatial relationship to determine the features that
have that relationship with the selection objects.

262

131
 1/30/2018

(Re) Classification –post processing

Vector reclassification can be performed with (or without) post-

processing.

With post-processing we mean that adjacent features that after

classification have the same class are merged together.

This is called spatial merging, aggregation or dissolving.

263

(Re) Classification –post processing

For raster datasets, post-processing is not

possible, only the value of the cell will
change.
Post-processing of vector data is only
sensible for lines and polygons.

264

132
 1/30/2018

(Re) Classification

3.1 User controlled classification

• Classification table
3.2 Automatic classification
• Equal interval technique
• Equal frequency technique

265

User controlled classification

In user-controlled classification we
indicate the classification attribute and
the classification method.
This is normally done via a
classification table.

266

133
 1/30/2018

Automatic classification

• User specifies the no. of output

classes.
• Computer decides the class break
points.
• Two techniques of determining the
class breaks are discussed:
• Equal interval
• Equal frequency
267

Automatic classification

• Equal interval is calculated as (Vmax–

Vmin) /n
• Vmax is the maximum attribute value, Vmin
is the minimum attribute value and n is
the number of classes.
• In our example: (10-1) / 5 ≈2
• Each class will have two values.

268

134
 1/30/2018

Automatic classification

Equal frequency, is also called

quantile.
Total number of features / number of
classes (n)
The objectives is to create
categories with roughly equal
number of features (or cells).
269

Automatic classification
The two techniques are illustrated here
again. On the horizontal axis, the original
values.
The classes are the blue vertical lines. In the
equal interval example, the blue lines are
equally spaced. On the vertical axis, the
frequency.
In the equal frequency technique, when the
frequency goes up, the classes become
smaller.
270

135
 1/30/2018

(Re) Classification -Summary

• Reclassification is used to reduce the complexity of a layer in

order to show patterns
• Two types of classifications, user controlled classification and
automatic classification
• In user controlled classification the user will create a
classification table.
• In automatic classification the user will determine the number
of classes and the classification technique.
• Two classification techniques were discussed:
• Equal interval
• Equal frequency
271

Overlay
Standard overlay operators take two input
data layers and assume they are geo-
referenced in the same system and
overlap in study area. If either condition is
not met, the use of an overlay operator is
senseless.
The principle of spatial overlay is to
compare the characteristics of the same
location in both data layers, and to
produce a new output value for each
location.
272

136
 1/30/2018

Overlay

Overview of the topics discussed:

4.1 Vector overlay techniques
• Intersection
• Clip by
• Overwrite by
4.2 Raster overlay techniques
• Arithmetic operators
• Comparison and logical operators
• Conditional expressions
• Decision table
273

Vector overlay -intersection

The standard operator for two layers of

polygons is the polygon intersection
operator.
The result of this operator is the collection of
all possible polygon intersections.
The attribute table combines the information
of the two input tables ( spatial join).

274

137
 1/30/2018

Vector Overlay Clip

Clip takes a polygon data layer and restricts its

spatial extent (the area that it covers) to the
outer boundary of a second input layer (clip
layer).
No other polygons from the clip layer play a role
in the result.
This technique can be used to reduce the area of
a thematic layer to that of the study area.
275

Vector Overlay Overwrite

The polygon overwrite creates a layer with the

polygons of the first layer except where polygons
exist in the second layer (as they take priority).
This operator can be used to overwrite a layer
with “updates“ stored in a second layer.

276

138
 1/30/2018

Vector Overlay

Vector overlays are usually also defined

for point and line data layers.
When a polygon layer is intersected by a
line layer the result will be a line layer (the
layer of the lowest order).

277

Raster overlay operators

Vector overlay operators are geometrically

complicated, raster overlays are mostly cell by
cell computations.
GISs that support raster processing have a full
language to express operations. This is called a
raster calculus.
The syntax used here is based on the ILWIS
raster calculus.
278

139
 1/30/2018

Raster overlay operators

New cell values are calculated using calculus -map algebra.
• Performed on cell-by-cell basis.
• No geometric calculation

Output_raster_name : = Raster_calculus_expression

279

Raster overlay operators

Overview:
• Arithmetic overlay operators
• Comparison and logical operators
• Conditional expressions
• Decision table

280

140
 1/30/2018

Raster overlay -arithmetic

Arithmetic operators
• +, -, *, /
• MOD (modulo division)
• DIV (integer division)
• Goniometric operaters: sin, cos, tan, asin, acos, atan.
• For example:
Raster2 : = Raster1 * 5

281

Raster overlay –comparison/logical

Comparison operators
• <, <=, =, >=, >, <>
C:=A<>B is true when the cell’s value in A
differs from the cell’s value in B. It is false
if they are the same.

Logical operators
• AND,OR,NOT, XOR (exclusive)
a XOR b is true if either a or b is true, but not
both.
282

141
 1/30/2018

Raster overlay – comparison/logical

The cell values of the output raster is

either true or false.
Other values are not possible.

283

Raster overlay –conditional expr.

Conditional expressions test a condition, if the

expression is true, the then_expression will be
evaluated, if it is false the else_expression.
A conditional expression can lead to many
different types of outputs, numeric value, but also
true/false.
Format: output_raster: = IFF (condition,
then_expression, else_expression).
284

142
 1/30/2018

Raster overlay – decision table

Decision tables are the same statement

as in a Conditional statement, but
presented in a different way.
• The decision table will guide the
overlay process.
• It lists all possible combinations of input
values, and the output values.
285

Overlay operations -RASTER

286

143
 1/30/2018

Overlay - Summary

Vector overlay techniques, intersection, clip by and overwrite by.

• Intersection is the fundamental operator, the attribute table is a
spatial join (fields from both input tables.
• Clip by is like a cookie cutter, cutting out the map extend of the
second layer.
• For types of raster overlay techniques, using arithmetic operators,
comparison and logical operators, conditional expressions and
decision table.
• Comparison and logical operators only evaluate to true and false.
• Conditional expressions and decision table lead to the same
result
287

Spatial Data Analysis Part II

Three topics will be discussed:

1. Neighborhood functions
2. Network analysis
3. Error propagation

288

144
 1/30/2018

Neighborhood functions

Will find out characteristics of the

vicinity (neighborhood) of a location.

289

Neighborhood functions

To perform neighborhood analysis, we

must state which target locations are of
interest to us, and what is their spatial
extent.
Define how to determine the neighborhood
for each target.
Define which characteristic(s) must be
computed for each neighborhood.
290

145
 1/30/2018

Neighborhood functions

Proximity computation makes use of the geometric

distance function.

Spread computation assumes that the phenomenon

spreads in all directions, but not necessarily equally
easily in all direction.
In seek computation the phenomenon will choose a
least-resistance path. 291

Neighborhood functions

➢ Buffer generation (Vector and Raster)

➢ Thiessen polygons (Vector)
➢ Spread computation (Raster)
➢ Seek computation (Raster)

292

146
 1/30/2018

Buffer zone generation

Principle is simple, we select one or more target

locations and determine the area around them.
• Buffer generation can be performed on vector
as well as raster data.

• Target locations can be point, lines or

polygons in a vector environment.
293

Buffer zone generation

• Buffers can be simple, or zonated.

• With a zonated buffer, the buffer
consists of multiple rings each
representing a different distance.
• In vector buffer generation, the buffer
will become a new polygon in the
output layer.
294

147
 1/30/2018

Buffer zone generation

Raster layers:
• Need target cell(s).
• The distance function applies the Pythagorean
distance between the cell centers.
• Using cell resolution as the unit.
• The distance from a non-target cell to the
target is the minimal distance one can find
between that non-target cell and any target
cell.
295

Thiessen polygons

Divide an area into polygons, so that each

polygon contains locations that are closer to
the midpoint than to any other midpoint.
It will generate a polygon around each target
location that identifies all those locations that
‘belong to’ that target.

296

148
 1/30/2018

Thiessen polygons

The easiest way to construct Thiessen

polygons is via a Delaunay triangulation of
the same input points.
The construction first creates the
perpendiculars of all the triangle sides.

297

Spread Computation

In spread computation the neighborhood of

a target location not only depends on
distance, but also on direction and
differences in the terrain.
Examples are pollutions and radio waves.
298

149
 1/30/2018

Spread Computation

Spread computation involves one or more target locations, which

are better called source location in this context, as they may be for
example the pollution source.

Spread computation also involves a local resistance raster, which

for each cell provides a value that indicates how difficult it is to
pass by that cell.

299

Spread Computation

While computing total resistance, the GIS

takes proper care of correct spread path
lengths.
The spread from a cell to its neighbor cell to
the east is shorter than to its northeast
neighbor

300

150
 1/30/2018

Spread Computation

The GIS computes the total resistance for

spreading from csrc to cn as ½(val (csrc) +
val(cn))
This is half of the resistance value of the
source cell (csrc), because this cell is only
traveled half, plus half of the resistance
value of the cell north of it
The value of the source cell of course is 0.
301

Spread Computation

The GIS computes the total minimal

resistance raster for a diagonal neighbor
as ½(val(csrc) + val(Cne)) * √2.
This is half of the resistance value of the
source cell plus half of the resistance
value of the cell to the north-east
multiplied by the square root of 2.

302

151
 1/30/2018

Spread Computation

Since the source material has the habit of

taking the easiest route to spread, we must
determine the minimal cost.
We must consider all possible paths to
reach the cell and assign the minimal
value.

Lowest Value

303

Spread Computation

Note that the accumulated resistance along

a path of cells is simply the sum of these
incurred resistances.
We can use the value 4.00 and add the
resistance for moving from this cell to its
north east neighbor cell.

304

152
 1/30/2018

Seek Computation

Seek computation applies when a

phenomenon does not spread in all
directions, but chooses a least-cost
path.
A typical example is a drainage
pattern in a catchment.

305

Seek Computation

Input for a seek computation is an elevation

raster.
For each cell, the steepest downward slope
to a neighbor cell is determined.
The direction of this downward slope is
stored in the flow direction raster.

306

153
 1/30/2018

Seek Computation

1. For each cell, first eliminate all cells

that are not downhill (have a higher
elevation value)
2. You have two types of neighbors: direct
neighbors and diagonal ones. For each
type, pick the steepest.
3. Compensate for the difference in path
length.
307

Seek Computation

From the flow direction raster, the GIS will

calculate the accumulated flow count
raster.
Cells with a high accumulated flow count
represent streams.

308

154
 1/30/2018

Seek Computation

The value of the accumulated flow for

each cell is the number of cells that
flow into this particular cell.
Cells with a value 0 have no other cells
flowing into them, and represent higher
areas.
309

Summary –proximity Computation

310

155
 1/30/2018

Network analysis

2.1 Types of networks

• Directed network and undirected network
• Planar network and non-planar network
• Two types of analysis
• 2.2 Optimal path finding
• Ordered
• Unordered
• 2.3 Network partitioning
• Network allocation
• Trace analysis
• 2.4 Turntables
311

Types of Networks

Network is a set of connecting lines

Network can represent rivers, roads,
pipelines, telecommunication lines etc.
Network analysis analyze the way
‘goods’ can be transported along these
lines.
Network analysis can be done in raster
or in vector.
312

156
 1/30/2018

Types of Networks

Networks can be directed, where

transportation is only in one
direction, for example rivers, or it can
be undirected, the goods can be
transported in both directions
(roads).
313

Types of Networks

Networks can be planar, this means they

are 2-dimensional.
Planar networks do not have overpasses
or underpasses.
Example of a planar networks are rivers.

314

157
 1/30/2018

Types of Networks

Non-planar networks have multi-level

crossings, underpasses and overpasses.
When they are modeled in 2D these
overpasses and underpasses should be
modeled in a special way.
Example of non-planar networks are roads.

315

Analysis on networks

Optimal path finding

Which generates a least cost-path on a network between a pair of

predefined locations using both geometric and attribute data.

Network partitioning

Which assigns network elements (nodes or line segments) to

different location using predefined criteria.
• Network allocation
• Trace
316

158
 1/30/2018

Optimal path finding

Optimal path finding is used when a least cost

path between two nodes in a network must be
found.
You need a cost function!
• Also called Impedance.
• One of the attributes in the feature
attribute table.
• Length, travel time, etc.
• The least-cost path is the one that has
the min. value of the total cost between
two nodes
317

Optimal path finding

Cost factors
• The cost can be defined on both lines
and nodes.
• For lines, the cost can be the same or
different along and against the line
direction.
• The cost on nodes is used to define the
turns.
318

159
 1/30/2018

Optimal path finding

Two costs, one for each driving

direction can be applied in rush
hour.
In the morning it takes much longer
to go into the city than it takes to
drive in the opposite direction.

319

Optimal path finding

Cost can be associated with line

segments, but also with nodes.
Passing a traffic light for example can
take a considerable amount of time.

320

160
 1/30/2018

Optimal path finding

Ordered optimal path finding: the

sequence in which the places have to
be visited matters.
Unordered optimal path finding: the
sequence does not matter.

321

Network partitioning

In network partitioning, the purpose is to assign lines and

nodes (parts of the network) to a number of target
locations (for example which part of the network belongs
to a hospital, school or fire station).

There are two types of network partitioning problems:

• 2.3.1 network allocation
• 2.3.2 trace analysis
322

161
 1/30/2018

Network Allocation

We have a number of resource

centers, and the problem is which part
of the network can be assigned to
which service center.
In principle this is a simple problem,
each line segment is assigned to the
service center that is the nearest.
323

Network allocation

Problems associated with network

allocation are:
• The capacity with which a centre
can produce the resources
• The consumption of the resources
which may vary amongst lines or
line segments.
324

162
 1/30/2018

Trace Analysis

Trace analysis is performed when

we want to understand which part of
a network is connected to the trace
origin.
A condition can be applied for
example, trace only in the direction
of the origin (upstream).
325

Summary –Network Analysis

326

163
 1/30/2018

Error propagation

How errors propagate:

• Errors already present in the input data will
propagate through the manipulations.
• New errors arise from the computer processing
(analytical operations performed).

327

Error propagation

Error propagation analysis:

• Testing the accuracy of each state by measurement
against the real world.
• Modeling error propagation, either analytically or by
means of simulation techniques.

328

164
 1/30/2018

Error propagation

Initially the complexity of spatial data led to the

development of mathematical models describing only
the propagation of attribute errors.
Modern models incorporate both spatial and attribute
errors.

329

CHAPTER 7
Data Visualization

330

165
 1/30/2018

Contents
GIS and maps
The visualization process
Visualization and strategies
The cartographic ‘toolbox’
• data characteristics,
• representation of these characteristics
Examples, how to map:
•qualitative data,
•quantitative data, (terrain or statistical) elevation, time

Map cosmetics: the finishing touch

Map dissemination 331

GIS and Maps

In a GIS environment, maps can be used

as:
•Input for GIS
•Communicate GIS results
•Support spatial analysis
Maps are not only final products (output)!

332

166
 1/30/2018

Map Characteristics

Main characteristics of maps

Maps provide answers (in graphical
form) to questions related to the 3
basic components of geographic
data
• geographic location (Where?)
Where did the cartography
students come from?’ • thematic attributes (What?)
• time (When?)
333

Map Characteristics

reality more abstract

Main characteristics of maps!!!!!
Maps offer abstract representations (models) of reality,
that are:

• simplified
map. • classified
• symbolized

334

167
 1/30/2018

Scale

Main characteristics of maps

Maps are representations at
scale
Scale: ratio between distance on
the map and corresponding
distance in reality.
335

Scale

Scale indications
• verbal e.g. one-inch-to-the mile
• representative fraction e.g. 1 : 100 000
• graphical (scale bar) suitable in digital
environments!

Digital Graphical Scale

Advantage of scale bar in digital environment is that

its length changes when zooming in or out.
336

168
 1/30/2018

Definition of a Map

A map is:
• a representation or abstraction of
geographic reality; a tool for
representing geographic information in a
way that is visual, digital.
• a reduced and simplified representation
of (parts of) the Earth’s surface on a
plane. 337

Types of Maps

Traditional distinction in maps

• topographic maps accurate representation of
the Earth’s topography.
• thematic maps one or more particular themes
are emphasized.

Less relevant distinction in a digital environment

338

169
 1/30/2018

Map Dimensions

You can distinguish types of maps based

on the number of dimensions used for the
representation.

• Flat
• Flat + Height
• Flat + Height + Time

339

The visualization process

Maps are the result of a visualization process.

340

170
 1/30/2018

Cartographic Tools

Visualization methods and techniques are

applied using cartographic ‘tools’:

• Functions (e.g. algorithms)

• Rules (e.g. generalization, cartographic
grammar)
• Habits or conventions (e.g. water is
represented in blue).

An algorithm can be used to smooth lines and improve the

appearance of features. 341

Visualization and strategies

‘Visualization’ has several meanings:

• generic: to make info visible (presentation in graphical form)
• more specific: to use sophisticated computer technology and
‘toolboxes’ to make data/ info visible for specific use:
visual exploration

this process is often called:

scientific visualization: meant to stimulate thinking
keywords: interaction, dynamics
342

171
 1/30/2018

Geo-visualization

Two main strategies of visualization:

exploration and presentation.

• private visual thinking: involves an

individual playing with the spatial data to
determine its significance.
• public visual communication: concerns
maps aimed at a wide audience.

If maps are visually explored, we also

talk about geo-visualization.
343

Geo-visualization

Geo-visualization is accelerated by:

• The possibility to generate maps at any stage in Geo-

information processing.
• Hard- and software developments
• New output media
• Changing needs / expectations of users.
• Availability of abundant data, from different sources.

344

172
 1/30/2018

Cartographic Visualization

The cartographic communication

process, based on “How do I say
what to whom, and is it effective?”
Information loss or gain:
Information derived by the map user
is not the same as the information
that the cartographic communication
process started with.

345

The Cartographic ‘Toolbox’

Analysis of the characteristics of data

What is the common ‘denominator’? Used

for the title of the map (theme, area, year)
What is the nature of the data or What are
the measurement scales ?
The common denominator refers to a common label for all the
attributes/attribute values that are mapped (here: geological units).
346

173
 1/30/2018

The Cartographic ‘Toolbox’

Measurement scales are linked to the way in which

people perceive visual variables (see further).

347

Nominal data exists as discrete, named values without a

natural order amongst the values.
Examples are the different soil types (e.g. sand, clay,
peat) or the different land use categories (e.g. arable
land, pasture).
Ordinal data are measured along a relative scale, based
on hierarchies. For instance, one knows that one value is
’more’ than another value, such as ’warm’ versus ’cool’.
Another example is a hierarchy of road types: highway’,
’main road’, ’secondary road’ and ’track’.
348

174
 1/30/2018

Quantitative data can be measured, either along an

interval or ratio scale.
For data measured on an interval scale, the exact
distance between values is known, but there exists no
absolute zero on the scale. Temperature is an example:
40oC is not twice as warm as 20oC, and 0oC is not an
absolute zero.
Quantitative data with a ratio scale have a known
absolute zero.
An example is income: someone earning 1000shs earns
twice as much as someone with an income of 500shs.
349

The Cartographic ‘Toolbox’

Basic elements of a map:

• Point symbols
• Line symbols
• Area symbols
• Text
These elements can all be varied in
appearance
350

175
 1/30/2018

Bertin’s Visual Variables

Bertin’s visual variables: an elementary

way in which, point, line and area
symbols can be graphically varied.
• Size
• Color
• Value (lightness)
• Grain/ texture
• Orientation
• Form/shape

351

Bertin’s Visual Variables

The visual variables enable observers to perceive:
• What belongs together, or is of equal importance (e.g. all red
symbols represent danger).
• Order
(e.g. the population density varies from low to high, represented
by light and dark color tints, respectively).
• Quantities
(e.g. symbols changing in size with small symbols for small
amounts).
• An instant overview of the whole representation.
352

176
 1/30/2018

Measurement Scales and Visual Variables

353

How to Map Qualitative Data

What is the common denominator of the

data? Watersheds
What is the nature of the data? qualitative
(nominal)
Solution: Colors of equal visual weight or
brightness which allow the user to quickly
differentiate between watersheds.

354

177
 1/30/2018

How not to Map Qualitative Data

355

How to Map Absolute Quantitative Data

What is the common denominator of

the data? number of inhabitants.

What is the nature of the data?

absolute quantitative (ratio)

Solution: symbols varying in size

356

178
 1/30/2018

How not to Map Absolute Quantitative Data

357

How to Map Relative Quantitative Data

What is the common denominator of the

data? Number of inhabitants/ km2.
What is the nature of the data? relative
quantitative (ratio).
Solution: Value has been used to display
the density from low (light tints) to high
(dark tints).

358

179
 1/30/2018

How not to Map Relative Quantitative Data

359

Test Yourself

360

180
 1/30/2018

Map Dissemination (Output)

Map design is influenced by:
• Data characteristics & user strategies (see before)
• Output medium, e.g.: paper or screen maps

Screen maps:
• Often smaller

Legend is not always visible and they enable:

• Access to a data base
• Links to other data
• Embedding in multi media
• Dynamics and interaction 361

How to Map the Terrain elevation

362

181
 1/30/2018

How to Map the Thematic Data in 3D

Statistical (socio-economic) data can

also be represented as an elevated
surface.
Here, the municipalities in the
province of Overijssel are elevated
proportionally to their number of
inhabitants.
The resulting map is called a 'prism
map' 363

How to Map Time Series

Single static map: specific graphic

variables and symbols are used to indicate
change or represent an event
Series of Static Maps: A single map in the
series represents a ‘snapshot’ in time.
Together, the maps depict a process of
change.
Animated map: Change is perceived to
happen in a single image by displaying
several snapshots after each other, just like
a video. 364

182
 1/30/2018

Map Cosmetics: The Finishing Touch

1. Additional information
marginal info (or metadata): makes
the map more usable.
2. Adding text improves the
identification of features.
3. Contrast improves overall map
legibility.

365

Web Maps

The Web as output medium increases the functions of maps:

• next to: insight in spatial data (traditional role)

• interface to additional information / services
• previews of data that can be acquired

Classification of maps on the Web

366

183
 1/30/2018

Summary

Role of map in a GIS environment: more than just output

Basics: how to translate characteristics of spatial data into symbols on a
map, with particular attention to:
• analysis of the measurement scale of the data,
• measurement scales can be linked to perception properties of
visual variables
• selection of those variables that best translate the nature of
the data.
After translation of the data, the map has to be finalized to make it usable
(taking care of visual contrast, adding text, marginal information)
Finally the map is disseminated to users, e.g. via the Web.
367

References
Arbia, G., Griffith, D., and Haining, R. Error propagation modeling in raster gis: overlay operations. International
Journal of Geographical Information Science 12, 2 (1998), 145–167.
Aronoff, S. Geographic Information Systems: A Management Perspective. WDL Publications, Ottawa, Canada,
1989.
Bertin, J. S´emiology Graphique. Mouton, Den Haag, The Netherlands, 1967. BIJKER, W. Radar for rain forest—A
monitoring system for land cover change in the Colombian Amazon. PhD thesis, International Institute for
Aerospace Survey and Earth Sciences, Enschede, The Netherlands, 1997.
Board, C. Report of the working group on cartographic definitions. Cartographic Journal 29, 1 (1990), 65–69.
Bonham-Carter, G. F. Geographic information systems for geoscientists : Modeling with GIS, vol. 13 of Computer
methods in the geosciences. Pergamon, Kidlington, U.K., 1994.
Burrough, P. A. Natural objects with indeterminate boundaries. In Geographic objects with indeterminate

END OF COURSE
boundaries, P. A. Burrough and A. U. Frank, Eds. Taylor and Francis, London, U.K., 1996, pp. 3–28.
Chrisman, N. R. Errors in categorical maps: testing versus simulation. In Proceedings AutoCarto (1989), pp. 521–
529.

Good Luck
Clarke, D. G., and Clark, M. Lineage. In Elements of Spatial Data Quality, S. C. Guptill and J. L. Morrison, Eds.
Elsevier Science, Oxford, U.K., 1995, pp. 13–30.
Date, C. J. An Introduction to Database Systems, seventh ed. Addison- Wesley, Reading, Ma, 2000.
Dibiase, D. Visualization in earth sciences. Earth and Mineral Sciences, Bulletin of the College of Earth and Mineral
Sciences 59, 2 (1990), 13–18.
ElmasrI, R., AND Navathe, S. B. Fundamentals of Database Systems, second ed. Benjamin/Cummings, Redwood
City, Ca, 1994.
368

184
 1/30/2018

Heuvelink, G. B. M. Error propagation in quantitative spatial modeling- Applications in Geographical Information

Systems. Nederlandse Geografische Studies. Koninklijk Aardrijkskundig Genootschap, Utrecht, 1993.
Iliffe, J. Datums and Map Projections for Remote Sensing, GIS and Surveying. Whittles Publishing, CRC Press,
2000.
Intergovernmental Committee on Surveying and Mapping (ICSM). Get in step with the geocentric datum. Tech. rep.,
Office of the Surveyor General, Land Information New Zealand (LINZ), 1999.
Janssen, L. L. F. Principles of Remote Sensing, vol. 2 of ITC Educational Textbook Series. International Institute for
Aerospace Survey and Earth Sciences, Enschede, The Netherlands, 2000.
Kerle, N., Janssen, L. L. F., AND Huurneman, G. C. Principles of Remote Sensing, third ed., vol. 2 of ITC
Educational Textbook Series. International Institute for Geo-Information Science and Earth Observation, Enschede,
The Netherlands, 2004.
KiiverI, H. T. Assessing, representing and transmitting positional accuracy in maps. International Journal of
Geographical Information Systems 11, 1 (1997), 33–52.
Knuth, D. E. The TEXbook. Addison-Wesley, Reading, Ma, 1984.
Kraak, M.-J., and Brown, A., Eds. Web cartography, developments and prospects. Taylor & Francis, London, U.K.,
2000.
Kraak, M.-J., and Ormeling, F. J. Cartography: Visualization of Spatial Data. Addison-Wesley Longman, London,
U.K., 1996.
Land Information New Zealand (LINZ). Strategic business plan. Tech. rep., Land Information New Zealand (LINZ),
October 1999.
LaurinI, R., and Thompson, D. Fundamentals of Spatial Information Systems, vol. 37 of The APIC Series. Academic
Press, London, U.K., 1992.
Leunen, M.-C. V. A Handbook for Scholars, revised ed. Oxford University Press, New York, NY, 1992.
369

Longley, P. A., Goodchild, M. F., Maguire, D. M., and Rhind, D. W., Eds. Geographical Information Systems:
Principles, Techniques, Management, and Applications, second ed., vol. 1. John Wiley & Sons, New York, N.Y.,
1999.
Mccormick, B., Defanti, T. A., and (eds.), M. D. B. Visualization in scientific computing. ACM SIGGRAPH Computer
Graphics—Special issue 21, 6 (1987).
National mapping Division, U. S. Geological Survey. Spatial data transfer standard. Tech. rep., U. S. Department of
the Interior, 1990.
Office of the Surveyor General of Land Information New Zealand. A proposal for geodetic datum development. Tech.
Rep. OSG TR2.1, Land Information New Zealand (LINZ), 1998.
Openshaw, S., Charlton, M., and Carver, S. Error propagation: a Monte Carlo simulation. In Handling geographical
information: methodology and potential applications, I. Masser and M. Blakemore, Eds. Longman, Harlow, U.K.,
1991, pp. 78–101.
Preparata, F. P., and Shamos, M. I. Computational Geometry—An Introduction. Springer-Verlag, New York, NY,
1985.
Robinson, A. H., Morrison, J. L., MuehrckE, P. C., Kimerling, A. J., and Guptill, S. C. Elements of Cartography, sixth
ed. John Wiley & Sons, New York, N.Y., 1995.
Samet, H. The Design and Analysis of Spatial Data Structures. Addison- Wesley, Reading, Ma, 1990.
Strunk Jr., W., and White, E. B. The Elements of Style, third ed. MacMillan Publishing Company, New York, NY,
1979.
Veregin, H. Developing and testing of an error propagation model for GIS overlay operations. International Journal
of Geographical Information Systems 9, 6 (1995), 595–619.
Zadeh, L. A. Fuzzy sets. Information and Control 8 (1965), 338–353.

370

185