Geographic Information System

Chapter 4 – Spatial Analysis

TRƯƠNG QUỐC ĐỊNH
Analytical GIS Capabilities
▪ There are many ways to classify the analytic
functions of a GIS. By following distinctions
in function classes, we have:
▪Measurement, retrieval, and classification
functions
▪Overlay functions
▪Neighborhood functions
▪Connectivity functions
Retrieval, classification and
measurement
▪ Measurement
▪Measurements on vector data
▪Measurements on raster data
Retrieval, classification and
measurement
▪ Measurements on vector data
▪Location, Length, Area, Minimal Distance
▪Minimal Bounding Box
▪ determines the minimal rectangle—with sides parallel to the axes of
the spatial reference system—that covers the feature.
Measurement
▪ Length (Lines)
▪by Pythagorean theorem 1
D

D= (x2 − x1 )2 + ( y2 − y1 )2 2

▪Area (Polygons)
▪by dividing the polygon into
triangles whose areas can
easily be calculated
Retrieval, classification and
measurement
▪ Measurements on raster data
▪The geometric information stored with the raster
data is:
▪ Horizontal and vertical resolution, and the location of an
anchor point so all other measurements by the GIS are
computed.
▪ The anchor point is fixed by convention to be the lower
left (or sometimes upper left) location of the raster.
Spatial selection queries
▪ Spatial selection by attribute conditions
Spatial selection queries
▪ Spatial selection using topological relationships

Inside
Intersect
Adjacent
In distance with
Classification

1 urban or built-up
2 agricultural
3 rangeland 41 deciduous forest

4 forest 42 evergreen forest

... 43 mixed forest

Overlay functions
▪ These functions (operators) are as follows
▪polygon intersection
▪spatial join
▪polygon clipping
▪polygon overwrite
Overlay
▪ A series of registered
data layers ‘overlaying’
each other
Overlay
▪ Derived from manual
cartographic overlay
using Mylar sheets
(transparent plastic)
that were physically
overlaid on top of one
another.
Overlay
▪ Overlay is a GIS operation that integrates
data from two (or more) datasets
▪Visual overlay
▪ Datasets are visually combined
▪ No new datasets are produced
▪Computational overlay
▪ New datasets are created
▪ New topology is created based on the intersection of features of the
input datasets
▪ Computationally extensive and time consuming
▪ Input datasets remain intact
Overlay
▪ The technique is used to determine how features
of the input datasets are located in relation to
each other
▪Which hotels are within 200 m from a main road?
▪Which wells are within abandoned military bases?
▪Which schools are within the flood zone?
▪ Types of vector overlays
▪Point-in-polygon
▪Line-in-polygon
▪Polygon-on-polygon
Point in Polygon Overlay
Trees Land Cover NewTrees
1 2
A A
C + = C
B B

Point Table Poly Table Point Table

ID Tree ID Cover ID Tree Cover
A Elm 1 Rural A Elm Rural
B Maple 2 Urban B Maple Rural
C Elm C Elm Urban
Line in Polygon Overlay
Streets Land Cover NewStreets

C A C
A
+ = D
B B
1 2

Line Table Poly Table Line Table

ID Street ID Cover ID Street Cover
A Race 1 Rural A Race Rural
B Race 2 Urban B Race Urban
C Arch C Arch Urban
D Race Urban
Polygon Overlay

Intersection (and) Union (or) Identity

Polygon Overlay: Intersection
Agriculture Land Cover <Intermediate>

A
A

B B

ID Owner ID Cover Area of

A Brown A commercial intersection
B Smith B industrial New node
Polygon Overlay: Union
Agriculture Land Cover <Intermediate>

A
A

B B

ID Owner ID Cover Area of

A Brown A commercial union
B Smith B industrial New node
Polygon Overlay: Identity
Agriculture Land Cover
(input layer) (identity layer) <Intermediate>

A
A

B B

ID Owner ID Cover Area of

A Brown A commercial identity
B Smith B industrial New node
Analysis of Raster data
▪ Raster datasets represent the Earth’ surface as a
grid
▪Features are represented by collections of cells.
▪Each cell contains a value.
▪ Analytical operations applied to these datasets
are based on
▪Mathematical operations used on arrays of numbers
and matrices.
▪Logical relationships between the entities in the input
datasets.
Measurement Functions
▪ In a raster data model,
there is more than way
to measure a distance
between two points
▪Euclidian distance
▪Manhattan distance
▪Proximity distances
▪…
Reclassification (Queries)
▪ Reclassification of a raster dataset results in
creation of a new dataset
▪New (output) values are calculated from the old
(input) based on a reclassification rule.
▪Other reasons to reclassify a dataset
▪ Replacing values based on new information.
▪ Grouping values together to simplify the information conveyed by
the dataset.
▪ Reclassifying values in several raster datasets to a common scale.
▪ Setting specific values to “NoData” or setting “NoData” cells to a
value.
Proximity Analysis
▪ Filtering
▪Changes the value of a cell based on the values of
the neighboring cells.
▪ Buffering
▪Used to identify a zone of interest around an
entity, or set of entities.
▪Based on calculating proximity as defined in
terms of
▪ Geographic distance
▪ Cost unit distance
Filtering
▪ Filter is a group of
cells around a target
cell:
▪The size and shape of a
filter is set by the user.
▪The new value assigned
to the target cell is
based on the filtering
algorithm applied.
Buffering (Neighbourhood 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.
Buffering (Neighbourhood functions)
▪ Proximity computation
▪Buffer zone generation
▪Thiessen polygon generation
Buffer
▪ Definition of what is within/without a given
proximity
Line buffer

Point buffer

Polygon buffer
Doughnut Buffer
▪ e.g. within 10 meters but not within 5 meters

Buffer
polygon
5
10

‘Hole’
Variable Buffer
ID Dist A B
Original line
A 3
B 2 C
C 5

6 4

10

Buffer polygon
Proximity Analysis
▪ Thiessen polygons (otherwise
known as Voronoi polygons or
Voronoi diagrams), are an
essential method for the analysis
of proximity and neighborhood.
▪ Thiessen polygons are used to
allocate space to the nearest
point feature. It defines an area
around a point, where every
location is nearer to this point
than to all the others (2D).
Proximity Analysis
Construction of Thiessen polygons
▪ A Thiessen polygon encloses all the space which is closer
to the associated center than to any other point. It is
obvious, that the borders of Thiessen polygons are the
geometric places, which have the same distance to two
centers.
▪ In order to construct Thiessen polygons, all the points are
triangulated into a triangulated irregular network (TIN).
For each triangle edge, the perpendicular bisectors are
generated, which form the edges of the Thiessen
polygons.
▪ The perpendicular bisectors are constructed by drawing
circles with radius d around the corresponding points. The
vertices of the Thiessen polygon are at the location, at
which the bisectors intersect.
Proximity Analysis
Construction of Thiessen polygons
Table 12.3 Computing water use based on land-use area
Total Land Unit
Node
Node Land Use Use Demand Demand
Node Total
Area Type Area (l/day/ha (l/day)
(l/day)
(ha) (ha) )
J-1 6.88 Industrial 6.88 11,200 77,100 77,100
Industrial 1.38 11,200 15,500
J-2 7.69 Commercial 0.92 4,700 4,300 60,200
Residential 5.38 7,500 40,400

Commercial 1.31 4,700 6,100

J-3 7.69 Residential 5.15 7,500 38,600 44,800
Undeveloped 1.23 0 0

Industrial 0.17 11,200 1,900

Commercial 0.10 4,700 470
J-4 8.50 20,800
Residential 2.45 7,500 18,400
Undeveloped 5.78 0 0
Industrial 6.48 11,200 72,500
J-5 8.09 80,100
Commercial 1.62 4,700 7,600
Industrial 0.20 11,200 2,200
J-6 4.86 Commercial 1.36 4,700 6,400 33,400
Residential 3.30 7,500 24,800
Map Algebra: Queries
▪ Logical Operations
▪Greater Then Tool
▪ Which areas are 1,000 m above sea level?
Map Algebra: Overlays
▪ A family of procedures that use various rules to
combine values in one or more raster datasets to
produce an output dataset.
▪ These rules are operationalized via various
operators and functions
▪Arithmetic operators
▪Relational operators
▪Special operators
▪Conditional expressions
▪Logical functions
Map Algebra: Overlays
▪ Arithmetic Operations
▪Addition
▪ Which meteo stations are within the resort boundary?
▪ Functions as point-in-polygon vector overlay
▪ Raster Calculator
Map Algebra: Overlays
▪ Logical Operations
▪Boolean AND
▪ Which areas are above 2,400 m and federal land?
▪ Functions as INTERSECT polygon-on-polygon vector overlay
▪ If both input values are true (non-zero), the output value is 1. If one
or both inputs are false (zero), the output is 0
Map Algebra: Overlays
▪ Logical Operations
▪Combinatorial AND
▪ What soil types are found in suitable areas?
▪ Functions as INTERSECT polygon-on-polygon vector overlay
▪ If both input values are true (non-zero), the output is a different
value for each unique combination of input values. If one or both
inputs are false (zero), the output value is 0.
Map Algebra: Overlays
▪ Special Operations
▪ DIFF
▪ Finds the differences between two rasters.
▪ Useful for performing time-series studies, like finding those places
where land use has changed between 1980 and 1990.
▪ IN {list}
▪ Compares the input values to those in the provided list and, if there
is a match, returns the value to the output cell, or otherwise returns
NoData.
▪ Useful for “selecting” cells based on their values and writing them
to a new raster.
▪ OVER
▪ All non-zero values in the first input are "pasted "over the
corresponding cells in the second input.
▪ Useful for updating one raster with another.
Convex Hull
▪ A region (set of points) in the plane is
convex if every line segment between two
points in the region is also in the region.
▪ The convex hull of a finite set of points P is
the smallest convex region containing P.
▪ Theorem: The convex hull of a finite set of
points P is a convex polygon whose vertices
is a subset of P. The convex hull problem is
finding the convex hull given P.
Convex Hull
Euclidean Minimum spanning
tree
▪ Given N points in the plane, find MST
connecting them. Distances between point
pairs are Euclidean distances.
Visualization
▪ The four measurements and their definitions are:
▪ Nominal Qualitative measurements (name, type, state)
▪ Ordinal Quantitative measurements with a clear order, but
without a defined 0 value (small, medium, large)
▪ Interval Quantitative measurements with a defined
beginning point (temperature, height, distance)
▪ Ratio Quantitative measurements that provide a
relationship between two properties where the 0 value
indicates the absence of the relationship (particulates
mg/m3, time to cover a distance, dissolved oxygen in a
liter of water)
Thematic Map Representation
▪ Thematic Maps are those map that represent
geographic distribution of a specific
phenomena.
▪ Thematic maps show us the weather
forecast, election results, poverty, soil types,
the spread of a virus…
Qualitative data
▪ The application of color would be the best
solution since it has characteristics that allow
one to quickly differentiate between different
geographic units.
Quantitative data
▪ The final map should allow the user to
determine the value per feature and also offer
an overview of the geographic distribution of
the phenomenon.
▪ To reach this objective, the symbols used
should have quantitative perception
properties. Symbols varying in size fulfill
this demand.
Relative Quantitative data
▪ The numbers now have a clear relation with
the area they represent
Visualization of a Map with
Time Series
▪ Can be visualized by:
a. Single map, in which tints represent age of the
built-up area;
b. series of maps;
c. animation.
Map Cosmetics
Temporal GIS
▪ Spatio-temporal data
▪Time can be considered as fourth dimension.
▪From temporality point of view, there are two
types of information namely: Static and dynamic,
which must be modeled under a temporal GIS.
Static information
▪ Most of phenomena in the real world are
dynamic in nature.
▪ The term static can only be used for objects
such as cartographic maps, roads, facilities,
utilities, etc, that may not change in a short
period of time, however, all of these objects
will change during a long period of time.
Dynamic information
▪ Dynamic information refers to that information of
geospatial objects that change in a short period of
time. The length of this period is defined according to
its usage fields.
▪ According to working domain, one can consider the
following dynamic aspects of spatial information:
▪ Geometrical changes of features over time (such as urban
expansion)
▪ Positional changes of features over time (such as car
movement)
▪ Change of features attribute over time (such as traffic
volume)
▪ Any combination of the above changes
Dynamic information

▪ According to the length of duration of time,

one can divide dynamic information as
follow:
▪Real time data
▪Near real time data
▪Time stamped data
Real Time Data

▪ The term real time, about GIS refers to

capability of management, visualization and
analysis of graphical and attribute
information as soon as they input to GIS.
▪ The term real time, about data refers to
those kinds of data (such as traffic volume)
that are collected and imported to GIS as
soon as an events occurs.
Near real time data

▪ Real time is related to momentary updating,

visualization and analysis.
▪ Because of possible complex processes and
analysis that may be needed for such data
before they can be used in GIS, applying such
data is so difficult and even in some cases is
impossible.
▪ Therefore in these cases phrase near real time
may be more adoptive.
Time stamped data

▪ Time stamped data refers to such data that a

time represented with it in some ways (for
example as an attribute).
▪ There are different components of time that can
be considered about an event as follows:
▪When an event occurred in real world (valid time)
▪Occurrence duration of an event
▪When information about an event imported to GIS
▪When those data retrieved and manipulated
(transaction time)
Temporal visualization and
representation
▪ Some approaches that are offered for
representation of temporal information in a
temporal GIS are as follows:
▪ Animation
▪ 3D representation
▪ Multimedia representation
Temporal visualization and
representation
▪ Visualization of temporal information in a
temporal GIS can be considered from the
following aspects:
▪Representation of changes in a database (changes
in traffic volume)
▪Representation of changes in a geometrical data
(changes in roads network)
▪Representation of changes in a database as well as
in graphics (movement of vehicles)
Time stamping spatio-
temporal modeling in GIS
▪ Some approaches of spatio-temporal
modeling in GIS as follows:
▪ Snapshot model
▪ Space time composite model (STC)
▪ The event-based model
▪ The space-time cube model
Snapshot model
▪ In the snapshot model, when an event occurs, new
layer is constructed and occurrence time is stamped to
the layer (all of the information, changed or not
changed, are stored in the layer).
▪ Inconsistency and data redundancy are major
disadvantages of this approach.
▪ The snapshot model is the most common one in the
Earth sciences, as satellite imagery is such an
important base data source for them. After image
classification of several images of the same area, we
essentially have obtained a field-based snapshot
sequence that might function as a basis for study with
time-related questions.
Space time composite model
(STC)
▪ The space-time composite model also starts from a two-
dimensional view of the study space at a given start time.
Every change of an object that happens later is projected
onto the initial data layer and is intersected with the
existing features.
▪ This leads to successive intersections, thereby creating
an incrementally built, finer polygon mesh. Over time,
more and more polygons will be stored in the data layer.
Every polygon in this mesh has its attribute history stored
with it.
▪ This model can be useful if the amount of changes is
limited, and changes are discrete steps, as is the case for
instance in cadastral applications, where parcels may be
split or joined.
The event-based model
▪ In an event-based model, we start with an initial
state and record events along the time line. Whenever
a change occurs, an entry is recorded. This is a time-
based model. The spatial and thematic attribute
domains are secondary.
▪ Our event records on the event-based model are such
that we can reconstruct the full spatial and non-spatial
history of our study area. This reconstruction will
require some or much computation. This, therefore, is
a model with low storage consumption but with high
costs in computation.
The space-time cube model
▪ Like the previous one, this model is based on a
two-dimensional view of the study space
(spanned by the x- and y-axis), in which
geographic phenomena are traced through time
(along the t-axis) thereby creating a three-
dimensional space-time cube.
▪ Multiple, concurrent object changes are even
more difficult to guard topologically, and the
rules of full topological consistency under
continuous change are not even well-
understood.