SPATIAL DATA MODELS

& NETWORKS
Dr P P Nageswara Rao
Outstanding Scientist, ISRO (Retd)
ppn1953@[Link]
 What is Data Model ?
Data ? : These are observations made from real world, organized and
processed to make meaningful information.
A data model describes how data
• Is represented in an information system.
• What relationships exists among them.
• Is structured and used.
• Common models include :
• Heirarchial model
• Network model
• Relational model
• Entity-relationship model
• Data may have three dimensions: thematic, spatial, and
temporal. e.g., temperature at 22° N 88° E was 35° C on 15th
May 2014. it has all three dimensions : temperature (theme), spatial
(location) and temporal (date /time)
 Types of Spatial Data Models
• What are spatial models ? The Real world is too complex for
understanding, so we create “models” or abstraction of reality that have some
similarity with selected aspects of the real world. A spatial model is a collection
of spatially referenced data that act as a model of reality.
• Conceptual data models: Here, spatial features may be
discrete (objects) like roads, buildings, water bodies etc. or
continuous (fields) like contours of elevation, temperature,
rainfall, etc.
• Logical data models: Spatial features are represented in two
forms : Raster and Vector. In this case, objects are put in the
form of vector data whereas fields are in the form of raster
data.
• Object-oriented data models: In this model, data has
properties such as coordinate system, attributes and capability
to respond to requests such as zoom, pan, query, etc.
How are spatial features stored in a
GIS?
 Raster Data Models
• A) Field –based Raster Models :
In this case, each cell contains a
single value representative of
the observation on the surface
that is covered by the cell (e.g.
Elevation).
• B) Object-based raster Models:
In this case, a point feature is
represented as a value in a
single cell., linear feature as a
series of connected cells that
portray length., an area feature
as a group of connected cells
portraying shape.
 Vector structure
The start and end points of a
line are called nodes, any
change of direction on a line
is a vertex. Lines are joined
by sharing a common node.
Polygons are joined
(adjacent) if they share a
common line.

Vector objects are all created from the general graphic

primitives of single x,y coordinates. points are single
coordinates, lines are strings of coordinates, polygons are a
closed string of coordinates (often composed of multiple lines).
 Vector Data Models
• Object-based vector models
– Spagehetti model : In this, data are
collection of points and line segments with
no real connection. These models are
inefficient in data storage and queries and
consistency checks are not possible.
 TOPOLOGICAL MODEL
The topological model is the most widely used method of encoding spatial
relationships in a vector based GIS.
The model is
termed Arc-Node data
model. In this , Point is
encoded as a single XY
co-ordinate pair.
an Arc is a series of
points that starts and
end at a node.
A Node is an
intersection point
where two or more arcs
meet. A Polygon is
comprised of a closed
chain of arcs that
represents the
boundary of the
area.
Topological relationship between objects
 Grid Cell Size and Data Volume
Raster model Vector model
The raster data is
characterized by
subdividing a
geographic space
into grid cells.

MMU = Minimum
Mappable Unit.

The general rule is

that the grid size
should be less than
half of the size of
the MMU.
The choice of grid cell size depend on the user needs and information content.
Grid cell size determine the scale of data compression. A coarser grid cell
structure require less data storage but will provide less accurate representation
of original data.
 File Formats of spatial Data
Raster Formats Vector Formats
BMP - Bit Map Image E00- Arc Export
JPEG – Joint photographic Expert Coverage – ARC/INFO Coverage
group
PNG- Portable Network Graphics MDB - Personal Geodatabase
TIFF- Tagged Image File Format DWG – Auto CAD Drawing File
GIFF- Graphics Interchange DXF – Data Interchange
Format (exchange) File
GeoTIFF – Georeferenced tiff DGN – Microstation Design Files
Image
MrSID – Multi resolution Seamless SHP – Arc View Shape
Image Database
 Raster Data Representation
Columns

Columns
Rows

Rows
Cell-by-Cell Encoding of Raster Data

Run-Length Encoding of Raster Data

If the map is divided in half in both the X and Y directions, four panels (quadrants) are
identified (level1), just two columns and two rows. The computer tests whether any
quadrant wholly contains a single map characteristic. It is further divided into its
quads (Level 2). At that point, there are seven of the 16 quadrants that wholly contain
a single characteristic. Their positions in the four-by-four grid are noted, and the
remaining nine mixed panels are divided into their quads (Level 3). Fourteen of these
are noted as completed, and the remaining 22 are divided for Level 4 of the quadtree.
The process is repeated until an appropriate quad level resolution is reached.
Quadtrees and run-length data structures are good at compressing raster data.
 Field –based Vector Model
• Although vectors are ideal for representing
discrete objects, it can also be used for field-
based continuous data such as elevation,
temperature, etc.
• Mass-points, contour lines / isolines, lattice, and
triangulated irregular network (TIN) are used to
represent elevation or other continuously
changing values.
• However, Raster like continuity cannot be
obtained by any of the above models.
 Point Model
• Multiple points are Pyramid

used to represent surface.

• For example, a pyramid can be
represented by a minimum of five
points, four for base and one for peak.
• Mass-point is a technique to represent
surfaces using several points in a very
dense manner.
• Point model is not a good approach to
represent surfaces, because it requires
the viewer to imagine the surface
joining the points.
Contour / Isoline Model

• Contour is a line on a map that

joins places of the same elevation
above sea level. Contour interval is
the difference in elevation between
two contour lines.
• Isoline is line on a surface,
connecting points of equal value
such as temperature, rainfall, etc.
• Contour / isolines can be
used to represent
surface.
 Lattice Model (wire mesh)

• Lattice is a representation of surface using an array of equally spaced

sample points (Mesh points) that are referenced to a common origin
and have constant sampling distance in the x and y directions. Each
mesh point contains the z value at that location which is referenced
to common base value, e.g., the sea level.
 Triangulated Irregular Network (TIN)

• TIN is used to create a DTM (Digital

Terrain Model) using height data.
The TIN method joins the height
observations (points) together with
straight lines to create non-
overlapping triangles. A triangle
consists of 3 lines connecting three
nodes.
• Shaded relief maps can be
generated by plotting the facets in
3-D and shading them as functions
of their slopes and aspects.
 Triangulated Irregular Network (TIN)

TIN is a vector model designed for elevation data. It uses a set of irregularly
spaced elevation measurements with intensive sampling in areas of complex
relief. A bit of computer wizardry is applied to determine the network of
triangular facets that best fits these data. Each facet has three interconnected
elevations and can be visualized as a tilted triangular plane. The direction
cosines of the plane identify its gradient (slope) and orientation ( aspect). The
average of the three elevations generalizes the plane's height. As shown in figure
, the XY coordinate (location) and the z coordinate (elevation) are stored, in a
features table by their three nodes and adjoining facets. The final link is to an
attribute table which contains descriptive information on each facet.
Raster versus vector Spatial
Raster Model Vector Model
data models
Advantages Advantages
 Simple data  Smaller file size
structure  Individual identity
 Easy and efficient for discrete objects
overlaying  Efficient for
 Compatible with topological
RS data relationship.
 Spatial variability  Efficient projection
is efficiently shown transformation.
 Used for  Accurate map
continuous data output.
 Easy to edit.
Disadvantages Disadvantages
 Large file size  Complex data
 Difficult to edit structure
 Loss of  Difficult overlay
operations
information when
 Not compatible with
large cell size used RS data.
 Not efficient for
continuous data.
 V to R and R to V Conversions

Basic procedures for direct calculation-based vector to raster conversion (VtoR) and
the reverse (RtoV).
Raster data Classified raster
An Example of R2V.
Ground Water
quality map of
Chitradurga district

Vectorized
 Object-oriented data Model
• The object-oriented approach incorporates the abstraction
mechanisms : generalization and aggregation.
• It involve the derivation of attribute values at different levels
of detail and abstraction.
• Two methods for the derivation of properties are : (1)
inheritance describing properties and methods of subclasses
in hierarchies, and (2) propagation deriving properties in
part of hierarchies.
• While inheritance acts in a top-down fashion along the
generalization hierarchy, propagation can derive values from
parts to the aggregates (bottom-up).
• The object-oriented data model is built on the four basic
concepts of abstraction : classification, generalization,
association, and aggregation.
 Example of Inheritance
• Agricultural Land
• Built Up – Crop land
– Urban • Kharif crop (June/July to
• Residential Sept/ Oct)
• Mixed built up • Rabi crop (Nov/Dec to
• Recreational (Playgrounds, Feb/Mar)
parks) • Two crop area
• Public and semi-public
(Schools, colleges) • Three crop area
• Communications (gas/oil • Zaid crop ( summer)
pipes, electric lines) – Fallow land
• Public utilities (water supply, • Current fallow
sewerage, etc)
• Commercial • Long fallow
• Reclaimed land – Plantations
• Vegetated area • Agriculture plantations
• Transportation ( railways, (tea, coffee, rubber)
roads, airports, ports, water • Horticulture plantations
ways (coconut, areacnut, fruits,
vegetables)
• Agro-horticultural plantations
– Aquaculture/ pisciculture
 Time-Space relationship in GIS
Spatial objects may change over time in terms of space
and content.
It may include a change in location, size, orientation, and form.
As a consequence, the spatial relationships among objects may
also change.
What are networks ?
What are networks in GIS context ?
Networks in a GIS
Basics of Network Analysis
Spatial Data Models in Transport Network
 Network Data Models
There are four basic application areas of network data models:
Topology. The core purpose of a network data model is to provide an accurate
representation of a network as a set of links and nodes with proper arrangement
indicating location, direction and connectivity.
Cartography showing symbols for each of the attributes. For instance, a highway
symbolized as a thick line with a label such as its number, while a street may be
symbolized as an unlabeled simple line.
Geocoding to pinpoint the location of an address, or any location along the
network, with reasonable accuracy.
Routing and assignment. Network data models may be used to find optimal paths
and assign flows with capacity constraints in a network.

Routing is concerned by the specific behaviour of a limited number of vehicles,

traffic assignment is mainly concerned by the system-wide behavior of traffic in a
transport network.

This requires a topology in which the relationship of each link with other
intersecting segments is specified. Impedance measures (e.g. distance) are also
attributed to each link and will have an impact on the chosen path or on how flows
are assigned in the network. Routing and traffic assignment in an urban area is
much more complex as it must consider stop signs, traffic lights and congestion,
in determining the impedance of a route.
Topology of a Network Data Model
Cartography of a network Model
Geocoding in a Network Model
Routing in a Network Model
Relational Database Representation of
a Simple Network
Creation of a Connectivity Matrix
with a link Table
A links table can easily be used to construct a connectivity matrix, a core
element for measuring accessibility. It can be achieved by using the simple
algorithm :The values of all the cells Cij in the connectivity matrix C are set to 0.

The default assumption that nodes are not linked.

For each record of the links table, the link becomes a connection (value of 1).
For each record of the links table, an inverted connection is created if the link is
bi-directional.

Turn Penalties at an
Intersection
Basic Representation in a network model
Constructing the geometry of a network depends on the mode and the scale
being investigated. For urban road networks, information can be extracted
from aerial photographs or topographic maps. Air transport networks are
derived from airport locations (nodes) and scheduled flights between them
(links).

Two fundamental tables are required in the basic representation of a network

data model that can be stored in a database:

Node table. This table contains at least three fields; one to store a unique
identifier and the others to store the node's X and Y coordinates. Although
these coordinates can be defined by any Cartesian reference system,
longitudes and latitudes would insure an easy portability to a GIS.

Link table. This table also contains at least three fields; one to store an
unique identifier, one to store the node of origin and one to store the node of
destination. A fourth field can be used to state if the link is unidirectional or
not.
 Layer-Based Approach
Among the most significant attributes that can be attached to network layers are:
Classification and labeling : classified into categories such as its function (street,
highway, railway, etc.), importance (number of lanes) and type (paved, non-paved).
Linear referencing system: One of the most common is the address system
where each segment is provided with an address range. Through linear
interpolation, a specific location can be derived (geocoding).
Segment travel costs. Can consider a vast array of impedance measures. Among
the most common is the length of the segment, a typical travel time or a speed limit.
Congestion can also be assessed, either as a specific value of impedance or as a
mathematical function.
Direction. To avoid unnecessary, and often unrealistic duplication of links,
especially at the street level, a directional attribute can be included in the attribute
table.
Overcrossing and undercrossing. A provision must be made in the attribute table
to identify segments that are overcrossing or undercrossing a segment they are
intersecting with.
Turn penalties : Conventionally in road transportation, a right turn is assumed to
have a lesser penalty than a left turn. The opposite applies for countries where
driving is on the left (e.g. UK).
 The TIGER
The TIGER (Topologically Integrated Geographic Encoding and
Referencing) model is a notable example of a layer-based structure which
has been widely accepted.

TIGER was developed by the US Census Bureau to store street information

constructed for the 1990 census.

It contains complete geographic coordinates and in a line-based structure.

The most important attributes include street name and address information,
offering an efficient linear referencing system for geocoding.

The layer-based approach is consequently good to solve the cartography

and geocoding issues. However, it is ill-suited to comprehensively address
routing and assignment transport problems.
 Object-Oriented Network Model

A transport network can be represented by two object classes : Link and Node.
The transport network, inherits the properties of its two subclasses. Because of the
topological attributes of a network, a connection relationship has to be established
between the link and node classes. It defines nature and extent of the geometric
connectivity. Object-oriented network models are stored in relational databases.
The above representation would require four relational tables : one labeled Transport
Network, which could be a simple identifier for the network, a Node and a Link table containing their respective properties,
and a Connection table where connectivity information is stored.
 Thank you

Any clarifications ?