11. DATA MODELS
Conversion of real world geographical variation into discrete objects is done through
data models. It represents the linkage between the real world domain of geographic
data and computer representation of these features. Data models discussed here are
for representing the spatial information.
Data models are of two types: Raster and Vector. In raster type of representation
of the geographical data, a set of cells located by coordinate is used; each cell
is independently addressed with the value of an attribute. Each cell contains a
single value and every location corresponds to a cell. One set of cell and associated
value is a LAYER. Raster models are simple with which spatial analysis is easier
and faster. Raster data models require a huge volume of data to be stored, fitness
of data is limited by cell size & output is less beautiful.
Vector data model uses line segments or points represented by their explicit x,
y coordinates to identify locations. Discrete objects are formed by connecting line
segments which area is defined by set of line segments. Vector data models require
less storage space, outputs are appreciable, Estimation of area / perimeter is accurate
and editing is faster and convenient. Spatial analysis is difficult with respect
to writing the software program.
Figure 3. Vector and raster data examples
12. DATA STRUCTURES
There are number of different ways to organize the data inside the information system.
The choice of data structure affects both; Data storage volume and processing efficiency.
Many GIS have specialized capabilities for storing and manipulating attribute data
in addition to spatial information. Three basic data structures are – Relational,
Hierarchical and Network.
Relational data structure organizes the data in terms of twodimensional tables
where each table is a separate file. Each row in the table is a record and each
record has a set of attributes. Each column in the table is an attribute. Different
tables are related through the use of a common identifier called KEY. The information
is extracted by relations who are defined by query.
Hierarchical data structure stores the data in a way that a hierarchy is maintained
among the data items. Each node can be divided into one or more additional node.
Stored data gets more and more detailed as one branch further out on the tree.
Network data structure is similar to hierarchy structure with the exception that
in this structure a node may have more than one parent. Each node can be divided
into one or more additional nodes. Nodes can have many parent. The network data
structure has the limitation that the pointers must be updated every time a change
is made to database causing considerable overhead.
13. ERRORS IN GIS
(i) Errors in GIS environment can be classified into following major groups
Age of data

 Reliability decreases with age

Map scale

 Nonavailability of data are proper scale or Use of data at different scales

Density of observation

 Sparsely dense data set is less reliable

Relevance of data

 Use of surrogate data leads to errors

Data inaccuracy

 Positional, elevation, minimum mapable unit etc.

Inaccuracy of contents

 Attributes are erroneously attached

(ii) Errors associated with processing
Map digitization errors

 due to boundary location problems on maps and Errors associated with digital representation
of features

Rasterisation errors

 due to topological mismatch arising during approximation by grid

Spatial Integration errors

 due to map integration resulting in spurious polygons

Generalization errors

 due to aggregation process when features are abstracted to lower scale

Attribute mismatch errors


Misuse of logic

