Chapter three
Data Structures and Analysis
By: Tekeste Demesie
1
Introduction
• From the discussion of Fundamental GIS Concepts we noted:
• two types of data: spatial data (where) and attribute data (what)
• two models for data representation: raster model and vector model
• two computing environments: traditional relational data base model and object-
oriented model
• as we search for software for our GIS project, ideally we seek
implementations which:
• handle both spatial and attribute data, and pay commensurate attention to each
• support vector and raster formats, including
• conversion between the two
• analytic capabilities for both
• utilize relational data base or object-oriented computing models
• how do software vendors in the market place meet these needs?
Data Input: Data Model
• 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.
• Two major categories of spatial data representation in GIS: raster and
vector.
• Raster approach: cells
• Vector approach: points, lines, and polygons
Tekeste D.
GIS Data Structures?
• Method of representing reality in GIS system
• Common data structures: Raster and Vector
Raster data
Reality
Vector data
Vector Data Structures
• In vector GIS, geographical data are three types;
points, lines and polygons.
• Features are separated from each other by defined
borderlines.
• Examples are rainfall stations, streams in a
watershed, agricultural plots in a given piece of
land
• The point hasno spatial extent and represented by a
single pair of co-ordinates
• Formats: .shp, .dbf, .prj, .xml, .sbn, .sbx, .cpg
Vector Data Structures
• Geometrical objects in the vector model are stored as one or
several pairs of coordinates
Vector Data Structures
Vector Data Structures
Tekeste D.
Vector Data Structures: Points
• 1 X,Y location
• Applies to features:
• No area
• Without area
• No length
• Too small to displayed at
• Display: marker symbol, current scale
size, colour
Vector Data Structures: Lines
• Series of X,Y points • Applies to features:
• No area • Without area but with length
• Too thin to represent as an area at
• Have length
the current scale
• Connectivity
• Display: Line stile,
thickness, colour,
offset
11 December 2022
Vector Data Structures: Polygons
• Enclosed region • Applies to features:
- Countries
• Area - Provinces
• Display - Catchments
• Fill pattern, fill color, outline
pattern and color
Vector Data Structures: Advantage
• Exact
• Fast for processing
• Good for visualization
• Compact in terms of data structure
• Often topological
• Discrete data is ideally represented with a vector
database
Topology
• The study or representation of spatial relationships between geometric
features — how points, lines, and polygons are connected, adjacent, or
overlap.
• Focus: Connectivity and relationships — not shape or elevation.
• Used in GIS/vector data to ensure data integrity (e.g., roads connect at
intersections, polygons don’t overlap).
• Examples:
• A river line must connect to a lake polygon.
• Adjacent land parcels share common borders.
• No gaps or overlaps between administrative boundaries.
• In short:
Topology = “how features relate to each other.”
Topography
• The physical surface features of an area — including its elevation, slopes,
hills, valleys, and landforms.
• Focus: Shape and elevation of the Earth’s surface.
• Used in maps, DEMs (Digital Elevation Models), and contour mapping.
• Examples:
• Mountain height: 2,500 meters
• Valley depth, hill slope, and contour lines
• In short:
Topography = “the shape of the land.”
Vector Data Structures: Disadvantage
• The location of each vertex needs to be stored explicitly
• Analysis functions are complex and may be processing
intensive
• Continuous data, such as elevation data, is not effectively
represented in vector form
• Spatial analysis and filtering within polygons is
impossible
Raster Data Structure
• A raster data represents the real- world
using cells or pixels (smallest unit of an image)
with no distinct border lines
• The same features represented in vector structure
can also be represented by raster structure
• Rainfall station, streams in a watershed,
agricultural plots in a given piece of land
• Attribute data is not common
• Formats: TIFF, JFIF, BMP, RLC, BIL, BSG, BIP,
GIF, IMG, RDS, GRASS
Raster Data Types
A. Satellite imagery :
B. Aerial photograph :
C. Scanned maps :
D. Photos :
Raster Data: Cell (Pixel)
Cell resolution implies dimensions of the cell
Raster Data: Cell (Pixel)
• A cell is identified using index
• For example, cell “7,5” = 7th
row and 5th column in a
matrix
• The cell consists the data value
at the center
The value is 2
What Does a Raster Contain
• Numbers, characters, or codes
1=Farm 2=Water
3=Green area
4=Pasture 5=Forest
6=Building
• Raster data requires a lot of storing space as it does
not allow for “empty” cells
Raster Data: “Mix-Cell”
• One solution is to take to the dominant area
within a cell
• Another method is to
code the cell with the
class found at the
center of the cell
Raster Data: Advantages
• Way of storing is quite simple and
easy to understand
• Is very efficient for representing
continuous surfaces
• Particularly suitable for combining
with remote sensing data
• Enable overlay analysis
• Raster structure is more
efficient for continuous data
Raster Data: Disadvantages
• Handling can be somehow slow if the cells are small and
larger areas
• Topological information unavailable
• Information about the exact location of linear
objects are lost
• Fuzzy boundaries
Format Issues
• Most GIS applications can utilize both vector and raster formats, and/or they
can convert between the two
• Converting from vector to raster is easy;
• Converting from raster to vector is difficult
Tekeste D.
Triangulated Irregular Network (TIN)
• Represents space using a set of non-overlapping
triangles that border one another and vary in size and
proportion
• Are created from a set of input points with x, y
coordinate values as well as z-values
• Commonly used
for representing
surface
Triangulated Irregular Network (TIN)
• Consists of:
–Points
–Lines (Edges)
–Areas (Faces)
• Mass Points – Irregularly distributed
sample points, each with an x,y
location and a z value, which are used
as the basic elements to build a TIN.
THANKS!
Any questions?
By: Tekeste Demesie
27