Geographic Information Systems:
an introduction
Week III
Data models cont’d
The Nature of Geographic Data
Last week
• Representing geographic data
– E.g. continuous fields, discrete objects
• Data models
– E.g. Raster, vector
Data Model Levels
Humanoriented
Reality
Conceptual Model
Logical Model
Computeroriented
Physical Model
Increasing
Abstraction
Logical Data Models &
Applications
•
•
•
•
•
•
•
CAD
Graphical
Image
Raster
TIN
Geo-relational
Object
•
•
•
•
•
•
•
Engineering design
Simple mapping
Image processing and analysis
Spatial analysis / modeling
Surface /terrain analysis / modeling
Geoprocessing geometric features
Features with behavior
Raster and Vector Models
• Raster – implementation of field conceptual model
– Array of cells used to represent objects
– Useful as background maps and for spatial analysis
• Vector – implementation of discrete object
conceptual model
– Point, line and polygon representations
– Widely used in cartography, and network analysis
Rasters and Vectors
• How to represent phenomena conceived as fields
or discrete objects?
• Raster
– Divide the world into square cells
– Register the corners to the Earth
– Represent discrete objects as collections of one or more
cells
– Represent fields by assigning attribute values to cells
– More commonly used to represent fields than discrete
objects
Legend
Mixed conifer
Douglas fir
Oak savannah
Grassland
Raster representation. Each color
represents a different value of a nominalscale field denoting land cover class.
Characteristics of Rasters
• Pixel size
– The size of the cell or picture element, defining the
level of spatial detail
– All variation within pixels is lost
• Assignment scheme
– The value of a cell may be an average over the cell, or a
total within the cell, or the commonest value in the cell
– It may also be the value found at the cell’s central point
Vector Data
• Used to represent points, lines, and areas
• All are represented using coordinates
– One per point
– Areas as polygons
• Straight lines between points, connecting back to the
start
• Point locations recorded as coordinates
– Lines as polylines
• Straight lines between points
Vector Data
POINTS
POLYGON
LINES
NODES
ANNOTATION
Vector - Land Records
Surveyed feature
20.37’
26.23’
13
12
30.5’
26.23’
GIS
Survey
Link
/
Survey point
/
/
/
/
/
9
Computation
Raster vs Vector
• Volume of data
– Raster becomes more voluminous as cell size decreases
– Raster is vaster
• Source of data
– Remote sensing, elevation data come in raster form
– Vector favored for administrative data
– Vector is more appropriate when storing large amounts
of attribute information.
• Software
– Some GIS better suited to raster, some to vector
Data Model Levels
Humanoriented
Reality
Conceptual Model
Logical Model
Computeroriented
Physical Model
Increasing
Abstraction
Data Models and Spatial Analysis
•
•
•
•
•
Raster v. vector?
Picture of a large burn
Picture of an urban landscape
Picture of burning houses
Map of burning houses
Geo-relational Model
Formats of feature data models in
ArcGIS
• Georelational models
– Coverages
– Shapefiles
• Database Management System (DBMS)
– Geodatabase
Coverages
Traditional model
for complex
(multiple feature
types)
geoprocessing
environments
Coverages
primary
.
composite
routes sections regions
secondary
annotation ticks links
Shapefiles
Simple, open,
“flatfile” based
format for single
data features
Geodatabase
Object-based
data model
where all features
are contained
in database;
features can
exhibit “behavior”
Compression techniques
No compression
Run length encoding (lossless)
ADRG - National Imagery and Mapping Agency (NIMA)'s ARC Digitized Raster
Graphics
BIL - Band Interleaved by Line (image format linked with satellite derived
imagery)
CADRG - National Imagery and Mapping Agency (NIMA)'s Compressed ARC
Digitised Raster Graphics (nominal compression of 55:1 over ADRG)
CIB - National Imagery and Mapping Agency (NIMA)'s Controlled Image Base
(type of Raster Product Format)
Digital raster graphic (DRG) - digital scan of a paper USGS topographic map
ECW - Enhanced Compressed Wavelet (from ERMapper). A compressed
wavelet format, often lossy.
ESRI grid - binary and ASCII raster formats used by ESRI
GeoTIFF - TIFF variant enriched with GIS relevant metadata
IMG - ERDAS IMAGINE image file format
MrSID - Multi-Resolution Seamless Image Database (by Lizardtech). A
compressed wavelet format, often lossy.
Vector formats
Geography Markup Language (GML) - XML based open standard (by OpenGIS)
for GIS data exchange
DXF - Contour elevation plots in AutoCAD DXF format
Shapefile - ESRI's open, hybrid vector data format using SHP, SHX and DBF files
Simple Features - Open Geospatial Consortium specification for vector data
MapInfo TAB format - MapInfo's vector data format using TAB, DAT, ID and MAP
files
National Transfer Format (NTF) - National Transfer Format (mostly used by the UK
Ordnance Survey)
TIGER - Topologically Integrated Geographic Encoding and Referencing
Vector Product Format - National Imagery and Mapping Agency (NIMA)'s format of
vectored data for large geographic databases.
GeoMedia - Intergraph's Microsoft Access based format for spatial vector storage.
ISFC - Intergraph's Microstation based CAD solution attaching vector elements to a
relational Microsoft Access database
Personal Geodatabase - ESRI's closed, integrated vector data storage strategy
using Microsoft's Access MDB format
Coverage - ESRI's closed, hybrid vector data storage strategy. Legacy ArcGIS
Workstation / ArcInfo format with reduced support in ArcGIS Desktop lineup
Grid formats (for elevation)
USGS DEM - The USGS' Digital Elevation Model
DTED - National Imagery and Mapping Agency (NIMA)'s Digital Terrain Elevation
Data
GTOPO30 - Large complete Earth elevation model at 30 arc seconds
SDTS - The USGS' successor to DEM
Other formats
Binary Terrain - The Virtual Terrain Project's Binary Terrain format
Well-known text (WKT) – ASCII spatial projection description (ESRI uses a *.prj
extension)
Well-known binary (WKB) - Binary spatial projection description
Geo-relational Model
Formats of feature data models in
ArcGIS
• Georelational models
– Coverages
– Shapefiles
• Object model
– Geodatabase
Coverages
Traditional model for complex geoprocessing
environments
•Topologically correct
•Data processing and spatial analysis
•Complex data structure
•Difficult interchange (E00)
•Multiple feature types
primary
.
composite
routes sections regions
secondary
annotation ticks links
Coverages
Shapefiles
Simple, open, “flatfile” based format for single
data features
•Single feature type within one file
•No topology
•Easy to transport
•Open
Shapefiles
Geodatabase
Object-based data model where all features are
contained in database; features can exhibit
“behavior”
•Multiple features stored as tables
•Data processing topology
•Object oriented instead of restricted to points, lines, polys
•Resides in traditional database
Geodatabase
Topology
• Field within mathematics.
• The study of the general abstract nature of continuity or
"closeness" on spaces.
• System administrators actually describe computer
networks as being topologic
– Bus topology
– Ring topology
– Star topology
• Leonhard Euler
–
–
The solution of a problem relating to the geometry of position 1736.
Euler was aware that he was dealing with a different type of geometry where distance was not
relevant.
Consider the problem of building a fusion reactor
which confines a plasma by a magnetic field.
The solution of a problem relating to
the geometry of position
Topology
• Science and mathematics of geometric
relationships
–
–
–
–
Simple features + topological rules
Connectivity
Adjacency
Shared nodes / edges
• Topology uses
– Data validation
– Spatial analysis (e.g. network tracing, polygon
adjacency)
Topology and GIS
• Topology in GIS is generally defined as the spatial
relationships between adjacent or neighboring
features.
• Two types of topology:
– Polygon topology
– Line topology
• three advantages of incorporating topology
in GIS databases:
data management, data correction and
spatial analysis
Data correction and storage
Topology and spatial analysis
• Adjacency
• Connectivity (network models)
• Containment
Polygon Topology Model
Polygon Topology Contiguity
Definitions
• A method of analysis is spatial if the results
depend on the locations of the objects being
analyzed
– move the objects and the results change
– results are not invariant under relocation
• Spatial analysis requires both attributes and
locations of objects
– a GIS has been designed to store both
Spatial Autocorrelation
• Spatial autocorrelation is determined both
by similarities in position, and by
similarities in attributes
– Sampling interval
– Self-similarity
Tobler
Spatial autocorrelation measures
n
i,j
zi
c i,j
wi,j
number of objects in the sample
any two of the objects
the value of the attribute of interest for object i
the similarity of i’s and j’s attributes
the similarity of i’s and j’s locations