BIT 2307
Geographic Information Systems
Geo-spatial Modeling
The Modeling of Spatial Entities
• During the design of spatial entities, the
GIS designer moves from the position of
observing the geographical complexities of
the world to one of simulating them in the
computer. The process involves;
The Modeling of Spatial Entities
• Identifying the spatial features from the real world
and deciding on how to represent them in a
conceptual model (e.g. as lines, points or areas)
• Representing the conceptual model by an
appropriate spatial data model. This involves
choosing between raster and vector models. In
many cases the equipment and GIS software
available may dictate this choice
• Selecting an appropriate spatial data structure to
store the model within the computer. The spatial
data structure is the physical way in which entities
are coded for the purpose of storage and
manipulation.
The Modeling of Spatial Entities
• There are a number of problems
associated with simplifying the
complexities of the real world into
2 dimensional entity types. These
include:
Spatial Modeling - Problems
1. The dynamic nature of the real world: the real
world is not static. Forests grow or are felled,
rivers flood and meander, and cities expand.
This poses 2 problems when defining an entity
i.e. selecting the appropriate model to represent
the feature under discussion (e.g. whether to
represent a forest as an area or a collection of
points). There is also a problem in representing
changes over time, e.g. a forest originally
represented as an area may gradually decline
until it is a dispersed group of trees better
represented by points.
Spatial Modeling - Problems
[Link] identification of discrete and continuous
features. One’s perception of the geographic
structures may differ according to different
application areas. E.g. sightseers may be
satisfied with roads represented as lines but to
the road construction and maintenance
engineers it may be more sufficient to
represent these roads as areas or even
surfaces. Areas with “fuzzy” boundaries such
as the woodland can also be problematic
when defining entities and also during
analysis
Spatial Modeling - Problems
[Link] scale at which a particular problem
needs addressing. E.g. road users may be
satisfied by roads being represented as a
series of line entities to which information is
attached about the status of these roads. But
to the road engineers additional entities and
information are required such as a set of
points to mark the location of road bumps
and potholes, area entities to mark lanes and
breadth of roads and simple line segments
used in the digital road map.
Raster Data Structures
• In a simple raster data structure different spatial
features need to be stored as separate layers.
• One of the major problems associated with raster
data sets is their size, since a value must be
recorded and stored for each cell in an image.
Thus a complex image made up of a mosaic of
different features (e.g. a soil map displaying 20
distinct soil classes) requires the same amount of
storage space as a similar map showing the
location of a single forest.
• To avoid this problem, a range of data compaction
techniques have been developed. These include
run-length encoding, block coding, chain coding
and quadtree data structures
Raster Data Structures
Run Length Encoding
• This technique reduces the data volume on a
row-by-row basis. It stores a single value
where there are a number of cells of a given
type in a group rather than storing a value for
each individual cell.
Raster Data Structures
Block Coding
• This technique extends the run length
encoding idea to 2-dimensions by using a
series of square blocks to store data.
Coordinates are required to locate the blocks
in a raster matrix. The top left-hand cell is
used as the location reference for the block.
Raster Data Structures
Chain Coding
• This method of data reduction works by
defining the boundary of an entity. This
boundary is defined as a sequence of unit cells
starting from and returning to a given origin.
The direction of travel around the boundary is
usually given using a numbering system (e.g.
0 = North, 1= East, 2 = South and 3 = West).
Raster Data Structures
Chain Coding
• The first line denotes where the coding started. On the
2nd line, the 1st line of each sequence represents the
direction and the number represents the number of
cells lying in that direction
N/B: The letters here are for clarity, i.e. numbers are used
to represent direction, e.g. 1,2, 1,5, etc.
Raster Data Structures
Quadtrees
• Works by recursively subdividing the cells in a
raster image into quads (quarters). The subdivision
process continues until each cell in the image has
been classed as having the spatial entity either
present or absent within the bounds of its
geographical domain. The number of subdivisions
required to represent an entity will trade between
the complexity of the feature and the dimensions
of the smallest grid cell
Raster Data Structures
Quadtrees
Modeling the Third Dimension
• In terms of display capabilities, the computer
screen is a 2-D display device, even though with
graphics it is possible to simulate the appearance of
3-D. This term is referred to as 2.5D.
• All the world features to be modeled have a third
dimension. To produce systems capable of
representing the complexities of the real world we
need to portray the third dimension in more than a
visual way. The representation of the third
dimension can aid in modeling the form of an entity
and associated spatial processes. This allows users
to evaluate the potential impact of the development
on the landscape
Modeling the Third Dimension
• Many GIS software products don’t allow
the modeling of the 3rd dimension, partly
because of the technical difficulties in
designing 3D models of geographical
space.
• This problem is however being addressed
with advancements in GIS technology
Surface Modeling
• Modeling surfaces e.g. height, population,
rainfall is problematic in GIS
• The following models are used to represent
such data;
- Digital terrain models (DTM)
- Digital elevation models (DEM)
- Digital surface models (DSM)
- Digital height models (DHM)
Digital Terrain Model
• Has wide application in GIS
• Describes a digital data set used to model a
topological surface (surface representing height
data)
• Created from a series of either regularly or
irregularly spaced x,y,z data points.
• May derived from many data sources e.g. contour
and spot height information in topographic maps,
aerial photography, satellite images &field survey
• The ‘resolution’ of a DTM is determined by the
frequency of observations used
Digital Terrain Model
California DTM
Digital Elevation Model
• Models surface elements like the DTM
• But DEM usually represents elevation of
ground above the surface of reference
Digital Surface Model
• Also used in various parts of the world to
represent various elements of elevation
• Accurate modeling of a surface requires the
storage of an infinite number of observations, this
is impossible
• A surface model approximates a continuous
surface using finite observations
• Thus an appropriate number of observations must
be selected together with their geographical
locations
• The DSMs display all of the measured points
collected by the sensor, as illustrated by the
irregular elevation spikes in the diagram.
Digital Surface Model
• Non-terrain elements are removed through
filtering to create a DTM. The DTMs simulate a
“bald-earth” and are traditionally preferred over
DSMs for topographic mapping purposes.
Digital Surface Model
Digital Height Model
• Represents the height of various objects
e.g. buildings, trees, etc.
Swiss topological DHM
Modeling the Fourth Dimension
• Any GIS database will reveal data layers
that represent various entities at a given
time. But as the production of a GIS
mostly includes entities captured at
different periods of time, the modeling of
time becomes more complex.
• There are several types of time that the
GIS designer can consider; work practice
time, database time and future time
Modeling the Fourth Dimension
Work practice time
• This is the temporal state of a GIS database
used by many people. Here there can exist
many different versions of the GIS in
existence at any one moment in time by
several people working to update
information, e.g. more than one person
working at the same time to update a
particular land parcel
Modeling the Fourth Dimension
Database time
• This is the period for which the database is
thought to be correct. Inevitably this lags
behind time because of the data capture
process. This problem can occur if there is
more than one version of the database being
used and there are no procedures governing
who updates it and when
Modeling the Fourth Dimension
Future Time
• Many GIS applications seek to model future time.
• Applications like forecasting require predictions
to be made.
• This can be complex especially when there are
several possible alternative scenarios, e.g. a GIS
used to model avalanches under different climatic
conditions will yield several different future
images that the GIS needs to handle.
Modeling the Fourth Dimension
•Problems with developing a data structure flexible
enough to depict time include; what to include, how
frequently to update old information, and when to
stop storing old information.
•Additionally, many organizations may need its GIS
to be accessible to individuals at different geographic
locations and for them to make changes to the time
status of an entity.
•Most GIS users represent different time states for an
entity by either creating a separate data layer for
each time period or by recording the state of the
entity at a given time as a separate field in the
database.
Modeling the Fourth Dimension
• There are four types of temporal (time)
events which provide an indication of
the types of changes that may affect an
entity. These include;
– Continuous: these events go on throughout
some interval of time
– Majorative: these events go on most of the
time
– Sporadic: these events occur some of the
time
– Unique: these events occur only once.