Serryn Eagleson

1
Hierarchical Spatial Reasoning Applied to
Administrative Boundaries Using GIS

• Serryn Eagleson
• Francisco Escobar
• Ian Williamson
• Department of Geomatics
• The University of Melbourne

2
Outline

• Background
• NSDI
• The problem uncovered
• Hierarchical Spatial Reasoning (HSR)
• Methods to solve the problem
• HSR - Applied to Administrative Boundaries
• Conclusion

3
The Problem

• Throughout history boundaries have been used to
  divide the space around us into administrative
  units
• In the majority of cases these boundaries have
  been established to facilitate administrative,
  political and economic purposes

4
Examples of Different Spatial Boundaries in
Victoria

• Land market - Cadastre
• Collection of statistics - Collection district
• Facilitate mail delivery - Postcode
• Government - Local Government Areas
• Natural boundaries - DNRE
• Political boundaries - Electoral boundaries
• etc...

5
East Gippsland
6
North West Melbourne Health Division
7
The Problem
8
Research Objective

• To identify the principles of Spatial
• Hierarchical Reasoning and apply
• them to Administrative polygons in
• Victoria using GIS

9
Possible Solutions

• Aggregation from point data to new boundaries
• Data Interpolation
• ?Reorganisation of Boundaries

10
Point Aggregation
11
Interpolation Process

• Data Interpolation
• Formula
• Vt ? Us(ast/?ast)
• Errors

12
Interpolation Errors
13
The Solution Boundaries Reorganised
14
Hierarchy

• Hierarchies have been used to break
• down a number of complex problems
• Quadtrees, Route Planning
• Properties
• Janus Effect
• Whole Part
• Near decomposability

15
Hierarchical Reasoning
Hierarchical Reasoning is based on the principle
to use the least detailed representation to
answer a question... (Timpf Frank 1997)
16

• Example
• In theory country boundaries can be deduced
• by processing every parcel in the cadastre
• Hierarchical Reasoning implies this to be a
• slow and inappropriate method of determining
• the country boundary
• Therefore we require a Spatial Hierarchy on
• which to reason
• (Timpf Frank 1997)

17
HSR Requirements

• Hierarchical structure and a method to
• transform the non hierarchical structure to
• the equivalent hierarchical one
• A set of rules on how to reason on the
• hierarchical structure
• (Car, 1997)

18
Development of the Spatial Hierarchy

• Organisational Requirements
• Key layers in the Hierarchy
• Rules for establishing the layers

19
Key Similarities Between Organisations

• Infrastructure Boundaries
• Railways
• Rivers
• Road Network
• Routing
• Identifiable on the ground

20
How can this system be implemented using GIS?
Spatial Search Thematic Search
21
Constraints to be considered

• Road Network
• Additional constraints
• contiguity
• shape
• area
• confidentiality
• density
• culture

22
Spatial selection approach
Thiessen Polygons - ( Martin 1991, Openshaw,
1992)
23
Thematic Selection

• Simplistically Algorithms determine the
  selection process based on attributes
• If (constraints are met) then
• add the Polygon to selection
• else
• Do Not
• end
• Union.Selected

24
Results
25
The ideal hierarchically ordered administration
system will enable

• Data Integration between organisations
• Rules for reasoning within the system
• Sufficiently precise information to the task at
  minimum economical cost

26
Conclusion

• The final objective will be a integrated
  hierarchical network of boundaries, at
  facilitating the collation of data into
  information.

27
Exam Questions

• Discuss the possibilities and issues of applying
  Hierarchical Spatial Reasoning to spatial
  boundaries
• List and describe three different approaches to
  solve the problem of data exchange amongst
  non-coterminous boundaries