Water Corporation take on Referential Topology

1
Water Corporation takeon Referential Topology

• WALIS Referential Topology Workshop April 09
• Presenter David Bulfield

2
Cadastral Re-alignment Project

• What is the Cadastral Re-alignment Proposal ?
• It is a misnomer for an asset re-alignment
  exercise to shift the locations of some 3million
  asset records to a new set of cadastral
  boundaries across the State.
• Land base is maintained independently by Water
  Corporation based on proposed subdivision
  Development activity
• Asset information is plotted in relation to
  proposed or actual parcel edges and road
  casements.

3
What is different with a Utility

• We deal almost exclusively with linear and point
  objects
• Generally, objects are related to but not
  necessarily linked (snapped) to boundary offsets.
• Every spatial shift of a parcel edge corner
  results in shift vectors applying to every
  related object.
• Changing lengths of linear objects cant always
  be reconciled.
• And we still have to contend with polygon slivers
  from static boundary layers (eg Loc Govt, MRS
  etc ..)

4
(not) Just another map!
Sewer
Water
5
Conflict in Our Business Needs
Data captured into GIS continually revised
updated from Survey information
Submit Development Application to DPI
DPI Refers to Agencies for comment
Applicant submits proposed asset details for
approval
Engineering/precalc survey revisions
Deposit Plan at Landgate
WAPC Final Approval
Landgate capture data from lodged survey plans.
Current lodged survey plans applicable until this
point
6
The Cadastral relativity problem

• Data was originally digitised from hardcopy plans
  at various scales with a resulting low degree of
  spatial accuracy.
• 1992, the Water Corporation stopped utilising
  spatial updates from the authoritative Landgate
  database and maintained data in isolation.
• Landgate continued to improve accuracy of its
  SCDB.
• Adhoc acquisitions of data from Landgate in
  outlying urban areas revealed a growing disparity
  with Core Land data.
• Recent expansion (explosion) of digital data
  requests and experienced GIS users indicates
  customers may not be as willing to warp our
  data to fit theirs.

7
Examples of current data variation
Figure 1 Example of variation of road centreline
to parcel edges. - Boyanup Road centreline -
Green Parcel edges - Black
Figure 2 Example of parcel edge variation
Landgate to Water Corporation. - Boyanup Landgate
- Blue Water Corporation - Black
8
Another example
Road centre-line
Our street location
9
Issues with Boundary Offset Relationships
Note change in pipe length !
10
So, Whats the point?

• If we never share our data keep it to
  ourselves, then we dont need to do anything.
• Why not just ignore the impact on others and our
  own data capture issues and continue as before.
• Why not just delete all the land and start again.
• Someone else has done this before why not ask
  them why they did it ( how)?

11
Our Project - Progress to date

• Phase 1 - Feasibility completed
• Identified a significant impact on the business
  when comparing our asset locations against other
  external datasets.
• Identified at least 17 prospective toolsets
  ostensibly covering the land base alignment
  process.
• Phase 2 - toolset evaluation pilot, cost and
  revised process discovery just started.

12
Other takeaways!

• Regardless of the known, accurate absolute
  location of our assets, we are obliged to relate
  asset location to cadastre.
• Use of Imagery greatly increases the perception
  of land base error.
• There is a product discrepancy between source
  data and users required accuracies.
• Dependent customer agencies should not be
  subjected to unnecessary shifts.

13
Questions ?
THANK YOU