How to establish a baseline

Written July 7th, 2014 by

Baseline image

The baseline is the most important element of an urban growth model.  Actually, this statement applies to any model. The baseline establishes the reference point from which a model will forecast future growth.  Not having an accurate baseline will result in your model either over or under allocating urban growth, especially in the early projection year cohorts.

Normally your baseline can be derived from the existing land uses of cadastral land parcels within the area that is subject to the urban growth model.  You might think establishing a baseline for a large cadastral model is a very challenging task.  Actually with today’s digital geospatial tools it is relatively easy. Although it is still time-consuming.

So how do you go about creating a baseline?

Creating the Baseline

Our approach to preparing a baseline for an urban growth model is to use GIS spatial tools.  First we create a GIS layer containing the all the land parcels within the area covered by our urban growth model. Normally this layer is established from the current Digital Cadastral DataBase (DCDB).

After the baseline GIS layer has been created, the next thing to do is identify all land parcels contained within road reserves, railway reserves and waterways. We remove the polygons for these land parcels from the baseline layer. This avoids this data adding to the size and complexity of the urban model when it is not required.

We then determine the parcels that cannot be developed. These parcels normally would have the land uses such as:

  • Open space such as parks and reserves.
  • Land utilised by utilities, such as electricity or telecommunications organisations.
  • Easements.
  • Community faculties such as cemeteries.
  • Land that makes it uneconomic or prevents building construction, such as steep sloping land.

Determine attributes

Next the attributes of existing residential dwellings and non-residential (retail / commercial / industrial) buildings that are currently “on the ground” need to be described and verified.  The methodology used to generate your urban growth model will dictate what attributes are required for your baseline. The Forecaz Modeller requires the following attributes:

  • For parcels that have residential uses:
    • If the dwelling is detached (i.e. house) or attached (i.e. duplexes, units, townhouses) or vacant land.
    • The existing use for those dwellings (i.e. short term accommodation, dwelling house).
    • For vacant land, the approved land use.
  • For parcels that have non-residential uses:
    • The gross floor area of the building(s). Determined from the building footprint and the number of floors in the building.
    • The existing land use (i.e. commercial office, retail shop).
    • For vacant land, the approved land use.
  • For parcels that have multiple land uses including mixed residential and non-residential uses, each use is recorded separately.

The determination of these attributes can be achieved through the interrogation, analysis and consolidation of a number of data sources.  Typically these data sources are:

  • Latest aerial photography.
  • Google Maps and Street View.
  • Local Government planning scheme GIS layers.
  • Building footprint GIS layers.
  • Development Application (DA) registers.
  • The property land uses used to decide Local Government land rates.
  • The number of property connections/meters.
  • Internet real estate web sites.
  • Drive-by of properties to photograph and record the details.

We find aerial photography and Google Street View to be our greatest friend with this task. It is worth purchasing the latest aerial photography if it is available, as the Google photography can be quite dated in many locations.

The challenge you will typically face in establishing the baseline is detecting and including recent development that has occurred (i.e. in the last 6 – 12 months).  You need to be on the lookout for vacant land created after completion of a sub-division and now has a dwelling or commercial building constructed on it.

Identify land with no development potential

Typically a large portion of properties in an urban growth model are fully developed and have no further development potential. For example: Low Density Zoned residential land with a single house.  We identify these properties and flag them in the baseline as land that cannot be developed further. This dramatically reduces the number of properties that our urban growth model needs to deal with. Fortunately, Forecaz Modeller provides functions that automate much of this task for us.

Some context

While this might seem a great deal of work, let us put some context around this task.  Here are some average statistics for various regional urban growth models we have worked on:

  • 4%-5% of the properties are located in areas that cannot be developed
  • With the remaining properties, 83-85% have no further development potential.
  • This leaves only 11-13% of properties that can be developed and contribute growth to the model.

Drilling down on the properties the can be developed:

  • 55-57% of the properties have only residential uses (6-7% of all properties in region)
  • 22-25% of the properties have only non-residential uses (2-3% of all properties in region)
  • 20-21% of the properties have both mixed residential and non-residential uses (2-3% of all properties in region)

These statistics show by focusing your efforts on the 11-13% of properties that have development capacity, will give you give you a superior baseline for your urban growth model. People are surprised by how many parcels cannot be developed or have no further development potential within an urban growth model. Also you will find determining the attributes of land parcels with residential uses is a lot easier than determining non-residential attributes.   These two factors help reduce the effort required to establish the baseline.

Hopefully describing the process we go through in establishing a baseline demonstrates the task is not a challenging has you might first think it is.

We are always interested in other people’s approach to establishing their baselines.  If you have some time, please outline your method in a comment below.


About the Author:
Bradley is the founder and Managing Director of Sizztech. He is an enterprise ICT consultant with over twenty years of experience in the Information Technology industry. In the last eight years Bradley has specialised in delivering web based geo-spatial solutions that focus on supporting capital works planning and delivery. This has lead to Sizztech developing Forecaz Modeller, an automated tool that performs urban growth modelling and forecasting.

Leave a reply