Housing Development Data

Indroduction

Housing Development Data (HDD) is a spatial dataset containing a highly accurate count of existing dwellings, vacant residential lots and residential developments. The data is collected annually on a lot by lot basis through an analysis of various spatial datasets including aerial photography, property boundaries, business registers, planning permit information and other key databases.

The data has been prepared for all individual lots across Melbourne where there is an existing dwelling or where the lot is located in a land use planning zone that provides for residential uses.

More information about the HDD dataset DPCD - Housing Development Data

Data collection

The principal data sets employed in this analysis are aerial photography and the cadastre. Aerial photography for metropolitan Melbourne is commissioned by the Victorian Government on a yearly basis and has been used in this analysis. The aerial photography was generally taken around late December of each year.

Customised GIS software has been developed to visually recognise built structures and therefore has the additional ability to recognise vacant allotments. The software has the ability to recognise via colour intensity and colour distribution associated with built structures and straight lines created by roof- lines. This combined with titling information allows the recognition of both built structures on a parcel by parcel basis, as well as the incidence of vacant lots.

< > Source orthorectified aerial imagery

Step 1. Source orthorectified aerial imagery, (35cm/pixel)
Step 2. Aerial imagery with detected edges
Step 3. Combining imagery with vector cadastral layer
Step 4. Processing built structures analysis
Step 5. Final HDD Stock layer with labelled information about total dwellings

Data assessment

Collected data for individual years is stored in stock layers. However comparing these stock layers for changes between consecutive years results in additional information about constructions and demolitions.

To illustrate how changes are calculated in project layers, an example polygon located in Reservoir is shown. An interactive map demonstrates the residential development activity within the 'Example Area' polygon. The stock of dwellings as well as constructions and demolitions are shown in the table.

To see the location of 'Example Area' in Google maps click here :  Example Area in Housing Construction Map.

Stock Layer

 < > Construction of 1 dwelling

Year Dwellings Projects Constructions Demolitions Change detected
2004 44 0 0 0 No change
2005 46 2 4 2 Construction of 4 dwellings and 2 demolitons
2006 76 1 30 0 Construction of 30 mulitlevel apartments
2007 75 0 0 1 Demolition of 1 existing dwelling
2008 76 1 1 0 Construction of 1 dwelling
2009 78 1 2 1 Construction of 2 dwellings
2010 78 0 0 0 No change
Total n/a 5 37 3

In summary the example above showed that between 2004-2010 there were 5 construction projects with a total of 37 dwellings constructed and 3 dwellings demolished, resulting in a net increase 34 dwellings.

The pictures and data are taken from 'HDD Extractor' which has been designed to automatically generate these types of statistics for any geographic area.

A full set of statistics generated for the 'Example Area' by HDD Extractor can be downloaded. It has more additional statistics about the stock of dwellings, vacant lots and development activity on an annual basis. Click here to download : HDD_04-10_for_Example_Area.xls

There are 4 types of projects that are illustrated in this example:

  • Replacement project - Old dwelling replaces new dwelling (Constructed 2008)
  • Backyard Infill - One or more dwellings built in back of the original property (Constructed 2009)
  • Minor Infill - Two or more dwellings built after original dwelling has been demolished (Constructed 2005)
  • Major Infill - 10 or more dwellings built with high residential density (Constructed 2006)

    • Types of Dwellings

    All private dwellings are counted within the assessment and preparation of the HDD regardless of zone and dwelling type. However, not all 'potential' dwelling structures that could accommodate permanent residents or households are counted. Structures that have been excluded from the HDD include:

  • Improvised dwellings;
  • Caravans, huts etc within caravan parks;
  • Motels, hotels and hostels for permanent residential purposes;
  • Potential dwellings attached to retail premises (generally pre 1970s). Many “older retail” shops have either a potential dwelling attached or a shop-top. The vast majority of these are used for non-residential purposes such as storage, rather than inhabited by permanent households. As such the decision was made to exclude such properties from the calculation of the dwelling stock unless unambiguously used as dwelling/s
  • Student accommodation that could be described as 'on-campus student halls of residence'
  • Supported residential care facilities. However, non-supported residential care facilities are included (independent living dwellings)
  • Serviced apartments that are generally for short-term stays.

    • go up