A carregar...
Publicação
Investigating the use of dasymetric techniques for assessing employment containment in Melbourne, Australia
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 4.3 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
This project studies employment containment in Melbourne, Australia.
Employment containment is a measure of the proportion of people that work in a
location close to their home. Recent urban planning policies in Melbourne have aimed
to improve employment containment in the city’s suburbs. While there has been
analysis of the rates at which people both live and work within broadly defined ‘local
areas’, little work has been done to investigate employment containment using smaller
and more uniform catchment areas as the unit of analysis. This research attempts such a
finer scale analysis using dasymetric downscaling techniques. A regression modelling
approach supported by land use data, alongside a binary dasymetric method, is used to
develop fine scale estimates of employment distribution, while binary and populationdensity
weighted methods are used to develop a fine scale estimate of working
population distribution. For the employment distribution estimate, the Poisson model
that distributed employment to employment-related land use classes produced the
smallest error. However, the error produced by this model is still high. For the working
population distribution estimate, the population-density weighted estimate is the more
accurate of the approaches, and overall produced low error. For the employment
containment analysis, a number of employment centres were randomly selected and an
employment containment catchment has been derived from a 5 km2 commuting
distance catchment. Commuting flows from an origin-destination matrix were areaweighted
to estimate flows into the employment centre from the 5 km2 catchment. The
method is found to be potentially useful; however inspecting the results of this
employment containment calculation highlighted flaws in the current estimates that
should be addressed before the measures can be used to further analyse employment
containment in Melbourne. Improvements to this method would support urban strategic
and transport planning analyses at a metropolitan-wide scale.
Descrição
Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies.
Palavras-chave
Australian Bureau of Statistics Central Business District Change of Support Problem Geographically Weighted Regression Local Government Area Modifiable Areal Unit Problem Ordinary Least Squares Root Mean Square Error Statistical Local Area