Residential intensification in a suburban fringe local government area, Casey, Melbourne Metropolitan area, Australia.

Phan, Thu ; Peterson, Jim ; Chandra, Shobhit 等

1. INTRODUCTION

Many cities and metropolitan areas are changing form due to enforcement of urban consolidation policies. These policies have been designed to reduce sprawl, and preserve farmlands and/or designated open spaces on the city fringe, to seek for efficient uses of infrastructure and less car-dependence, and to improve quality of life. They have featured in planning policy debates, especially after the adoption of sustainability as a guiding principle in the 1990s. While some researchers offer support for a certain level of higher density development (Buxton and Tieman, 2005), critics of higher density development have been able to refer to social and cultural arguments, including one about infill development in the existing urban areas threatening the status of prevailing residential streetscapes (Birrell et al., 2005b). Additionally, it is noted that the adequacy of current infrastructure and services can be less than needed to accommodate increases in population and dwelling numbers (Troy, 1996, Mees et al., 2008). Burton (2000, p.1970) stated that one of the main problems in advancing debate about the compact development model is a lack of empirical evidence to support either claims or counter-claims. Talen (2003) added that although "smart growth theory" (the term usually used in the United States or Canada) has emerged, there is still a lack of quantitative measurement, representation and evaluation of urban form and its associated impacts. Thus neither formulation nor implementation of urban compact city policy may be as well served as advocates would want (Talen, 2003).

Urban consolidation has prevailed in urban planning policies in major Australian cities for over 20 years (Searle, 2003). Recently, urban consolidation has been integrated in state-wide strategic plans released between 2002 and 2005 in Melbourne, Sydney, and Adelaide (Randolph, 2006). These urban planning policies have surely been long-enough implemented for useful research results about their impacts to have emerged.

However, a detailed geography of settlement intensification has not yet been systematically identified (Buxton and Tieman, 2004, Birrell et al., 2005b, Holloway and Bunker, 2003), and therefore, spatial relationships between increased population density and local infrastructures and services, not to mention the associated social costs, are not well documented. Additionally missing, is a metropolitan scale (constituent city-by-city), appraisal of the relative significance of intensified dwelling development forms. Detailed datasets (such as at land parcel level of mapping), have been called for, so that monitoring and management of urban development can take place at the local as well as at the metropolitan scale of mapping. As Phan et al. (2008) demonstrate, integration of selected spatial datasets can offer a geography of residential infill development that is detailed enough for decision support at local government level. In this paper, we present a result of applying that methodology to identify location of residential intensification in the City of Casey, Melbourne, Australia between 2001 and 2006. We chose the study period as of 2001-2006 for a number of reasons:

1. its average population annual growth rate of 1.3 percent, which is higher than that of 1.1 and 0.6 percent in 1996-2001 and 1991-1996, respectively (DSE, 2007a); meaning that dwelling development could increase to accommodate the population growth

2. the period covers the time before and after the implementation of the recent Victorian state strategic planning policy, Melbourne 2030, so that implications for urban planning policy can be discussed from the residential urban form changes documented.

3. the period covers two census dates: 2001 and 2006, so that many socioeconomic factors from the ABS can be incorporated and integrated in the analysis.

4. previous studies (Buxton and Tieman, 2005, Buxton and Scheurer, 2007, Mitchell, 1999, Birrell et al., 2005b) are old enough for the appraisal to be re-visited.

2. LITERATURE REVIEW OF RESIDENTIAL INTENSIFICATION AND URBAN FORM QUANTIFICATION

Generally, census data (the ABS census of people and housing) is the key dataset used by urban geographers in Australia. Aggregated census collection districts (e.g. suburban level or local government area level) (DPCD, 2007b, Randolph, 2006, Roberts, 2007, Yates, 2001) form the main analysis mapping unit. Using such aggregated datasets has three major disadvantages if urban intensification is defined as the increase in population or in number of dwellings per unit area. First, even the smallest census unit (census collection district: CCD)) is aggregated from 220 dwellings (ABS, 2006), thus it does not represent the precise location of features of urban intensification. Secondly, there are problems associated with aggregated data, such as the "modifiable areal unit problem" (Openshaw, 1983). Thirdly, in the census data from the Australian Bureau of Statistics (ABS), multiple detached or attached houses subdivided from previous residential land lots are classified as "separate house". Failure to be counted as higher density developments, the lack of these developments in the dwelling count underestimates of rate of urban consolidation (Buxton and Tieman, 2005).

Lately, a growing concern for documenting change patterns has seen recognition of the utility of temporal and spatial data of finer scale and the availability and accessibility of analytical tools such as geographical information systems (GIS) (Clifton et al., 2008). Input data can now include the cadastre (in time series) or fine-scale land use land cover (LULC) maps generated by the research team from analysis of increasingly available image data. However, studies have been limited to (1) manual plotting of infill features (Birrell et al., 2005b), (2) descriptive analysis of urban consolidation patterns, focussing on densification in existing urban areas in three municipalities in the middle and outer regions in Sydney (Holloway and Bunker, 2003, Bunker et al., 2002), and (3) comparative analysis of urban consolidation patterns for four inner municipalities in Melbourne (Buxton and Tieman, 2004).

Techniques to quantify and classify urban form have been advanced from a range of disciplinary perspectives (Clifton et al., 2008). Cuthbert and Anderson (2002) used a kernel estimate method to characterise residential and commercial land intensification in Halifax Regional Municipality. Song and Knaap (2007) presented an approach to characterise and classify neighbourhood for new single-family homes in the Portland (USA) area. Hahs and McDonnell (2006) used a gradient analysis approach (based on the distance from central business district) to study landscape differentiation spatially in Melbourne. Herold et al. (2003), using North American examples, applied a number of landscape metrics to examine changes in the spatial-temporal urban environment. Batty and Longley (1988) used a fractal analysis technique to measure urban morphology in a UK village, land parcel by land parcel, the perimeter and area of each land parcel treated as indicators of urban form. Buxton and Tieman (2004) examined the contribution of dwelling supply from greenfield developments on the fringe of the Melbourne Metropolitan under different development density scenarios.

Research about the influence of strategic urban land use policies on local and regional development in Australian metropolitan regions only recently emerged (Randolph, 2006). Many spatial datasets and techniques of analysis have been advanced; as a result, there is a scope for spatial data integration, analysis and visualisation of the patterns of residential intensification and its local impacts.

3. URBAN CONSOLIDATION-POLICY CONTEXT IN MELBOURNE IMPLICATIONS OF THE SURVEY FINDINGS

Previous studies (Buxton and Tieman, 2005, Birrell et al., 2005a) have provided comprehensive description and analysis of the evolution of state planning policy on urban development since 1970s. To set a context for our study, we summarise the key time periods marking the changes or transitions in urban planning policies in Melbourne either at local or state level. Generally, there are four main periods of planning policy changes in Melbourne in intervening and controlling the urban consolidation policy.

First, from the early 1970s to early 1990s, urban consolidation was controlled by each municipality under its "flat codes" (in terms of type, design, location and density). These were non-statutory and varied across municipalities. However, the state planning agency of the time, the Melbourne and Metropolitan Board of Works, introduced dual occupancy provisions into the Melbourne Metropolitan Planning Scheme (MMPS) in Amendment 150 in 1981 (Buxton and Tieman, 2005). Initially, dual occupancy was in the form of additional dwellings in the backyard of a suburban house, and later there were examples of house replacement by two detached houses (Birrell et al., 2005a).

Secondly, the Victorian Government introduced The Victorian Code for Residential Development (Multi-Dwellings), named VicCode 2, in December 1993 to control medium density developments (Buxton and Tieman, 2005). Under this policy, all metropolitan councils were required to have their provisions regarding planning permits for 3+ dwellings or dual occupancy in any urban zone or reserved land. VicCode 2 was reviewed in mid 1994 and The Good Design Guide for Medium Density Housing was then introduced to provide further requirements for local governments in approving medium density development applications. Specifically, a planning permit is required for a lot for the use and development of

* 2+ dwelling development, other than a moveable dwelling unit, not exceeding four storeys, or

* 1 dwelling on a lot less than 300[m.sup.2]

Thirdly, Rescode, a medium density code, was introduced by the Victorian Government in 2001 to become a statutory tool for controlling medium density development. Requirements and standards are given for each dwelling development application category, namely:

* Permit not required

* Single dwellings

* Multiple dwellings on a lot, and

* Subdivision.

Rescode was developed based on previous planning tools: VicCode 1 and The Good Design Guide (Buxton and Tieman, 2005). It added some new standards, though some of them were not statutory. Historically, the success of urban planning policy was little due to "the discretionary nature of much content, the use of qualitative measures and the lack of clarity" (Buxton and Tieman, 2005, p.141).

Fourthly, in October 2002, the Victorian State Government introduced its thirty-year strategic urban land use planning policy: Melbourne 2030. It was released as a strategic plan for accommodating a projected (2030) MMA population increase of one million while improving economic efficiency, the environment and maintaining the community's liveability. This strategic plan seeks to change radically the traditional pattern of Melbourne's low density urban form to "a more compact city" model, stated as Principle one in the Melbourne 2030 document. Principle one aims to (DoI, 2002. p.45):

Locate a substantial proportion of new housing in or close to activity centres and other strategic redevelopment sites that offer good access to services and transport.

Specifically, according to Melbourne 2030, location of new residential housing should be (DoI, 2002, p.57):

* In or around the central activities district (CAD);

* In or within easy walking distance of Principal or Major Activity Centres;

* In or beside Neighbourhood Activity Centres that are served by local public transport;

* Abutting tram, train, light rail and bus routes that are part of the Principal Public transport network and close to principal or major activity centres;

* In or near major modal public transport interchanges that are not in principal or Major Activity Centres; and/or

* In major redevelopment sites, that is able to provide 10 or more dwelling units, close to activity centres and well-served by public transport. Being a fringe city (that is 40km from the CBD), it is the last five points above that apply to the City of Casey.

Additionally, the first audit of the strategic plan, Melbourne 2030, was undertaken in 2006 and 2007 in two stages. Stage one was conducted by Department of Planning and Community Development (DPCD), an agency looking after the implementation issues of Melbourne 2030. This stage involved (1) meetings between DPCD with senior planning staff in each of 31 Local government councils, (2) and a report: Melbourne 2030 Audit: Analysis of Progress and Findings from the 2006 Census, which presents development trends and patterns from the 2006 census data, reviews progress in implementing Victorian Government actions to put Melbourne 2030 into practice, and summarises the feedbacks from key stakeholders. After that, stage two was conducted by four independent experts, which considered outcomes from stage one, submissions from all stakeholders and provided advice to the Minister for Planning on the ongoing implementation of Melbourne 2030 (DPCD, 2007b). It is found that there is a little evidence of achievement in terms of plan design objectives in the last five years (2001-06). In particular, a contrasting trend between policy and practice in household development is seen. For example, Greenfield development was recorded at 48.3 percent of household growth between 2001 and 2006 whereas the policy target for Greenfield development is 31 percent of total household growth by 2030. The audit report suggests that a ten year timeframe might be long enough for progress from plan to implementation to be seen (DPCD, 2008, p.40 & p.59). However, Buxton (2008) asserted and argued that radical actions need to be undertaken and specific measures should have been recommended in order to meet key principles of Melbourne 2030. They refer to redirection of planning from car-based, single use suburbs dominated by detached housing on the urban fringe, because such plans lead to high energy use and affordability problems and a need for continuous public transport extension.

In this study, to assess the influence of Melbourne 2030 on locations and patterns of residential infill development, increased dwelling numbers (per 2001 land parcel) will be identified and their relative location vis a vis designated activity centres, public transportation networks and selected services are examined and analysed.

Although the concept of higher density development is not clearly defined in Melbourne 2030, it is generally agreed by researchers that higher density development relates to population intensification, and that it can be seen by the increased number of dwellings, built as (1) redevelopments on land parcels that were previously occupied by a single dwelling with a generous garden space, (2) semi-detached or attached housing in single or multiple storeys. In other words, land supply for higher density development includes (1) already established residential areas, (2) brownfields, and (3) greenfields (Buxton and Tieman, 2004). The actual development pattern in terms of location, extent and dwelling types has been found to vary by each local government area (Buxton and Tieman, 2005). As such, analysis should be undertaken for each local government area before an attempt is made to document the metropolitan pattern of urban form change. In this study, we particularly examine the location and extent of residential intensification in the first and the third forms.

It is generally agreed that the decision support that is served by mapping residential urban form change refers to:

(1) monitoring the current state of urban intensification for each local government area;

(2) examining the driving forces of urban consolidation: planning policy, market influences on location, development types and housing prices, residential reactions and opposition; and historic development legacy in each local government area (Bunker et al., 2002); and

(3) examining need for infrastructure improvement or any risk associated (e.g. increased stormwater runoff or flooding) in areas which are subject to higher infill development.

To complement a previous study (Phan et al., 2008) of part of the MMA that has long been a built-up area, this paper presents results of mapping and examining the growth in the number of dwellings in a city much closer to the MMA boundary: the City of Casey.

4. STUDY AREA

The City of Casey (813.3 [km.sup.2]) is one of most rapidly growing regions in the fringe regions of the MMA (Figure 1). Early European settlement occurred in the late 1830's, at much of the same time as settlement of the Melbourne CBD. Over many years, pastoral activities dominated the local economy. Any local urban development referred to neighbouring shires, Cranbourne and Berwick (proclaimed as cities in 1993 and 1994, respectively). The modern extent of the City of Casey refers to a 1994 amalgamation from the former City of Berwick, the former City of Cranbourne, (formed in the late 1860s) and a small part of the City of Knox (City of Casey, 2008).

The City of Casey is one of the key growth areas in the MMA. Its population increased by 37,481 persons (21 percent) between 2001 and 2006 (calculated from the ABS data (ABS, 2007)). In absolute terms, Casey showed the largest Australian inter-censal population growth (2001 and 2006) (DSE, 2007a). The concomitant dwelling number increases are attributed to many factors, some of which are: strong national economy, growth in household incomes, low interest rates, and population growth (DPCD, 2007a).

The City of Casey is located in the Southern Region in the MMA (Figure 1). Regional environmental amenity and liveability attributes include proximity to the Dandenong Ranges, bays and beaches, rural areas, and accessibility to the townships and settlements within green wedges and developments such as golf courses (Southern Regional Housing Working Group, 2006).

Figure 2 shows a gradual increase in the number of private dwellings between 1981 and 2006, at a five year interval, in selected local government areas (LGAs) in the East, North and West Regions of the MMA: Casey, Moreland, and Wyndham, respectively. Casey clearly shows the highest number of new dwellings between 2001 and 2006; but at a decreasing rate between 2002 and 2006 (Figure 3). In terms of dwelling structure, it can be seen from Figure 4 that such a rapid increase in total dwelling numbers can be attributed to a big increase in the number of separate houses. Medium density development (in the form of semi-detached, row or terrace house; and flat, unit, or apartment), also shows a growth; but it is at a lesser rate.

5. METHODOLOGY

The primary datasets used to identify the pattern of residential intensification in the City of Casey (Figure 5) include planning scheme maps, cadastral maps, and address point data. They are provided by Department of Sustainability and Environment (DSE) and used for identification of residential land and examination of residential intensification, respectively. The 2005 aerial photo mosaic was provided by Casey City and, together with Google Maps (http://maps.google.com/) (aerial photos were archived on 16 February 2006), was used for data quality checking and results validation.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 4 OMITTED]

The cadastre is a register of the precise location, extent, use and ownership of land, these being non-spatial attributes in land parcel datasets. It is maintained, for among other reasons, for taxation purposes (DSE, 2006). The Vicmap Address dataset (VADD) is a fully geocoded digital street-address dataset. The records include the spatial relationship of each address to the relevant polygon in the cadastre (DSE, 2007c). Urban addresses in the MMA are assigned 8 metres back from the property road frontage mid-point (DSE, 2007c). Unit and house number are assigned by each Local Government Area either after approval of land subdivision, plan registration at Land Registry or at the time of sale of the property (Salmon, 2008). VADD is regularly validated with local government property records with 90% of completeness and accuracy (Salmon, 2008). Because in 2006 all dwelling non-spatial attributes (for instance, unit number, house number, or street name) are in VADD, we used both the property dataset and VADD for analysing the 2006 dwelling patterns. Other datasets, such as planning schemes and land mark datasets (including non-residential properties, such as abattoirs, camping grounds, caravan parks, car parks, cemeteries, parks, recreation areas, showgrounds, and sports areas) (DSE, 2007b), were also used to differentiate residential land parcels from other non-residential land parcels.

In this study, we mapped the location of increased dwellings in each 2001 land parcel (Figure 5). They were then classified by the number of increased dwellings between 2001 and 2006 in five classes:

* 1-2 dwellings;

* 3-10 dwellings;

* 11-36 dwellings;

* 37-76; and

* 77+ dwellings.

The first class (1-2 dwellings) represents dual occupancy dwelling stock change while the middle class (such as 3-9 dwellings) can represent medium density development in the form of multi attached units or townhouses and apartments. The final class (77+ dwellings) represents a mix of dwelling development types. In this study, we consider urban consolidation as the process of increased dwelling density over the land parcel unit, like other studies (Birrell et al., 2005b, Phan et al., 2008), therefore we regard the uses of primary datasets (i.e. cadastre data and VADD) to count the increased dwelling number per 2001 land parcel as reasonable.

[FIGURE 5 OMITTED]

Proximity analysis has been widely deployed in analyses of access to facilities: for example, proximity has been used as an explanatory variable to examine the influence of transportation network on location and prices of housing stock (Cho et al., 2005) or to examine the influence of planning policy on the pattern of residential development (Buxton and Tieman, 2005). We used a proximity analysis tool (buffer in ArcGIS 9.2 (ESRI, 2007)) to analyse the extent and amount of higher dwelling density development around public transport networks and designated activity centres.

6. RESULTS AND DISCUSSION

6.1. Extent and scale of current residential intensification

Figure 6 shows the location, extent and amount of urban consolidation pattern in the City of Casey in terms of increased dwelling numbers per 2001 land parcel. Only 10.6 percent (970 in total) of them refer to dual occupancy or infill of residential land parcels: 55 percent of the new dwellings refer to Greenfield developments. The rest is of other categories, including multi attached units or multi storey dwelling complexes (Table 1). As shown in Table 1, the green field developments (resulting in subdivision of large number of dwellings: 77+) account for most of the land cover change.

These indicate that much residential intensification in the City of Casey occurred as urban sprawl, mostly the transformation of a Greenfield site to residential land use featuring multi attached and detached houses or units/flats. This is a contrast to the results of an analysis of City of Monash dwelling intensification (Phan et al., 2008) that showed that dual occupancy or redevelopment of old housing stock for replacement with two to seven dwelling units was the most common change. This reflects the different or even contrasting patterns of residential development between the fringe and suburban areas. Indeed, the suburban area, City of Monash, has been developed since the 1960s. As a result, in 2001-06; there are not many big land parcels available for further development or subdivision. In contrast, the young fringe area: City of Casey, is far from the Melbourne CBD, relatively lacking in access to public transportation networks, and, even today, has land cover that is mostly agricultural. Thus, until recently, this area has been less preferable for settlement than areas closer to the CBD. Previous studies for municipalities in middle and outer regions of Sydney also found a similar trend: increased dwelling density mostly occurs in small land parcels of mature cities whereas less mature or younger residential areas tend to have Greenfield sub-division during phases of residential intensification (Bunker et al., 2002).

[FIGURE 6 OMITTED]

In section 3, data analysis from the ABS, shows that the category "separate house" accounts for a big proportion of new dwellings in the City of Casey in the 2001-2006 period. It is observed on the aerial photos mosaics (in 2005 from Casey City Council and in 2006 viewed in Google Earth 4.3) that the "separate house" class in Casey includes both detached and attached single houses. Although the method presented in Figure 5 allows defining precise location of residential intensification in terms of increased dwelling number per 2001 land parcel, currently it does not offer differentiation between detached/attached single house and medium density development (such as multiple-storeys flat or apartment). In a previous study in Sydney, Holloway and Bunker (2003) could disaggregate residential intensification types: dual occupancy, small lot housing, housing for aged and disabled, townhouses and villas; and flats and units because these were the dwelling type attributes attached in their primary dataset (i.e. development applications provided by local councils).

6.2. Overview of Demographics and Housing Prices trend

A snapshot of City of Casey demographic changes over three census periods: 1996, 2001 and 2006 (Figure 7) shows that young people aged 5-14 years old and middle aged people aged 35-44 years old, account for the largest proportion of population in the City of Casey. In fact, these two age groups represent young families. The dominance of young families might reflect the nature of housing affordability for first home buyers. It is reported that between 1991 and 2005, house and apartment prices in Casey increased at 137 and 149 percent, respectively (Southern Regional Housing Working Group, 2006, p.29). This rate is about 10-15 percent lower than those for the whole MMA and is much lower than those of Inner or coastal local government areas. However they are slightly higher than those of adjacent local government areas, such as Cardinia, Frankston or Greater Dandenong.

[FIGURE 7 OMITTED]

Being an area with no more than moderate housing price increase, and offering new houses via development of Greenfield areas, Casey City planners have had to decide which housing density value might meet living and open space requirements. Buxton and Tieman (2004) suggested that increased development density on the fringe from 10 dwellings/ha to 15 dwelling/ha can save a substantial amount of land while at the same time giving greater density of access to the public rail network. Such theoretical analysis assumes the provision of sufficient and accessible infrastructure.

6.3. Accessibility to public transport networks

It is generally agreed that settlement locations or transit locations should be within walking distance of the public transport network (up to 800m, which is equivalent to a 10 minute walking). In the City of Casey, there are two railway lines and a number of principal bus line networks to connect local residents to activity centres and other places of interest. Proximity analysis of new dwelling development between 2001-2006 in Casey (Table 2 and Table 3) shows that within 400m of public transport, there is mostly a small number of increased dwellings per 2001 land lot (i.e. 1-2 dwellings). In terms of dwelling supply, larger numbers of increased dwelling per 2001 land parcel (77+ dwellings) provides a significant proportion of the dwellings, particularly in areas close to railway lines. In an aggregate number, the proportion of dwelling supply from areas within 400m of bus lines and railway lines accounts for only about 10-12% of the total dwelling supply in the 2001-06 period. These figures suggest that a large proportion of new dwellings are still disadvantaged in access to the public transport network. A recent study found that of major Australian cities, Melbourne saw the largest increase in car driving and the largest declines in car pooling, public transport and walking for travelling to work over the last three decades (1976-2006) (Mees et al., 2008). Therefore, without improvements in public transportation networks, intensified dwelling development and corresponding likely increased population settlement in the fringe city areas such as Casey might increase the number of cars involved in the journey-to-work, not to mention other activities. This subsequently threatens the success of urban compact city policy.

6.4. Accessibility to activity centres

Activity centres are classified under a hierarchical system, in which Principal Activity Centres (PACs) play the more significant role than others in provision of and accessibility to services for the wider extent of neighbourhood communities (Goodman and Coote, 2007). There are six local activity centres proposed in Melbourne 2030, including:

* Principal Activity Centres (PAC): Cranbourne and Narre Warren, Fountain Gate; and

* Major Activity Centres: Casey Central, Berwick, Hampton Park, and Endeavour Hills.

According to Melbourne 2030, the location of Activity Centres will influence and facilitate the development of different housing types, including forms of higher-density housing (DoI, 2002). Proximity Analysis (Table 4) shows that most infill development within 800m of a designated Activity Centre refers to small increased dwelling numbers per 2001 land parcel (i.e. 1-2 dwellings). It provides only about 2 percent of the total number of new dwellings from infill development in Casey in 2001-2006. Altogether, infill development within 800m of Activity Centres in Casey City accounts for only 6 percent of total new dwellings in 2001-06. This again indicates that new dwelling development is not necessarily close to designated Activity Centres as preferred in the Melbourne 2030 objectives. In fact, dispersed infill subdivision and green field developments are the two main types of high density development in City of Casey between 2001 and 2006.

6.5. Proximity to other services

A place of interest database (e.g. church, schools, or shopping centres) is not available in the current Victorian Spatial Dataset. However, places of interest are freely viewed in Google Earth. These datasets are provided by TrueLocal[TM]. We selected all places of interests in the City of Casey (schools and hospitals) and then transformed them to a ArcGIS format. The Hawth's Analysis Tool (Beyer, 2004), an extension of ArcGIS, was then used to calculate the nearest distance between each address point in areas of new dwelling developments to those places of interest. Table 5 shows that nearly half of the new dwellings (48 percent) are within 1 km of their nearest schools. Locations of hospital or medical centres are further from new dwellings than schools: about 35 percent of new dwellings are outside the 2km buffer of any nearest hospital or medical centre.

Although our analysis currently does not disaggregate the proportion in terms of school types, or medical centre or hospital types, these figures still provide a certain measure of infrastructure provision to serve the increased dwelling density. To improve accessibility to these services by walking or using public transport, infrastructure needs to be upgraded or improved in areas of those residents living outside the 2km buffer of any nearest hospitals or medical centres.

7. CONCLUSION

Integration of spatial datasets: cadastres, planning schemes, and aerial photos allow identification and visualisation of patterns of residential subdivision between 2001 and 2006. This case study from a portion of the expanding MMA fringe shows a marked contrast to the pattern of urban residential densification that characterised another city (Monash) twenty km closer to the CBD but responding to the same urban development/re-development policy. Proximity Analysis: 400m buffer of the public transport network (railway and bus lines) and 800m buffer of Activity Centres in proximity analysis showed that between 10 and 6 percent of new (2001 and 2006) dwellings developed within buffers of the public transport networks and Activity Centres, respectively. Additionally, nearest distance analysis found that 95 percent of high density developments are within 1km of the nearest schools while about 35 percent of them are outside the 2km buffer of the nearest hospitals. These figures imply that the ultimate goal of Melbourne 2030 to plan for new development while maintaining liveability of residents has not yet been achieved. According to Whitzman (2008), liveability means housing affordability, tackling disadvantage, low crime rates, a cosmopolitan or multicultural way of life, and 'room to grow'. Thus, there is clear scope to bring policy and practice closer together, for instance, by improving access to infrastructure. Part of any programme to bring this about could involve encouragement for infill development near existing infrastructure. Furthermore, multiple storey attached units should be developed in established urban areas in the fringe to increase the number of house supply while reducing the amount of land consumption (Buxton, 2008, Buxton and Scheurer, 2007).

ACKNOWLEDGEMENTS

The authors would like to thank Faculty of Arts and Monash Graduate Research School, Monash University for financial support of this project. We thank Centre for GIS, School of Geography & Environmental Science, Monash University for providing necessary facilities and datasets. Also, great thanks to staff at Victorian Spatial Data Infrastructure, Department of Sustainability and Environment and the City of Casey for granting access to data and for their help in understanding the data lineage and structure. Comments from the reviewer in the early version of this paper are also greatly appreciated.

REFERENCES

ABS (2006) 1216.0 Australian Standard Geographical Classification (ASGC), 2001, Spatial Unit Accessed 8 February 8 May 2006 Available at http://www.abs.gov.au/ausstats/[email protected]/0/B03C37D9D89F7C19CA25 6AD4007F67F8?Open&Highlight=O,urban

ABS (2007) 2006 Census Community Profile Series Time Series Profile Casey (C) (LGA 21610).

Batty, M. and Longley, P. A. (1988) The morphology of urban land use. Environment and Planning B: Planning and Design, 15, pp. 461-488.

Beyer, H. L. 2004 Hawth's Analysis Tools for ArcGIS http://www.spatialecology.com/htools/tooldesc.php.

Birrell, B., O'Connor, K., Rapson, V. and Healy, E. (2005a) Melbourne 2030: Planning Rhetoric Versus Urban Reality, Monash University ePress.

Birrell, B., O'Connor, K., Rapson, V. and Healy, E. (2005b) Residential infill and its threat to Melbourne's liveability. Melbourne 2030: Planning Rhetoric Ijersus Urban Reality. Monash University ePress.

Bunker, R., Gleeson, B., Holloway, D. and Randolph, B. (2002) The Local Impacts of Urban Consolidation in Sydney. Urban Policy and Research, 20, pp. 143-167.

Burton, E. (2000) The Compact City: Just or Just Compact? A Preliminary Analysis. Urban Studies, 37, pp. 1969-2001.

Buxton, M. (2008) Why deniers can't be implementers The Melbourne 2030 Audit and the goverment response. Planning News, 34 (6-9), p. 28.

Buxton, M. and Scheurer, J. (2007) Density and Outer Urban Development in Melbourne. Urban Policy and Research, 25, pp. 91-111.

Buxton, M. and Tieman, G. (2004) Urban Consolidation in Melbourne 19882003: The Policy and Practice. RMIT Publishing: Melbourne.

Buxton, M. and Tieman, G. (2005) Patterns of Urban Consolidation in Melbourne: Planning Policy and the Growth of Medium Density Housing. Urban Policy and Research, 23, pp. 137-157.

Cho, S.-H., English, B. C. and Roberts, R. K. (2005) A spatial Analaysis of Housing Growth. The Review of Regional Studies, 35, pp. 311-335.

City of Casey (2008) History of the Casey region Accessed 15 October Available at http://www.Casey.vic.gov.au/history/article.asp?Item=919

Clifton, K., Ewing, R., Knaap, G.-J. and Song, Y. (2008) Quantitative analysis of urban form: a multidisciplinary review. Journal of Urbanism: International Research on Placemaking and Urban Sustainability, 1, pp. 17-45.

Cuthbert, A. L. and Anderson, W. P. (2002) An examination of urban form in Halifax-Dartmouth: Alternative approaches in data. Canadian Journal of Urban Research, 11, pp. 213-237.

DoI (2002) Melbourne 2030: Planning for sustainable growth. Department of Infrastructure: Melbourne.

DPCD (2007a) Census 2006 Info Sheet 3: Dwellings change in Melbourne Accessed 15 January Nov 2007 Available at http://www.dse.vic.gov.au/CA256F310024B628/0/SBA2EBF4F8F58B0 1CA257341001D8D79/$File/DPCD+3-Dwellings+change.pdf

DPCD (2007b) Melbourne 2030 Audit: Analysis of Progress and Findings from the 2006 Census. Department of Planning and Community Development.

DPCD (2008) Audit Expert Group Report. Department of Planning and Community Development, Accessed 31 Jan 2008, Available at http://www.dse.vic.,gov.au/CA256F310024B628/0/E3FA9C74C4FA6A 22CA25744E0004719F/$File/Audit+Expert+Group+Report.pdf,

DSE (2006) Vicmap Property-Metadata Accessed 29 Jan. Available at http://www.giconnections.vic.gov.au/content/vicgdd/record/ANZVI080 3002683.htm

DSE (2007a) Summary of population change 2001-2006 Accessed 15 January. Available at http://www.dse.vic.gov.au/CA256F310024B628/0/200181861 F6398A4 CA2573E0001 A5234/$File/Census+2006+ +Summary+of+population+change.pdf

DSE (2007b) Victorian Spatial Data Directory Landmark Area Polygon 1:25,000--Vicmap Features (IN LANDMARK AREA POLYGON/). Department of Sustainability and Environment, http://www.land.vic.gov.auNSDDcategory.htm.

DSE (2007c) Victorian Spatial Data Directory Vicmap Address (VICMAP ADRESS/) Metadata. Department of Sustainability and Environment, http://www.land.vic.gov.au/VSDDcategory.htm.

ESRI (2007) ArcGIS 9.2 http://www.esri.com Goodman, R. and Coote, M. (2007) Sustainable Urban Form and the Shopping Centre: An Investigation of Activity Centres in Melbourne's Growth Areas. Urban Policy and Research, 25, pp. 39-61.

Hahs, A. K. and McDonnell, M. J. (2006) Selecting independent measures to quantify Melbourne's urban-rural gradient. Landscape and Urban Planning, 78, pp. 435-448.

Herold, M., Goldstein, N. C. and Clarke, K. C. (2003) The spatiotemporal form of urban growth: measurement, analysis and modeling. Remote Sensing of Environment, 86, pp. 286-302.

Holloway, D. and Bunker, R. (2003) Using GIS as an Aid to Understanding Urban Consolidation. Australian Geographical Studies, 41, pp. 44-57.

Mees, P., O'Connell, G. and Stone, J. (2008) Travel to Work in Australian Capital Cities, 1976-2006. Urban Policy and Research, 26, pp. 363378.

Mitchell, K. (1999) Dual Occupancy and Its Impact on Metropolitan Growth in Melbourne Metropolitan Growth in Melbourne (1986-1992). Master of Arts in Town Planning Faculty of Arts. Melbourne, Victoria University of Technology.

Openshaw, S. (1983) The modifiable areal unit problem: Concepts and techniques in modern geography. Geobooks: Norwich.

Phan, T. H., Peterson, J. A. and Chandra, S. (2008) Urban infill: extent and implications in the City of Monash. People and Place, 16, pp. 23-31.

Randolph, B. (2006) Delivering the compact city in Australia: Current trends and Future Implications. Urban policy and research, 24, pp. 473-490.

Roberts, B. H. (2007) Change in Urban Density: Its Implications on the Sustainable Development of Australian Cities. In Forster, C. (Ed.) Proceedings of the State of Australian Cities National Conference 2007 (SOAC).

Salmon, C. (2008) VICMAP Address Custodian Spatial Information Infrastructure, Information Business Technology Department of Sustainability and Environment Email Communication on 25 June 2008

Searle, G. (2003) The limits to Urban Consolidation, Urban Frontiers Program. Issues Paper No. 14. University of Western Sydney.

Song, Y. and Knaap, G.-J. (2007) Quantitative Classification of Neighbourhoods: The Neighbourhoods of New Single-family Homes in the Portland Metropolitan Area. Journal of Urban Design, 12, 1-24.

Southern Regional Housing Working Group (2006) Southern Regional Housing Statement.

Talen, E. (2003) Measuring Urbanism: Issues in Smart Growth Research. Journal of Urban Design, 8, pp. 303-303.

Troy, P. (1996) The Perils of Urban Consolidation, Sydney: Foundation Press.

Whitzman, C. (2008) Planning for living well. Planning News, 34 (9).

Yates, J. (2001) The rhetoric and reality of housing choice: the role of urban consolidation. Urban policy and research, 19, pp. 491-527.

Thu Phan

PhD Student, School of Geography and Environmental Science, Monash University, Melbourne VIC 3800.

Jim Peterson

Associate Professor, School of Geography and Environmental Science, Monash University, Melbourne VIC 3800.

Shobhit Chandra

Professional Officer, School of Geography and Environmental Science, Monash University Melbourne VIC 3800.

(1) This paper was presented at the 32"a ANZRSAI Conference held in Adelaide from 30t' Nov-3rd Dec 2008.

Table 1. Summary of residential intensification in City
of Casey, 2001-06.

Class             Count   % of occurrence  Dwelling
                          of subdivision   supply

1-2 dwellings       970             69.3       2104
3-10 dwellings      201             14.4       1260
11-36 dwellings     130              9.3       2908
37-76 dwellings      50              3.6       2751
77+ dwellings        48              3.4      10862

                    % of total     Total
                      dwelling     block
                       supply      size (ha)

1-2 dwellings           10.6          392.4
3-10 dwellings           6.3          332.7
11-36 dwellings         14.6          448.8
37-76 dwellings         13.8          255.4
77+ dwellings           54.6         1356.8

Notes: Residential intensification is classified by number
of subdivided dwellings.

Table 2. Infill development within 400m of principal public transport
network bus lines and its proportion of total infill development in
the City of Casey in 2001-06

                                                      % of total
                         % of occurrence   Dwelling   dwelling
Class            Count    of subdivision   supply     supply

1-2 dwellings      299              21.4        659          3.3
3-10 dwellings      55               3.9        306          1.5
11-36 dwellings     22               1.6        424          2.1
37-76 dwellings      5               0.4        280          1.4
77+ dwellings        4               0.3        760          3.8
Sum                385              27.5       2429         12.2

Table 3. Infill development within 400m of railway lines and its
proportion of total infill development in the City of Casey in 2001-06

                          % of occurrence  Dwelling   % of total
Class             Count    of subdivision  supply     dwelling
                                                      supply

1-2 dwellings       107              7.6        238          1.2
3-10 dwellings       12              0.9         70          0.4
11-36 dwellings       2              0.l         50          0.3
37-76 dwellings       4              0.3        201          1.0
77+ dwellings         8              0.6       1412          7.1
Sum                 133              9.5       1971          9.9

Table 4. Infill development within 800m of Activity Centres and its
proportion of total infill development in the City of Casey in 2001-06

                                                       % of total
                          % of occurrence   Dwelling    dwelling
Class             Count   of subdivision     supply     supply

1-2 dwellings       164              11.7        369          1.9
3-10 dwellings       17               1.2         93          0.5
11-36 dwellings       4               0.3         71          0.4
37-76 dwellings       3               0.2        167          0.8
77+ dwellings         3               0.2        478          2.4
Sum                 191              13.7       1178          5.9

Table 5. Nearest distance between new dwelling developments
to places of interest

            Hospitals/      %         Schools    %
            Medical
Distance    Centres

<500m             1401       6.3        3823      17.3
500m-1km          3066      13.8        6955      31.4
lkm-2km           9866      44.5       10327      46.6
2km+              7819      35.3        1047       4.7

Figure 3. Total dwelling units approved in City of Casey between 2002
and 2006

2002   4114
2003   3711
2004   3250
2005   2435
2006   2406

Source: Australian Bureau of Statistics (ABS), Time Series Dataset,
Casey Local Government Area (ABS, 2007).

Note: Table made from bar graph.