Elements of Urban Form

Abstract

This chapter provides a common platform for the research presented in this book and is divided into two parts. The first section examines the elements of urban form identified for the purposes of the research and explains how they were measured. The second section profiles the five case study cities and fifteen case neighbourhoods which were the focus for the empirical research discussed in later chapters. These profiles provide an outline of both the urban form and socio-economic characteristics of the areas studied. The chapter concludes with a review of the urban form features of the case study cities and neighbourhoods, and shows how the different physical elements integrate together with socio-economic characteristics.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

This chapter provides a common platform for the research presented in this book and is divided into two parts. The first section examines the elements of urban form identified for the purposes of the research and explains how they were measured. The second section profiles the five case study cities and fifteen case neighbourhoods which were the focus for the empirical research discussed in later chapters. These profiles provide an outline of both the urban form and socio-economic characteristics of the areas studied. The chapter concludes with a review of the urban form features of the case study cities and neighbourhoods, and shows how the different physical elements integrate together with socio-economic characteristics.

Elements of Urban Form

The term ‘urban form’ can be used simply to describe a city’s physical characteristics. At the broad city or regional scale, urban form has been defined as the spatial configuration of fixed elements (Anderson et al., 1996). Features of urban form at this scale would include urban settlement type, such as a market town, central business district or suburbs. However, urban form is closely related to scale and has been described as the ‘morphological attributes of an urban area at all scales’ (Williams et al., 2000). Characteristics therefore range from, at a very localized scale, features such as building materials, façades and fenestration, to, at a broader scale, housing type, street type and their spatial arrangement, or layout.

It should be noted that urban form does not simply relate to physical features, but also encompasses non-physical aspects. One can see this in the example of density. Simply put, density is used as a measure of the number of people living in a given area: it is not just a physical, tangible element. Density is also closely linked with the configuration of the social environment and interaction within residential neighbourhoods: flats and apartments are examples of high-density housing whereas detached and semi-detached properties tend to be of lower densities. There are therefore non-physical economic, social and political processes in place which are physically manifested in housing, schools, parks and other services and facilities.

The scales at which urban form can be considered or measured include the individual building, street, urban block, neighbourhood and city. These levels of spatial disaggregation influence how urban form is measured, analyzed and ultimately understood. The issue of scale is discussed throughout this chapter (and the book) as it constitutes an underlying dimension of any examination of urban form.

Urban form generally encompasses a number of physical features and non-physical characteristics including size, shape, scale, density, land uses, building types, urban block layout and distribution of green space. These are categorised here as five broad and inter-related elements that make up urban form in a given city (Fig. 2.1).

Fig. 2.1

Elements of urban form

These elements of urban form have been identified on the basis that they are claimed to influence sustainability and human behaviour. They are considered in more detail below. These elements relate to developed, and not developing, countries. For this reason, infrastructure (e.g. water, roads, gas etc.) is not discussed here as an element of urban form; however, it is acknowledged that infrastructure would form an important part of examinations of urban form in developing countries.

Density

Density is a deceptively complex concept with a number of inter-related dimensions. While it may provide an objective, spatially-based, measure of the number of people (living) in a given area, it is also assessed subjectively; it is a social interpretation dependent on individual characteristics and so may differ from resident to resident (Churchman, 1999). For example, while the density of Trafalgar Square in London may be reported as low (density usually being a measure of residential occupancy), the perceived density, and extent of crowding, may be very high (after Rapoport, 1975).

Density entered the consciousness of UK policy makers in the nineteenth century when urban areas were growing rapidly and overcrowding and appalling living conditions were prevalent among the poor (Jenks and Dempsey, 2005). There is also a more cultural dimension to density, where the densities at which people live may be considered as relative. Current English housing policy states that new residential building should be at a minimum of 30 dwellings/ha which for some may be an unacceptably high density (DCLG, 2006). In Hong Kong however, a minimum of ten times that density would be considered low (after Jenks, 2000; Breheny, 1997; Jenks and Dempsey, 2005).

Density is also closely associated with other elements of urban form, such as land use and access to services – for example, for a service or facility to be viable, it needs to serve a population of a particular size. Density on the one hand can be seen as an outcome of the competition between land uses within a given urban transport infrastructure and its associated pattern of accessibility. On the other hand it is a policy goal as it is also an input into the quality of urban life through the viability of services provision and availability of public and private space. Density has therefore been used as a tool to measure the viability of public transport infrastructure and other service provision, the feasibility of certain land uses, particularly commercial and service, in urban design and construction. At what point density becomes high (or too high) is unclear, but in recent years planning policy and practices in many countries have been attempting to increase the average density of new development.

Land Use

Broadly speaking, the term land use is used to describe the different functions of the environment. Within the urban context, the dominant land use tends to be residential but a functional urban area requires industrial, retail, offices, infrastructure and other uses. The spatial (micro) pattern of land uses is crucial to the arguments about the efficiency of a city and potential ‘sustainable’ urban forms in influencing urban travel patterns and the quality of life, for example through the existence of green space. There are also certain ‘locally-unwanted land uses’ such as prisons (Grant, 2002), airports, or landfill sites claimed to be undesirable in residential mixed-use areas (Healey, 1997). Planners have traditionally attempted to separate land uses because of potential undesirable externalities but are now in favour of mixed use developments. For example current UK policy promotes easily accessible services and facilities for residents (DCLG, 2006); both ‘horizontally’ – at ground floor level – and, increasingly in new city centre developments, ‘vertically’ – within the same building (DETR, 2000). However, land use patterns are dynamic rather than static phenomena and are subject to real estate market forces.

A key component of local land use is the availability of local neighbourhood services. The provision of services and facilities is dependent on the resident population’s requirements so a particular land use mix therefore differs from neighbourhood to neighbourhood (Urban Task Force, 1999). The local urban context and the requirements of the population are therefore important in this matter. It is not however clear which services and facilities can and should be provided at which spatial scale. An ‘everyday eight’ local neighbourhood services and facilities identified by Winter and Farthing in the UK context includes post office, supermarket, primary school, newsagent and open space (1997, p. 127). Other services to which residents need local access, albeit on a less frequent basis, include a doctor’s surgery (Barton et al., 2003; Urban Task Force, 1999), chemist; bank (Burton, 1997); and community centre (Aldous, 1992). There is extensive prescriptive UK guidance on what those land uses should be for a given neighbourhood however there is no consensus (Dempsey, 2008).

Accessibility and Transport Infrastructure

Transport infrastructure is closely associated with accessibility as it determines the ease with which buildings, spaces and places can be reached. The level of accessibility describes the area residents and users are able to reach, as well as the extent to which they have the means to access places, services and facilities that are outside their local area (after Talen, 2003).

Accessibility is actually a layered concept and is not simply proximity as distance is just one contributor. It is dependent on a number of factors including the location of potential destinations relative to an individual’s starting point, how well the transport system connects to spatially distributed locations, how the individual uses the transport system, and the characteristics of, for example, the services and facilities that the individual plans to use (Liu and Zhu, 2004). A key accessibility relationship is between home and the city centre. Different aspects of the concept encompass access in terms of what is available within walking distance of home (sometimes referred to as ‘pedshed’), or access in terms of the means to get to, for example, services and facilities which are located further afield (Barton et al., 2003; Schoon, 2001). It is therefore closely linked to land use and layout: the services, facilities, open space, how they are arranged within a city or neighbourhood and the means of getting to them all contribute to how accessible a place or service might be described.

Urban Layout

Layout describes the spatial arrangement and configuration of elements of streets, blocks and buildings, often referred to at the street scale, such as grid or tree-like (cul-de-sac) street patterns. Layout has an important influence on pedestrian movement and the way in which different places and spaces are connected to each other (ODPM, 2005; CABE and DETR, 2000). The layout, whether or not it is ‘permeable’ and easy to find the way, controls access and movement for pedestrians, and could influence other aspects of urban form such as land use or density (Hillier and Hanson, 1984; Hillier, 1996).

The layouts of today’s cities are largely artifacts of their historical development and planning and building regulations. The configuration of the street network, in terms of its urban block sizes, their overall location within the city, pedestrian and vehicular connectivity, can affect the functioning of a city by, for example, influencing the location intensity of activities (Penn et al., 1998; Porta et al., 2008).

The connectedness and permeability of urban layouts are claimed to determine the nature and extent of routes between and through spaces which in turn has an influence on how lively and well-used a space is (Cowan, 1997). Streets which are well-connected to services, facilities etc. and/or the means for the pedestrian of reaching them, are argued to be more frequently used than deserted or quiet options (Gehl, 2001; Gehl et al., 2004).

Housing and Building Characteristics

The characteristics of housing and other buildings in urban settlements can have an important bearing on everyday living: it has already been noted that residents living in low-density detached dwellings with large gardens will have a distinct experience of the urban environment from high-rise city centre apartment dwellers. However, the influence of building characteristics extends beyond the density of urban living. Factors such as building type, height and age may have an effect on a number of issues. These might include a building’s orientation and exposure to sunlight and daylight (Mardaljevic, 2005) and the potential for modifications, such as changes to living space to work space or individual room conversion to continue accommodating an ageing resident as in the ‘lifetime homes’ model (Holmes, 2007).

Other factors such as the amount of living space in dwellings, number and types of particular rooms and lowest level of living space may also have significant influences on the efficiency of buildings in terms of its embodied, operating and life cycle energy (Newton et al., 2000).

Integrated Elements

While it is useful to examine these elements separately, it is also clear that they are inter-connected and interdependent. For example, the accessibility within an urban settlement is very closely linked to its density and the layout of, and extent of mixed uses within. A neighbourhood is not accessible without viable services and facilities available for residents to use. Neither is it accessible without pedestrian, cycling and public transport networks through which the neighbourhood is connected both to its own services and to services outside. When planning and constructing new residential areas, housing type and size may be dictated by the proposed density of a site, which for example, in line with recent UK policy, will also provide a range of land uses (including different services and facilities and open space) and a connected and permeable urban layout. This message is endorsed by a series of US studies that have sought to develop measures of urban sprawl (Ewing et al., 2002; Cutsinger et al., 2004).

The interrelated associations between these elements have wide-reaching implications for the research. Firstly, there is a need for compatibility in how the different elements of urban form are measured. This is to ensure that, secondly, the statistical (and other) analyses conducted can account for the individual effect that each element may have on a particular aspect of sustainability, as well as the collective influence of the elements of urban form. In this way, the main research question which looks to explain how urban form affects sustainability can be answered in as robust and reliable a manner possible. The next section examines outlines the methodological approach used in the research to measure these elements of urban form.

Measuring Urban Form

A largely quantitative approach was adopted in this research, although qualitative methods were used in parts of the project and are detailed in the relevant chapters. To measure robustly the features of urban form outlined above, a two-pronged process of data collection was followed. Firstly, existing datasets such as the 2001 Census, the Valuation Roll and Ordnance Survey data were examined to provide information on, for example, initial density measures, non-domestic properties and their location. Secondly, after ascertaining gaps in the data, e.g. building heights and information on transport infrastructure (e.g. bus stop location and type of car parking), a site survey was conducted in the study neighbourhoods on a street-by-street basis.

The site surveys were undertaken by researchers using an innovative method employing PDAs (personal digital assistants) with GPS (global positioning system) modules which allowed geocoded survey data to be directly downloaded into a GIS (geographic information system) platform. A useful innovation was the production of prototype software to automatically link separately sourced data tables from Ordnance Survey and Valuation Data. This allowed site surveyors easily to identify, locate and check business and mixed use premises. The data collected in this way included details about: buildings such as condition and height; land uses; the presence of litter and graffiti; position of bus stops and shelters.

The research design was a cross-sectional one, where data are collected at one point in time providing a ‘snapshot’ approach (Gray, 2004). While the project did not detail the extent to which a changing urban form can contribute to sustainability over time, it offers new and valuable data on how and to what extent current urban forms in UK cities can be described as sustainable. The following sections outline how each element of urban form was measured in the research.

Measuring Density

A wide range of different measurements have been used to calculate the density of a given area, such as persons per hectare (pph), dwellings per hectare (dph), bed spaces per hectare and habitable rooms per hectare (Woodford et al., 1976). Employing a number of density measures has been argued to be more robust than using one single density indicator which cannot accurately measure the density of a given area (Jenks and Dempsey, 2005).

A range of density indicators was selected to provide as complete a picture as possible of the overall density of the case studies, while accounting for the different scales of urban form (the city, neighbourhood, ‘sub-area’ and street). Examples of these indicators are presented in Table 2.1. It should be noted that the indicators of net residential density used here are based on a definition of residential which includes outdoor space such as gardens, but excludes streets and footpaths. This method of calculation results in density figures which are higher than those usually reported.

Table 2.1 Indicators of density

It is clear from the table that the indicators measure physical density and not perceived density. Aspects of perceived density are measured in Chapters 9.

Measuring Land Use

In order to measure the extent of mixed land uses in the case study neighbourhoods, a number of appropriate land uses were selected, excluding land uses which are not relevant for the purposes of the research, e.g. telecommunications, energy and waste infrastructure. Data on a number of particular services and facilities are not always specified in secondary data sources (e.g. name of supermarket, newsagent and children’s nursery), indicating a need to conduct primary data collection.

It is necessary to account for any ‘edge effects’, where residents may be using local services and facilities outside the case study boundaries identified for the research purposes. A ‘buffer zone’ of approximately 400m (approx. 5 min walking distance) is applied around each case study area to account for any ‘edge effect’. Households living on the edge of a chosen case study area might be closer to facilities just outside the boundary and therefore may choose to use those rather than the ones initially identified in the research. By applying a buffer zone around the neighbourhood, the researchers are able to capture and investigate usage of particular services and facilities including food shops, post office and GP surgeries. This land use information is then mapped using a GIS-based platform (Fig. 2.2). Examples of these land use indicators used in the research are presented in Table 2.2. The categories used are based on the National Land Use Database (NLUD) developed by the then ODPM (2003).

Fig. 2.2

Example of GIS-generated land use map (Oxford) (© Ordnance Survey)

Table 2.2 Indicators of land use

Measuring Accessibility and Transport Infrastructure

As indicated earlier, measures of accessibility can refer to different aspects of the concept. The indicators used to measure accessibility here cover transport infrastructure for pedestrians and cyclists as well as public and private transport. Table 2.3 shows some of the indicators used to measure accessibility which include characteristics of public transport infrastructure and journey times and distances. These indicators (and others such as socio-economic characteristics of residents and employment location) were included in a transport model (discussed in detail in Chapter 3) to provide as accurate a picture of accessibility in the case study neighbourhoods.

Table 2.3 Indicators of accessibility

Measuring Housing/Building Characteristics

It is not possible or desirable to measure the characteristics of every building in the case study neighbourhoods. An efficient method of measuring the characteristics involves the identification of predominant housing types per street, and highlighting where there were exceptions to this. It is also useful to make use of the household questionnaire in measuring these characteristics. While the focus of the questionnaire is to measure aspects of sustainability, it proved to be a useful tool to collect urban form indicators, such as housing type, lowest level of living accommodation and a household’s access to a garden/residential outdoor space.

The characteristics of non-domestic buildings are collected which understandably overlap with the indicators measuring land use. Indicators of maintenance are included in this category of indicators, such as the condition of buildings (where considered to be poor relative to other buildings in the street) and levels of litter and instances of graffiti and vandalism (Table 2.4).

Table 2.4 Indicators of housing/building characteristics

Measuring Layout

Urban layouts are difficult to quantify. Spatial network analysis is one of the ways in which spatial layouts can be objectively described and quantified to identify similarities or differences. Typically in a spatial network analysis, relationships between spaces in a city/settlement/building are represented as relational graphs similar to social network graphs. These graphs can then be analyzed to identify patterns and to quantify the relationships between spaces.

For this research, Multiple Centrality Assessment (MCA) is employed to measure layout (see Porta et al., 2006). MCA operates on a standard (or primal) graph representation based on street network systems which accounts for metric distances while analyzing the relationship between spaces. The final output provides a set of simple and compound measures of centrality numerically expressing the relative importance of a space in relationship to other spaces in the city. The output also includes a network map that shows the location and clustering patterns of spatial centrality (Fig. 2.3). A more detailed description of MCA and a table of the indicators employed to measure urban layout can be found in Appendix 1.

Fig. 2.3

Example of an MCA-generated map of Leicester

Measuring Overview

The elements of urban form link the constituent and distinct parts of this research together. This is the first time empirical research has examined the effect that urban form has on sustainability in a holistic manner. Indicators employed to measure urban form include both simple and complex measures collected or derived from secondary sources, primary data collection or detailed computer modelling. The range of indicators described above allows us to determine the relative influence that differing elements of urban form – land use, density, accessibility, housing/building characteristics and layout – have on economic, environmental and social sustainability.

The relationships that the elements of urban form, both individually and as a whole, have on the different aspects of sustainability are analyzed and outlined later in the book. The next section describes the case studies in detail and outlines the features of urban form in each area.

Case Study Areas: Profiles

The research in this book is based on empirical analysis in five British cities and a small number of neighbourhoods within each of these cities chosen for more in depth study. This part of the chapter describes these places in some detail, providing information about their urban form, housing, socio-demographic characteristics, history and economic profile. The case study cities are described first, followed by the profiles of the case study neighbourhoods.

Case Study Cities

Five provincial British cities provide the focus for the CityForm research: Leicester, Oxford and Sheffield in England and Glasgow and Edinburgh in Scotland (Fig. 2.4). The cities are all university towns and cover a variety of geographical and economic situations. It is important to note, that while the cities are varied in their economic, demographic and physical make-up, they cannot be claimed to represent UK cities as a whole. As a result, care must be taken when interpreting the results of this research.

Fig. 2.4

Location of the case study cities

General Characteristics of the Five Cities

Glasgow and Sheffield are the largest of the cities, with populations of more than 500,000. They are both traditional industrial cities that grew dramatically in the nineteenth century and now have important commercial roles. Edinburgh is the next largest with a total population of approximately 450,000 and is one of the few large provincial cities in the UK whose population is growing. It is the capital of Scotland and an important administrative and financial services centre. Both Leicester and Oxford are significantly smaller; Leicester with a population of 280,000 and Oxford with a population of 134,000. Leicester has quite a diverse economic structure and a thriving ethnic minority community that accounts for more than a third of its population. Oxford is most well-known as a university city, but is also a thriving tourist and business centre.

Demographically, the five cities tend to have a younger than average population with a high proportion of people in their twenties. This may be linked to their status as university towns, with the effects most noticeable in Oxford where one quarter of local people are aged between 20 and 29 and the average age of Oxford residents is 35 years 4 months. The most affluent of the five cities are Edinburgh and Oxford and they also have the highest proportion of young adults. Glasgow and Sheffield residents tend to be slightly older (38 and 39 years respectively), while Leicester has the youngest population profile with an average age of 35 years – more than 3 years younger than the UK average.

All of the case study cities – with the exception of Edinburgh – have a lower proportion of owner occupiers than the UK average (70%). Sheffield has the lowest proportion of private renters (10%), while in Oxford almost a quarter of households are private renters. Social renting on the other hand is most prevalent in Glasgow, where 40% of households have a social landlord, while just over one in six households in Edinburgh are in this situation.

Driving to work by car or van is the most popular mode of travel in all five cities with more than half of commuters using a car in Leicester and Sheffield. Glasgow is the only one to have an underground system and 4% use it while Sheffield has a tram network which attracts 3% of commuters. More than a quarter use the bus in Edinburgh but the city with the highest proportion of commuters using (all types of) public transport is Glasgow. Another notable difference is the high percentage, 15%, of people who travel to work on bicycle in Oxford, which is much higher than the other four cities (3% in Edinburgh).

Some Physical Characteristics of the Five Cities

Density provides a standard international indicator of urban form and enables both a simple comparison of individual cities and a worldwide perspective on the five case study cities. The figures presented in Table 2.5 reveal a significant variation in population density and dwellings per hectare. Sheffield is the least densely populated city of the five, both in persons per hectare and dwellings per hectare. Leicester has the greatest number of persons per hectare and Glasgow has the greatest number of dwellings per hectare. Edinburgh has a comparably low density, similar to that of Sheffield. These statistics also show that there is no clear divide between England and Scotland despite the distinctive tenemented housing stock prevalent in both Scottish cities. Edinburgh and Glasgow both have much higher numbers of flats than the English cities with around 60% and 70% of households living in flats in these two cities. However, there are differences between the English and Scottish cities studied in the number of rooms per dwelling. The English cities all had an average of more than 5 compared with 4 in Glasgow and 4.6 in Edinburgh.

Table 2.5 Density and housing types: case study cities

As Table 2.5 shows the densities of these cities range from 29 to 38 persons per hectare which places them at the lower end of the international spectrum of urban densities but much higher than most North American cities. The densities are at the low end of the range of densities of European cities and substantially below the densities of Asian cities that are typically over 200 people per hectare (Bertaud, 2003).

Case Study Neighbourhoods

The starting point for the empirical research is the neighbourhood and fifteen case studies were chosen to represent inner, between and outer neighbourhoods within the five cities. The case study neighbourhoods are chosen to provide a slice through each of the cities and to represent a wide range of neighbourhoods. The overview of these neighbourhood characteristics also provides a useful insight into the spatial socio/demographic and housing stock structure of (British) cities.

Each case study neighbourhood includes at least 2000 households, a mixture of land uses, a range of housing types and street patterns, nearby public transport and households with a range of socio-economic backgrounds. These neighbourhoods are used as the principal spatial unit of study and where appropriate are also split into sub-areas. Sub-areas are defined using maps and local knowledge, to identify natural physical sub-divisions respecting obvious major boundary features and to reflect relative homogeneity of urban form within their boundaries. Each case study neighbourhood is divided into 6–7 sub-neighbourhoods, giving a set of 97 in total (a few of which contain only a small number of households, because they are dominated by non-residential land uses).

Characteristics of Neighbourhoods

A range of information was collected about the case study neighbourhoods. This includes a site survey capturing information about the built environment and a questionnaire survey providing information about household characteristics, behaviour, feelings about the neighbourhood and use of local facilities (e.g. parks, shops, public transport). An overview of the 15 neighbourhoods is profiled at the end of this chapter (Table 2.6), and more details on individual localities are included in Appendix 2.

Table 2.6 Profile of the case study neighbourhoods

At one level these areas represent a set of diverse neighbourhoods but a number of regularities can also be discerned. Gross population densities follow a negative gradient spatial structure from inner to outer neighbourhood. There are also consistent spatial demographic patterns: younger people with few children living in inner areas and older households and families predominating in outer areas. Between areas are more diverse. Private rented housing is focused in the inner areas and outer areas are almost exclusively owner occupied. Social housing in British cities is spatially concentrated, and this is reflected in our case study areas, being located mainly in a few inner areas. These findings show the spatial structure of the five cities, as given by the characteristics of the case study areas, conform to an urban system diffusing from a central core, and that the different physical urban form elements integrate together with socio-economic characteristics.

Layouts of Neighbourhoods

The spatial analysis is carried out at the city level, neighbourhood, sub-area and street level. Most of the comparative analysis to determine the performance of various types of urban form is at the neighbourhood and sub-area level. The spatial characteristics of the neighbourhoods reveal a range of layout types ranging from the predominantly gridded to those that are largely culs-de-sac. The characteristics are quantified using MCA analysis as it permits the calculation of many indices that measure street networks and allow systematic comparison of the case study neighbourhoods (see Appendix 1 for details). There are measurable differences between the neighbourhoods and none can be said to represent one layout type or another, as all have a complex mixture of layout forms (Fig. 2.5).

Fig. 2.5

Shows the spectrum of complexity index scores across the fifteen neighbourhoods. Lower values indicate higher choice-grid like pattern that are predominantly in inner neighbourhoods and higher values indicate low choice-tree like pattern in the suburbs. (light blue = suburban; medium blue = between; dark blue = inner city)

The MCA analysis also identified the density of street intersections, the relative degree of complexity of the street networks, their interconnectedness, and how efficient the networks are – related to actual distances between intersections. In addition, the degree of compactness or sprawl is measured by considering the ‘fractal’ dimension of street patterns in the case study neighbourhoods. The fractal dimension index ranges from 1 to 2 – a score of 1 would be a perfect linear system where intersections lie on a straight line, while a score of 2 would be a system where intersections are distributed evenly throughout the space covering the whole neighbourhood. The range in the neighbourhoods from a more compact form in Sheffield inner area to the sprawling layout of Glasgow outer has a relatively narrow spread of values is again due to the fact that many case study areas have both compact and spread areas within them (Fig. 2.6).

Fig. 2.6

Examples of compact and sprawling neighbourhoods

Summary and Conclusions

At one level the elements of urban form are relatively simple – land use, density, accessibility defined by transport infrastructure, characteristics of the built environment and layout. Although there is an expanding literature that seeks to quantify urban form using complicated measures and advanced statistical techniques the findings of these studies have emphasised the overlapping nature of the elements and the results have not justified the complexity. The study has chosen in general to apply as simple measures as possible as the best way to elucidate the issues on sustainability. The analysis reported in this book is based on five cities in the UK. The urban form characteristics of these cities in terms of population density are very similar within the spectrum of world cities. These densities are at the low end of the range of densities of European cities and substantially below the densities of Asian cities.

The fifteen case study neighbourhoods selected for deeper study and located at inner, middle and outer points to represent slices through each of the five cities. There are a number of pointers that arise from the various measures of the elements of urban form of these neighbourhoods:

• The inner neighbourhoods tend to be well connected and complex, with predominately grid-like structures;

• The inner neighbourhoods tend to have more compact layouts in comparison with suburban neighbourhoods;

• The suburban neighbourhoods have tree-like structures with a single or limited number of main roads acting as a spine or trunk with culs-de-sac;

• The gross population densities also follow a similar spatial structure; a broadly consistent negative gradient from the city centre;

• There are consistent spatial demographic patterns: younger people with few children living in inner neighbourhoods and older households and families predominating in outer neighbourhoods. Between neighbourhoods are more diverse.

• Private rented housing is focused in the inner neighbourhoods and outer neighbourhoods are almost exclusively owner occupied.

• Social housing in British cities is spatially concentrated, and this is reflected in our case study areas, being located mainly in a few inner neighbourhoods.

These findings show the spatial structure of the five cities, as given by the characteristics of the case study neighbourhoods, conform to an urban area diffusing from a central core, and that the different physical urban form elements integrate together with socio-economic characteristics. The following chapters investigate the sustainability of different dimensions of this urban system drawing on more detailed study of these case study neighbourhoods.