X08. London prototype: Dynamics data capture and urban morphology - colouring-cities/manual GitHub Wiki

Author: Polly Hudson, 2020.

1.1 Introductory remarks

In this third section on data types the value of data describing dynamic behaviour in stocks not just at building, but also at plot and sub-building level, is explored, as is the importance of involving urban morphologists in this process.

1.2 Urban morphology and the study of urban change

1.2.1 Overview

Urban morphology developed as a scientific discipline in the late 19th century, in Central Europe, out of geography, to distinguish, characterise and explain urban landscapes (Whitehand, 2007). Karl Kropf, in his Handbook of Urban Morphology – a much needed guide to this complex and multifaceted subject – describes urban morphology as involving ‘the study of human settlements, their structure and the process of their formation and transformation’ (Kropf, 2017, p. 9), where tools are developed to answer the question, ‘what is really going on in the built environment?’ (ibid., p.5). As in the work of Jacobs and Alexander, interest in the impact of the built environment on people is central to urban morphology research.

Anne Moudon describes the three basic components studied in urban morphology as being ‘form’, ‘scale’ and ‘time’; time/historical context are seen as critical to understanding the process of formation and transformation of urban form (Moudon, 1997, p.7). Cycles and hierarchies, and relationships and interdependencies between urban components, are core areas of study, as are inheritance, persistence, and capacity for adaptation, and the form, evolution and mutation of building typologies over time. These are studied through the processes of analysis, comparison and synthesis (Kropf, 2017). Spatial information on building age, and on changes occurring to building form at plot level, are also of specific interest, ‘The positioning of a building in an urban nucleus not only depends on the building date of the manmade construction that still exists, but largely derives most of its characteristics from the first building date from which it inherits its position and the size of the built lot’ (Whitehand, 2003).

Classification of urban form and fabric are undertaken principally at building, plot, street block and street level to provide ‘a consistent and rigorous descriptive language’ (Kropf, 2017, p.10). Together these components form what is known as the ‘urban tissue’, which itself creates ‘a pattern of patterns that lies at the heart of the built environment as a form of organised complexity’ (ibid., p.15). Here, knowledge is combined of ‘process’; ‘configuration’ (the way in which parts are arranged); ‘hierarchy’, and ‘modularity’ (where the same pattern or configuration of elements is consciously reproduced) to create ‘types’ (ibid, p.14).

1.2.2 Patterns, hierarchies and cycles of change

Within urban morphology, much time is spent studying patterns and hierarchies of change. This work is now overlapping with findings from sustainability science, within mortality studies (as described in Chapters 2), and urban science, within the study of urban complexity and microdynamics (as discussed in Chapter 3), and is also relevant to the development of 3D ‘procedural’ city models as discussed further below.

Examples of overlaps can be seen, for example, with regard to ‘rules’ identified in relation to relative rates of change of urban components. In urban morphology, Brenda Case Scheer describes how site topography and hydrology will change more slowly than streets and land parcels, followed in turn by infill streets, plots and buildings and finally vegetation (ibid., 2017). She also notes that streets are ‘notoriously difficult to change, because they are the boundaries of property and they constitute access to property that is essential for development’ (Scheer, 2010, p.56). M.R.G Conzen observes how land uses change faster than plots, with plot boundaries often surviving for hundreds of years, and street networks able to persist for over a thousand years (Conzen, 1960). Jeremy Whitehand also describes how very old street systems ‘constitute a framework that powerfully influences the long-term historical development of the city’s conformation. Land and building utilization, in contrast, tends to be much more ephemeral. Buildings are, on average, intermediate in their resistance to change’ (Whitehand, 2007, ii–06).

In urban science, Batty and Stanilov, have also illustrated the long lasting impact of the preurban street structure on urban growth (and different rates of change between land uses) (Batty and Stanilov, 2011), while in sustainability science, Kohler and Yang also record that infrastructure, including street networks, changes more slowly than the stock (Kohler and Yang, 2007). Dijst et al. comment on ‘the very slow processes of changing physical transport, communication and utility infrastructures and distribution of land uses’ (Dijst et al., 2018, p. 190). In Reyna and Chester’s study of urban metabolism in Los Angeles, the hypothesis was put forward that different demolition rates, and thus different lifespans, could occur for the same land use (Reyna and Chester, 2015). This is confirmed within urban morphology by Whitehand’s finding that detached houses have a faster rate of change than semi-detached (owing to the greater opportunity for developers to profit from densification) (Whitehand, 2001).

Findings in urban morphology, with regard to relative rates of change between the domestic and non-domestic stock are also now being supplemented with those from sustainability science. In urban morphology, Whitehand’s research into building cycles, identified that non-domestic building booms coincide with domestic building slumps, and that non-domestic buildings are demolished at a faster rate than domestic (Whitehand, 1994). In sustainability science, higher rates of non-domestic demolition have been recorded in Japan (Tanikawa and Hashimoto, 2014), Switzerland (Aksözen et al., 2016; 2017), Germany (Bradley and Kohler, 2007), Finland (Huukha and Lahdensivu, 2014) and China (Liu et al. 2014). This is explained by Aksözen et al. as the result of a ‘faster pace of change of use in non-domestic building and by the generally higher flexibility of domestic buildings’ (Aksözen e al., 2017, p. 269). Kohler and Hassler observe that domestic buildings generally survive three time longer than non-domestic (Hassler, 2009). In Finland, non-domestic demolitions between 2000 and 2010 were four times higher than domestic, with commercial demolitions twice as high as for industrial and public buildings (Huuhka and Lahdensivu, 2014). In Germany, non-domestic stock in Ettlingen was found to be demolished at ten times the domestic rate (Bradley and Kohler, 2007). Relative rates of change were also identified as relating to location, with, Finnish and Chinese studies finding, for example, buildings to have shorter lifespans where located in central business districts (Huuhka and Lahdensivu, 2014; Liu, 2014). In terms of age, Swiss and Finnish studies found (in the case of mature cities) that newer buildings were demolished at a faster rate than old (Huuhka and Lahdensivu, 2014; Aksözen et al., 2017), with some age cohorts, such as early post-war buildings (as discussed in Chapters 1 and 7) identified as having a higher probability of demolition than others.

Findings however also differ, for example on whether smaller buildings might be more likely to be demolished than large, with this appearing to depend on building type (Porta and Romice, 2010; 2015; Hassler, 2009; Liu, 2014; Aksözen et al., 2017). Aksözen et al., also raise the interesting question of whether combinations of building attributes might increase or decrease the probability of building survival (Aksözen, 2017).

1.2.3 The importance of data on plots

In urban morphology, detailed information on plots and parcels, and streets, not just buildings, is considered essential to describe urban form and its dynamic behaviour. Moudon, like Batty and Xie, talks of the ‘cells’ of a city, the smallest being the ‘individual parcel of land, together with its building or buildings and open spaces. The characteristics of the cell define the urban form’s shape and density, as well as its actual and potential use over time’ (Moudon, 1997, p.7). Kropf highlights that no clear consensus exists for the definition of plots (or lots), but that a plot is commonly considered a small piece or area of land; described by M.R.G Conzen as a parcel of land representing a land-use unit defined by boundaries on the ground, and by units of property holding (Kropf, 2018). Jeremy Whitehand sees the plot as the basic element in land division patterns (Whitehand, 2001) and Scheer as the ‘game board’ of all action: ‘Streets, rights of way and property lines will endure for centuries; even a cataclysmic event… will not alter the underlying and invisible web of property boundaries’ (Scheer, 2010, p. 47). Peter Bibby in his work on land use notes that once the land has been enclosed, parcel boundaries ‘tend to endure even when the land cover and activity changes’ (Bibby, 2009, p. 53). Frank Brown also found land parcels and property boundaries to be ‘an essential element of continuity in the landscape, tying together the building patterns of successive periods’; acting as constraints to development, and able to reveal the underlying logic of urban systems (Brown, 1985, p.82). Whitehand observes how historical land-use patterns, embedded in the tissue of streets and plots, can also serve as a constraint on more drastic change (Whitehand, 2001). Data on plot and parcels, and land ownership are therefore extremely important in understanding constraints on, and forecasting change to, building stocks. They will therefore need to be made accessible within open building attribute data platforms, alongside data on buildings.

1.2.4 The importance of data on adaptability and typology evolution

Within urban morphology, detailed observations of change are made at the microlevel. Kropf differentiates between the two interrelated dynamic processes of ‘development’ and ‘evolution’ (Kropf, 2017, p. 37). ‘Development’ is described as relating to the life history of a building and the generic processes of foundation, incremental extension within its plot, internal transformation, and abandonment or destruction (ibid., p. 37). ‘Evolution’, on the other hand, is described as relating to the generation and mutation of building types, and the ‘reproduction, development and modification of many individual examples of the type over time’ (ibid., p. 36). Types of data required to analyse each of these processes, and methods of their capture, for incorporation in open data platforms, form the focus of the remaining section of this chapter. Data relating to the process of ‘development’ within plots is first described.

1.3 Tracking incremental development within plots

1.3.1 M.R.G. Conzen, historico-geographic methods

In 1960, a year before Jacobs’ Death and Life of Great American Cities was released, M.R.G. Conzen, published his study of the English town of Alnwick (Conzen, 1960). Many of Conzen’s interests overlapped with those of Jacobs. Indeed his method of tracking physical change within plots provides a formal description and illustration of the process of age diversity creation, of which Jacobs writes so eloquently. Conzen and Jacobs were both interested in the repetition and persistence of identifiable urban phenomena, and both considered that the logic of urban areas could not be understood simply through a set of aggregated statistics; instead study was required at a range of scales, including at building level. Conzen commented on the ‘crudeness’ with which the evolution of towns was studied in Britain in the mid-20th century, and how geographers ‘took account merely of the broad stages of outward growth and missed the variety of the phenomenon they cover, as well as the significant modern changes inside the street blocks…’ (ibid., 1960, p. 4). He noted the way in which focus was placed on streets and street spaces rather than on the structure of blocks and the form and uses of buildings themselves (ibid.).

Conzen pioneered methods of visualising and recording change to the urban fabric, in 2D, at plot level. Some of the first examples were generated by early proponents of a rigorous, scientific approach to the study of the built environment (ibid.), including Walter Geisler, who in 1918 produced colour-coded maps of land use and building storeys for Danzig (Gdansk), and the Austrian geographer Hugo Hassinger, whose 1916 map of Vienna, shown in Figure 1.9, was one of the first to use colour to show the relationship of building age to physical form (ibid.). Their work was to form the foundation of the scientific discipline of urban morphology (discussed further in the context of dynamics data in Chapter 4), and developed in Britain by M.R.G Conzen. In his early career in Germany in the 1930s, Conzen mapped building types for twelve towns (Whitehand, 2007). Here the power of colour in data visualisation was experimented with; with, number of storeys shown mapped by colour depth (ibid.).

As well as advancing work into the classification of buildings, and the use of colour, Conzen, in his study of Alnwick, demonstrated through the application of micro-visualisation techniques, how age and form are mixed, over time, to create areas that are ‘strictly unique’ (Conzen, 1960, p. 6). He revealed – through historical research and meticulous comparison of historical maps – hierarchies and cycles of change, relationships between buildings, plots, blocks and streets, and an underlying logic to the urban fabric. He sought, through the study of processes at different geographic scales and across long periods of time, to generate ‘a theoretical base yielding concepts of general application’ (ibid., p. 3); ‘By investigating a specific case which promises results of general significance and by adopting an evolutionary viewpoint, [the method] seeks to establish some basic concepts applicable to recurrent phenomena in urban morphology’ (ibid., p. 4).

Historico-geographical methods developed by Conzen, and advanced within the British School of Urban Morphology, have been seen as crucial by Jeremy Whitehand, in helping planners understand not only the composition and history of the urban fabric, but also constraints on its operation. In his 2007 paper, Whitehand stated that ‘awareness just of the existence of historical features is not enough. How they fit together is critical’ (Whitehand, 2007, ii–04). He also argued that greater integration of science and technology was needed in the study of the built environment, as well as the accommodation of diverse methods of analysis (ibid.). In 1974, Whitehand set up the Urban Morphology Research Group at the University of Birmingham, as the main centre for urban morphology research in Britain in the historico-geographical tradition (Pinho and Oliveira, 2009). In 1994, Whitehand and Peter Larkham founded the journal of the International Seminar of Urban Form (ISUF), Urban Morphology, to widen research and practice in fields concerned with built form (International Seminar on Urban Forum, 2020). Originally comprising around twenty geographers, historians, architects and planners, ISUF was recorded in 2020 as having over six hundred individual and institutional members representing more than fifty countries (ibid.). It therefore curates a unique repository of knowledge on the structure and dynamic behaviour of the built fabric. This knowledge base is set to become increasingly important as interest in stock dynamics grows, and as advances in AI and machine learning, and microsimulation, are made. Ensuring that open building attribute data platforms are useful to, and designed in consultation and collaboration with ISUF members, is therefore important.

1.3.2 Conzen and the burgage cycle

In his Alnwick study, Conzen visually recorded the step-by-step process by which building plots, in urban centres, are densified over long periods of time. His research led to his theory of the ‘burgage cycle’, a variation of this general phenomenon in which plots are filled in as a result of development pressure (Whitehand, 2001, p.105). Here, by mapping incremental additions, Conzen showed how, plots in city centres, were gradually built over (where incremental development was possible) until such time as the plot was completely or almost completely filled. At this point, constraints on further adaptation imposed by property boundaries resulted in complete demolition, with the development cycle then beginning again (ibid.). In 2002, M.R.G Conzen’s burgage cycle theory was tested by M. P. Conzen in the US. In Figure 5.1, a 200-year plot cycle from the Alnwick study, is compared with a 200-year plot cycle for Cincinnati. Patterns of subdivision, amalgamation and finally clearance can be seen to be remarkably similar in both, despite their geographic distance, providing evidence for the cycle being a general phenomenon of urban densification (Conzen, 2002). For historical and current incremental development and plot densification to be analysed, data on the geometry and dimensions of plots and the ratio of building to plot, need to be accessible. Spatial data on incremental change for multiple time points are also required.

1.3.3 Peter Larkham and 2D graphic representation of change

In the 1990s Peter Larkham, published a very different visual method of quantifying change at plot level (Larkham,1996). Here, historical planning applications were assessed for a number of British high streets, spanning more than twenty-five years. From these, data were manually extracted on relative amounts of demolition, refurbishment, facade alteration, internal alteration, and new build, and on whether schemes had been implemented or not. These were then recorded on charts as illustrated in Figure 5.2. Larkham’s method is particularly interesting in terms of the type of data captured and the scale at which it is collected. His identification of the importance of information on whether or not planning schemes have been implemented also raises the question of whether live planning data could potentially be streamed into open data platforms, and colour-coded depending on the stage of the planning process reached. (This is further discussed in Chapter 5). The way in which data are illustrated is also of relevance to platform design in the context of the possible future inclusion of data analytics dashboards. Figure 5.2 shows how similar graphics could potentially provide users with live insights into the metabolism of a city and into relative rates and quantities of adaptation, demolition and new build occurring.

1.3.4 Micromorphological methods of quantifying change in 3D

Torma et al. In Torma et al.’s longitudinal research into suburban high street development in London, construction, demolition, plot merging and plot generation are recorded in 3D for five time points between 1880 and 2013, as shown in Figure 5.3 (Torma et al. 2016). The study in part explores the morphological properties of buildings that lend themselves to adaptation. Through the use of Conzenian and Space Syntax methods, the authors demonstrated how close spatial integration between urban components, and high accessibility (in areas such as town centres) results in higher rates of adaptation and densification over time. Here the term ‘changeability’ is used, instead of adaptability, to describe the capacity of an urban system to undergo transition in a way that allows it to remain fit for purpose and resilient (ibid., p.5).

Owing to the labour-intensive nature of accessing and working with historical data, as noted in Chapter 2, only small sections of urban tissue were able to visualised and analysed; a key problem for historico-geographical studies in general. Three main data types were identified as necessary by the authors: demolition data, building modification data and change of use at ground-floor level. The first two were derived from historical map comparisons and on-site surveys. Current age data were accessed from local authority records, and spatial change within plots from changes to footprints as shown on historical maps. Historical business directories and aerial and street photos were also used (Torma et al. 2016). The approach allowed for the number of changes, and type of change, since original construction, as well as change to plot coverage, to be tracked and measured, and for analysis to be also undertaken on changes to the accessibility of plots. Small plot divisions, with high permeability, good access, and space for incremental development were shown to enable buildings to remain relevant and fit for purpose for very long periods. For this, data on street networks, plot and building area, plot perimeter, and street frontage width were also all required.

As well as identifying data types of importance in studying ‘development within plots’, the study is also interesting in its finding, as also observed by Jacobs, that to support long-term vitality, some demolition must occur, but that the speed and scale at which this happens is critical. Width of streets was identified as an important indicator of vitality, with high-accessibility roads found to generate too much demolition and change over time, to maintain sufficient diversity of form, and what the authors describe as an ‘equilibrium’ or balance of continuity and change. Street width was therefore also noted as important to collect. Predictors of the capacity of buildings to adapt were identified as: a) small plot size b) a history of repeated use change, c) the ability to accommodate non-domestic use in domestic buildings, d) the presence of adjacent non-domestic buildings, e) the number of historical modifications on the same street segment, and f) the amount of choice, in terms of land use, within a specific distance. These, data types closely compare to those used in Porta and Romice’s work (Porta and Romice, 2015) on Plot Based Urbanism (PBU). They are all considered necessary to be collected within open data platforms and are incorporated into Table 5.1 below.

1.3.5 Micromorphological methods of quantifying change in 3D Kostourou The last, and most recent study discussed, relating to the process of ‘development’ in stocks, was undertaken by Fani Kostourou, and quantified incremental development and densification within the suburban residential area of Cite Ouvriere, in Mulhouse, France. Here adaptation was analysed and visualised, not in terms of precise historical modifications (as in the case of Torma et al. and M.R.G Conzen), but by studying types of dynamic change occurring over a long period of time (Kostourou, 2021). The research involved 1,253 dwellings and four domestic typologies, with change to plot analysed for 520 plots, for sixteen time points over a 165 year timespan. To understand the dynamic process, historical planning applications, historical maps, texts and images, archive photographs, and testimonials were all drawn from. Through meticulous observation of change between each time point, eight ‘dynamic mechanisms’ facilitating adaptation were able to be identified. These were defined as ‘join’, ‘extrude’, ‘extend’, ‘subdivide plot’, ‘add shed’, ‘alter roof’, ‘change entrance place’ and ‘chamfer corner plot’ (ibid.).

Kostourou’s visualisation of these mechanisms, as abstracted forms, is shown in Figure 5.4. Her method also employed ‘Space Matrix’, a tool developed by Meta Berghauser Pont and Per Haupt in Sweden to investigate the relationship of built form to density. This uses four variables: ground coverage of plot area, built intensity of plot area, average number of storeys and open space ratio (accounting for pressure on unbuilt on land). These variables can be plotted on a 2D chart able to graphically represent the mathematical relationship between each, and leading to the automatic grouping of similar types of built form (Kostourou, 2021.; Berghauser Pont et al. 2019). Growth within individual plots was also calculated by Kostourou in terms of the capacity of the 3D unbuilt-on space to accommodate change in future, with ‘potential volume’ afforded by building regulations compared to actual volume. Through this process Kostourou was able to show the impact of individual ‘dynamic mechanisms’ on the intensification of open space consumption, and the cumulative impact on the density and compactness of the area as a whole (Kostourou, 2021). Owing to the study’s 165 year timespan, insights were also able to be gained into the long-term effect of specific policies and planning constraints on the way buildings changed in plots, and typologies muted (Kostouoru and Psarra, 2017).

Kostourou’s unique visualisations illustrate with unprecedented clarity the way in which small-scale updates to domestic buildings are able to occur on an as-needed, when-affordable basis, provided there is sufficient space and the right to extend. These also show how plots can accommodate change and remain fit for purpose (as M.R.G Conzen had also shown) for very long periods of socio-economic and technological change. Her ‘dynamic mechanisms’ are also important in illustrating how areas of homogenous age can increase their age diversity, and transform themselves from a ‘top-down, all-at-once, monotonous plan to a bottom-up, piecemeal, and formally diverse scheme’, provided the above conditions are in place, and that sufficient time is made available for this process to occur (Kostourou, 2019, p.91).

Data on building adaptability in plots therefore also need to be collected within open data platforms to support resilience and lifespan/mortality analysis. This includes information on the type of ‘dynamic mechanism’ operating in plots and, ideally, their date of construction. These mechanisms are also relevant to microsimulation models discussed in Chapter 3, to rule-based ‘procedural’ typology models discussed below, and to the visualisation of diversity within ‘static’ tissue though their colour coding.

1.3.6 Summary of data requirements for quantifying incremental adaptation within plots

Based on this small selection of studies, an additional list of data types to those identified in Table 2.2 (Chapter 2) is developed to inform platform prototype design as shown in Table 5.2. The table also groups data types into a small number of main categories to begin to try to structure information on the stock in a clear and logical way. For example data on plots is included within a main category named ‘Plot & Street’ and height and storey data under ‘Size’. ‘Dynamic mechanisms’ are placed under ‘Dynamics’, also considered suitable for data on historical construction and demolitions required for lifespan calculations. ‘Age’, ‘Land use’ and building ‘Type’, are seen as sufficiently important to warrant their own categories. (Subcategories to be collected within ‘Type’ are tabulated at this chapter’s end). Whether the data type is likely to be included with Stage 1 of platform development (i.e. within the scope of the thesis), or Stage 2 (beyond the thesis period), is also recorded. Building attribute data considered easier to generate or access are prioritised in the first stage, with plot and adaptation data, for example, assigned to the second.

1.4 Building typologies and dynamic classifications

1.4.1 Introductory remarks

In this second section of the chapter, the importance of collecting data on typologies, and their mutation over time, is discussed and illustrated by a diverse selection of urban morphology studies. Kropf describes how the process of evolution of types of element within the built fabric generates complex urban systems (Kropf, 2017). Scheer defines building types – known as typologies when systematically classified – as components having common formal characteristics, within which a wide range of variations may occur (Scheer, 2010). Moudon explains how similar basic typologies can be identified across cities and countries; ‘basic elements of urban form are the same and formative and transformative processes share the same basis’ (Moudon, 1997, p. 9). Steadman makes a clear distinction between ‘activity types’, which describe land uses in buildings, and ‘the geometrical forms in which activities can occur’, which he describes as ‘built form types’ (Steadman, 2014, p.1). These are viewed as ‘historical entities, made up of institutions-in-buildings that are reproduced though copying, with variations’, and which ‘may appear at some date, grow in numbers over time, persist, or else decline and disappear’ (Steadman, 2014, p.3).

In urban morphology, the description of typologies and the study of their mutation has largely been undertaken as part of the ‘typo-morphological’ tradition led by the Italian school of urban morphology. This developed in the 1950s out of the work of Saverio Muratori, whose theories on typologies were largely informed by his work, after the Second World War, coordinating the renovation of Italian historic cities and towns (Cataldi et al., 2002). Muratori felt, at the time, that Italian planners and architects had failed to see the city as an organism (for which an understanding of scale and of the evolution of form was vital) and that because of this the ‘complexity and originality’ of Italian architecture was in danger of becoming lost (ibid., p.4). He set in motion a bottom-up approach to planning, described by Shuyi Xie as ‘redevelopment by tradition’ through ‘a slow and organic piece-by-piece development process’ (Xie, 2019, p.1).

In the late 1970s, Gianfranco Caniggia and Gian Luigi Maffei built on Muratori’s work in their book Interpreting Basic Buildings, in which they explored ‘the dialectic relationship between supporting types and fabrics and the changes that occur at the same time in different places… and the same place at different times’ (Caniggia and Maffei, 2001, p.9). Here, basic building forms are first identified from the study of characteristics of the locality (ibid.). Typologies are then abstracted in 2D, or parameterised, with dimensions removed. Through a step-by-step ‘procedural’ process, driven by what is described as a generative code, derived types are generated, from the base type, through a process of ‘diachronic mutations’. This, as shown in Figure 5.6, mainly involves the repetition of elements horizontally and vertically, with mutations also compared across cities. Through this step-by-step process, complex derived types are created in a way considered to lead, in terms of a block or locality, to a coherent whole (ibid.). An illustration of the practical application of this approach, to support area conservation/reuse/ lifespan extension, is provided by Xie’s 2019 study of Schio (Xie, 2019).

5.4.2 Dynamic urban tissue types

In 2010, Brenda Case Scheer, in her book The Evolution of Urban Form called for a more sophisticated understanding, in architecture and planning, of the way in which types of urban components originate, evolve and transform, to support the development of more sustainable urban design strategies (Scheer, 2010). Whereas Caniggia and Maffei focused on the stages of mutation of specific typologies, Scheer divided the urban tissue as a whole into just three ‘dynamic tissue’ types. The purpose of this dynamic classification was to show how specific configurations of plots, buildings and streets in cities behaved, and were likely to behave over time (ibid.).

Like Batty, Scheer described the way in which most urban designs are based on a ‘static understanding’ of the built environment which she saw as wholly inadequate to support sustainable strategies that need to take account of structural patterns of change (ibid, 2010, p. 2). Like Alexander and Steadman, her particular interest lay in identifying and defining underlying constraints on the development of urban form in order to allow designers and planners to optimise design quality and maximise efficiency. Scheer divided the built fabric into three dynamic tissue types which she called ‘elastic’ tissue, ‘static’ tissue and ‘campus’ tissue (Scheer, 2010). Each is described below.

1.4.2.1 ‘Elastic’ tissue ‘Elastic’ tissue is defined by Scheer as a ‘fundamentally disordered’ tissue that commonly forms along arterial routes. Within this tissue type, commercial, and particularly retail buildings are closely packed, with plot shapes and building typologies of diverse form, size, age and orientation (ibid, 2010, p.53). Her description compares closely to Stanilov’s definition of ‘strip’ development in the US, where high concentrations of non-domestic buildings cluster owing to the ease of access to them along these high-access route (Stanilov, 2002). Scheer describes ‘elastic’ tissue has having no clear pattern since it is ‘created directly by subdivision of individual farm fields or aggregation of small pre-urban houses, without the establishment of a road network’ (Scheer, 2010, p.53). It is ranked as the least stable of the three dynamic tissue types, with buildings described as varying significantly in age due to rapid turnover and obsolescence. However demolition cycles, cited by Scheer as being as short as 20 years in the US, can also, as shown by M.R.G. Conzen, M.P. Conzen and Torma et al., be much longer along this type of commercial route. Stanilov, like Torma et al., observes higher rates of change are likely to occur on wider routes, and views commercial land uses as ‘the leading form-generating element in the evolution of metropolitan form’ (Stanilov, 2002, p. 193).

1.4.2.2 ‘Static’ tissue ‘Static’ tissue, in contrast, is defined by Scheer as a homogenous tissue relating primarily to residential stock found in suburban locations, and is composed of small plots often of uniform size (Scheer, 2010). Streets in ‘static’ tissue, have a high connectivity to buildings, but a lower rate of footfall than along commercial, ‘elastic’ routes (ibid.). Masucci et al. have shown how in London, smaller ‘space filling’ residential road networks, with which this tissue type is associated, will locate themselves in the gaps between, and created by these ‘elastic’ tissue routes (Masucci et al., 2013). This is further discussed in the context of London in Chapter 7.

‘Static’ tissue (seen represented in Kostourou’s study) is characterised by small single-family houses. Plots and streets are ‘planned together, surveyed at about the same time, and originally built on within a period of 10 to 20 years’ (Scheer, 2010, p.51). Scheer, like Kostourou, identifies this type of tissue as highly adaptable owing to the capacity of buildings to extend within plots, at low cost, ‘the simplest buildings, divided into increments of small ownership…have nearly infinite adaptability due to their building configuration, and it does not take enormous capital to alter them’ (Scheer, 2010, p.86). Scheer argues that the capacity of ‘static’ tissue for adaptation, the regular spacing, repetition, and small size of plots, and the fact that development in plots is often controlled by individual owners creates ‘the most stable urban form that exists’ (ibid., p.51) and a ‘persistent element of urban form’ (ibid., p.51). She also argues that the consistency and conformity of the tissue acts as a brake on change, inhibiting radical, large-scale change and discouraging subdivision and churn (and that though domestic land functions may convert and combine, they will not become obsolete, as may occur with non-domestic uses), ‘redevelopment that is inconsistent with the existing fabric is discouraged because it is less marketable and has a chilling effect on nearby properties’ (Scheer, p. 51).
The use of the term ‘static’ refers to this inertia, though it belies the scale of adaptation and plotsprawl that may be occurring in this tissue type, as illustrated, in Chapter 7, in the context of London.

1.4.2.3 ‘Campus’ tissue Scheer’s third dynamic tissue type is ‘campus’ tissue’. This comprises large land parcels containing isolated buildings and internal roads, disconnected from the street network ‘except at limited entrances’ (Scheer, 2010, p.52). Additional characteristics are that parcels are not subdivided when new buildings are added or changed, and are normally under the control of a single individual or body (ibid.). Land use may be domestic or non-domestic. ‘Campus’ tissue is typical of large government-controlled sites such as prisons, offices and housing estates; large private developments such as housing estates, shopping estates and business parks; independently run, public facing institutions such as museums, law courts and universities; and core infrastructure uses, such as airports and public transport buildings and utilities, where a complex public/private ownership/delivery structure may exist.

In Stanilov and Batty’s London study, residential, commercial and industrial sites are described as subject to incremental, organic growth, (in line with Scheer’s ’elastic’ and ‘static’ classifications above), while large hospitals, prisons and airports, typical of ‘campus’ tissue, are described as controlled by centralised decisions and as appearing ‘instantaneously in the landscape and change little over time’ (Stanilov and Batty, 2011, p. 261). Here demolition may be sudden and transformative, causing radical change to an urban area far more quickly than in ‘static’ or ‘elastic’ tissue. Bibby describes decision-making in relation to these types of land use as ‘autonomous’ (Bibby, 1997, p.98). In contrast, where ‘campus’ tissue, is controlled by independent institutions such as museums, law courts, and universities, these follow a ‘general pattern of quasi-organic development that occurs widely on major institutional sites’, and are represented by several morphological periods (Whitehand, 1994, p. 12). The impact of land use and ownership on the scale and speed of development within ‘campus’ tissue parcels, is further discussed in the context of London in Chapter 7.

Scheer observes how in the centre of cities ‘campus’ tissue is less common and parcels smaller (Scheer, 2010, p.84). Simon Jenkins has described the way in which large detached buildings in large plots in central London have, historically, been vulnerable to transformative change (Jenkins, 1975), as pressure to demolish will be higher where higher land values can be realised by subdivision into smaller plots. Scheer, like Jacobs, comments on how merging plots to create large parcels can reduce their resilience and effectiveness of operation, ‘tearing apart of the existing fabric to create a large project, the owner may actually harm the ability of the city to heal itself incrementally through small organic changes. The large parcel may become a long-term problem, since it now requires a powerful or very resourceful entity to adapt it to a new life when it comes to do so’ (Scheer, 2010, p.59–60). She warns that ‘Huge change is risky. Given the choice it might always be better to subdivide rather than to aggregate.’ (ibid., p.87).

1.4.2.4 Computational classification of dynamic tissue types Scheer’s dynamic tissue classification offers a simple method, able to be used within open building attribute data platforms, to spatially group buildings, streets and plots within cities according to their dynamic behaviour. For this data on land use, building type (especially adjacency) and specific street characteristics (e.g. location of arterial routes to locate ‘elastic’ and ‘static’ tissue; and of parcels containing access roads to locate ‘campus’ tissue) are required. Once dynamic tissue types have been geolocated, a dynamic classification can be assigned to each building in the city to indicate likely change to physical form over time, and the rhythm and relative speed at which this is predicted to occur. The potential value of dynamic classification data, for stock forecasting models, and for procedural models discussed below, is considerable. Dynamic classifications can also be verified using recent historical planning data, and/or longitudinal data (e.g. vectorised footprints as described in Stanilov and Batty’s 2011 study). In Chapter 8, semi-automated methods of geolocating dynamic tissue types for London are tested.

1.5 Typology subcategories proposed for open data platforms

The chapter concludes with a second draft table, Table 5.3 (which in Chapter 6 this table is merged with Table 5.2 above), which proposes subcategories for the ‘Type’ category, within the prototype open data platform, based on discussion in this second section of the chapter. Owing to the category’s complexity and the need for further detailed consultation with organisation and disciplines represented within the workshop, the main development of this category is postponed to the prototype second development stage.