X07. London prototype: Data needs in urban resilience and complexity science context. - colouring-cities/manual GitHub Wiki

Author: Polly Hudson, 2020.

1. Introduction

In this second section identifying data types to be included in prototype platform design attention is turned to the importance of spatial data on the physical composition of stocks at the microscale, and on age diversity within them can support stock resilience, and help anticipate negative locked in patterns of dynamic behaviour.

1.2 Resilience, adaptability, diversity, and the speed of change

Resilience is a field of research of increasing interest to many sectors and disciplines, and a core area of study within sustainability science (Hassler and Kohler, 2014). Here, fragilities in systems are assessed as is their capacity to ‘respond, adapt and evolve to different forms of pressure’ (ibid., p. 119). Moffatt, in his work on ecological urban design and long-term energy strategies, views resilience as representing a more sophisticated, dynamic view of the future than sustainability (Moffatt, 2014. Within resilience planning, risks, uncertainty and surprises are assumed, and strategies to address these put in place. Urban resilience is described as the inherent capacity of the built environment, and is related socio-economic system, to maximise the potential of stocks to adapt and recovery from change and loss (ibid.). C.S. Holling’s research into resilience in the context of ecology in the 1970s formed the foundation for much, later, resilience research (Nicol and Knoepfel, 2014). Speed of transformation is viewed here as a critical factor in system optimisation (as also referred to in Chapter 2), with a succession of slow-moving risks and disturbances over long time periods of time seen as more beneficial than fast-moving, extreme, events (Holling, 1973). In his 1973 paper, Resilience and stability of ecological systems, Holling points out that in ecology, important insights can be gained by studying the probability of the permanent loss of elements, and conditions needed for their persistence. He concludes that the recognition of both our ignorance in these areas, and the fact that unexpected events will always occur, are critical for resilience management (ibid.).

In the context of urban systems, Moffatt also states that where a system, or a component within it, ‘fails to persist for a “reasonable” length of time or fails to recover losses within a “reasonable” time period then it is neither sustainable nor resilient’ (Moffatt, 2014, p. 202). In this context, demolition, and lifespan shortening, can be reasonably seen to represent the failure of an urban area, urban tissue, building typology or individual building to be sufficiently resilient to absorb specific types of threat or shock. As such demolition can be viewed as an indicator of vulnerability and lack of resilience. Romice and Porta describe resilient built form as having ‘a high capacity to absorb change, to assimilate transitions, without having to renounce what gives them character and structure’ (Romice and Porta, 2015).

Longitudinal spatial data for sites able to provide information lifespans and survival rates, as well as capacity life extension through adaptation are therefore extremely important to urban resilience research, and to understanding the impact of loss of specific types of building to the stock as a system. Methods of collecting these data types, along with data on current composition, within open data building attribute platforms therefore need to be a priority. Moffatt argues that the speed at which urban systems change is central to their capacity for adaptability and resilience. He highlights the importance of knowledge, held and remembered, about and within urban systems, and of the slow process of trial and error and of age diversity, ‘It is the mix of elements with differing lifetimes that helps the system as a whole survive the exogenous forces of change. Some species respond quickly to change and shocks, others more slowly… All durable and sustainable systems appear to have this sort of structure; it is what makes them adaptable, robust and resilient..’ (Moffatt, 2014, p. 210). Moffatt also stresses that time (required for the generation of diversity) is a critical parameter in the development of resilient urban systems, and that long-term planning horizons appear ‘to be continuously shrinking’ with shorter building lifespans becoming increasingly common (ibid., p. 209).

Scott Page also comments on the close links between resilience and diversity; ‘diversity and complexity lie at the core of many of the challenges that we currently face, managing ecosystems, organisations and economies’ (Page, 2010, pp. 1–2). As well as enhancing system robustness, both Page and Andy Stirling argue that diversity has the capacity to provide insurance and security, improve productivity, spur innovation, generate collective knowledge and sustain further diversity (Page, 2010, p. 3; Stirling, 2007). For Stirling, like Moffatt, understanding diversity in complex systems requires knowledge of how components combine, for which he argues that data on ‘variety’, ‘balance’ and ‘disparity’, within and across component types, are needed (Stirling, 2007, p. 709); these properties of diversity are set out in Table 4.1. Stirling also highlights how the pooling of diverse ideas can accelerate innovation, and generate robust systems able to sustain themselves if and when component parts fail (ibid.), as also argued by Jane Jacobs, discussed below.

The integration of diversity into system design is also of relevance to open data platforms, to increase their resilience and protect against failure. It also confirms the importance of a multidisciplinary and transdisciplinary, collaborative approach . Romice and Porta argue that resilient places are able to absorb change without having to relinquish their character and structure (Porta and Romice, 2015). Like Moffatt, and Jane Jacobs and Christopher Alexander, discussed below, they conclude that resilience in the urban tissue is closely linked to diversity generated through the incremental development of individual parts. Urban form is described as being made up of ‘relatively small components that can adapt, assemble and reassemble’ and that depend ‘on a system of units that maintain their own identity even when combined into greater wholes.’ (ibid.). Issues of adaptation and diversity generation at plot level are discussed in Chapter 4.

1.3 Jane Jacobs: diversity, building age and time

1.3.1 Diversity and cities

The most detailed theoretical framework setting out the relationship between diversity, resilience, and the speed of change, in the context of building stocks, and demonstrating the importance of comprehensive microspatial data on building age and form, was produced in the mid-20th century by the writer, journalist, theorist, and activist Jane Jacobs. Her book, The Death and Life of Great American Cities (Jacobs, 1961), is described by Batty and Yu Xie as ‘a work of almost unparalleled prescience’, in which Jacobs argues ‘that apparent disorder was but a symptom of deeper order and complexity, and that naturally growing organic cities provided much richer and more workable environments than those imposed by city planners’ (Batty and Xie, 1994, p. s32). Here, Jacobs argues that diversity is essential to the generation of vitality and economic success in cities as well as to their sustainability and resilience, and (like Moffatt and others working other in sustainability science) that understanding the role of time in this process is essential, as is studying urban components at the microscale (Jacobs, 1961). Jacobs asked ‘How can cities generate enough mixture among uses – enough diversity – throughout enough of their territories to sustain their own civilization?’ (Jacobs, 1961, p. 144). She explains that for cities to be able to do this, certain conditions or rules have to be met. The first is that more than one primary land use exists in streets, to attract different users to the area at different times of day; the second, that street blocks must be short to maximise points of intense interaction; the third, that density needs to be high enough to stimulate intensive use; and the fourth, that buildings should be of diverse age, ‘closely mingled’, with a good proportion of old (ibid, p. 150–1).

1.3.2 Jacobs’ views on the role of diversity of building age

Jacobs stated that ‘Cities need old buildings so badly it is probably impossible for vigorous streets and districts to grow without them. By old buildings I mean not museum-piece old buildings, not old buildings in an excellent state of rehabilitation – although these make fine ingredients – but also a good lot of plain, ordinary, low-value buildings, including some run-down old buildings’ (ibid., p. 187). Her description of the importance of knowledge of building age; of how age diversity is generated, and how it can sustain successful cities, can be summarised as follows: Diversity of building age provides diversity of building form. This in turn generates a wide variety of types of space, and of affordability of space, offering possibilities for inclusion of a variety of potential uses (i.e. cheaper rents for older, more rundown buildings will be able to accommodate new, small, independent businesses, more easily than new build can; as here rents will need to cover investment on building costs, and will be affordable to fewer, wealthier users e.g. chains). Diversity of building use in turn attracts diverse local populations – importantly, at different times of the day. This means that more people use the streets for a longer period, resulting in areas that are safer, and therefore, in turn, more likely to attract people to them. Mixing of land uses, within streets and buildings, through this process, also facilitates knowledge spill-over and the cross-fertilisation of ideas, thus driving innovation and producing social and economic benefits for the city as a whole. Page also records the way in which mixing, driven by forces of ‘mutation and recombination’, are ‘well-known primary sources of innovation’ (Page, 2010, pp. 8–9).

Jacobs wrote extensively about the role played by older buildings and existing stock in this process. She argued that it was here that diversity was held and generated, and that substantial loss within this incubator and reserve would have significant implications for future diversity generation, and thus city success. As described by Moffatt, Jacobs also stressed that diversity of age, and of form, and consequently of vitality, could only be generated through slow, incremental development, in which there is constantly ‘a mixture of buildings of many ages and types’ (ibid, 189) with ‘what was once new in the mixture eventually having time to become what is old in the mixture’ (ibid., 1961, p. 189). The only harm of old buildings in cities, she argued, is when there is ‘nothing but old age’ (ibid, p.188). Jacobs stated that greater attention needed be paid to the ‘economics of time, not hour by hour through the day… but by decades and generations’ (ibid., p. 189), as time makes the high building costs of one generation ‘the bargain of another’, causes some structures to become obsolete, and allows others to adapt to new uses; ‘The economic value of old buildings is irreplaceable at will. It is created by time. This economic requisite for diversity is a requisite that vital city neighbourhoods can only inherit and then sustain over the years’ (ibid., p. 199). Only through understanding the economics of time, she believed, could the implications of the presence or absence of older buildings be fully understood. In the National Trust for Historic Preservation’s 2014 report Older, Smaller, Better, (which used property tax data relating to building age, use and size for three US cities to test Jacobs’ theories), the strongest economic performance in commercial areas was found where small-scale older buildings of mixed age existed, where areas had been left to evolve slowly over time (National Trust for Historic Preservation, 2014).

1.3.3 Organised complexity and the need to collect data at the microscale

Jacobs also argued that urban problems were very particular kinds of problem, classified as those of ‘organised complexity’ (Jacobs, 1961, p.429); these she said required a specific approach to problem-solving. She described how the term had been used by Warren Weaver, in 1948, in the context of the life sciences, to describe problems ‘which involve dealing simultaneously with a sizable number of factors which are interrelated into an organic whole’ (Weaver, 1948, p. 539). In biological systems, and urban systems, multiple variables need to be considered at once, as these will be ‘varying simultaneously and in subtly different ways’ (Jacobs, 1961, p. 433). Through painstaking empirical study, at street and city level, and across US cities, Jacobs realised, without the aid of computers, that these variables, though appearing to be operating in apparent disorder, could in fact be seen as symptom of order and complexity. She described how three ‘tactics’ are necessary to understand organised complexity in cities and to begin to solve complex problems (ibid, p. 440). The first is to identify and understand processes; that is operations or actions taking place in the city over time. The second is to find ‘unaverage’ clues using volume data. Urban problems were considered unsolvable to Jacobs through the use of statistical averages and small samples alone. She specified that significant amounts of data representing the average in cities were needed to pinpoint these statistical outliers, seen as ‘announcers of the way large quantities are behaving, or failing to behave’ (ibid., p. 443). The third tactic, or rule, was to work at the microscale, always reasoning from the particular to the general, informed by bottom-up knowledge based on empirical study at local level. Though cities ‘cannot go under the same microscope’, in the same way as the life sciences, ‘tactics for understanding both are similar in the sense that both depend on the microscopic or detailed view…’ (ibid., p. 439).

Jacobs also called for the rigorous collection, description and classification of information on stocks. She also advocated increased study into the long-term dynamics of urban systems (and components within them) to gain the necessary detailed perspective, and to see how components are ‘interrelated into the organic whole’ (ibid., p.433). Where her call fundamentally differed from spatiotemporal data collection being undertaken within the scientific discipline of urban morphology at the time, was that it argued for the capture of these data at city scale. Understanding aggregate or more abstract properties of cities could, Jacobs believed, only be grasped in this way. These tactics, relating to processes, citywide datasets, and observation at the microscale, are all applicable to prototype open building attribute platform design, as described in Chapter 8.

1.3.4 Problems with homogeneity of building age and locked-in cycles

The Death and Life of Great American Cities was in large part written in response to Jacobs’ first-hand experience of loss of diversity and uniqueness in her home city of New York, and her voluntary work fighting against this with community planning and historic environment groups. Figure 4.1 illustrates the type of mass housing redevelopment, overseen by Robert Moses, against which Jacobs fought (Auchincloss and Lynch, 2016). In Death and Life., she discusses the socio-economic impacts of destroying diversity by ‘wiping out old buildings in great swathes’ (Jacobs, 1961, p. 198) and their replacement with ‘large swatches (sic) of construction built at one time [which] are inherently inefficient for sheltering wide ranges of cultural, population and business diversity’ (ibid, p.191), and which are inadequate substitutes for the ‘inherent efficiency, in cities, of mingled age and inherently varied overhead’ (ibid., p. 192).

To understand the potential long-term impact of loss of physical diversity in cities, Jacobs studied change to New York’s physical fabric over more than a century. Through local area analysis she was able to identify the way in which cycles of deprivation and ill-health not only repeated themselves (as shown by Dorling and Noble for London) but were also perpetuated by cycles of physical redevelopment. Jacobs argued that the inability of single-age areas to adapt, and their lack of resilience to demolition, could be seen as directly linked to the loss of building age diversity caused by the speed and scale of construction of new development. Where large-scale rapid change occurs within neighbourhoods, she argued, it creates an homogeneity of age and ‘a strange inability [for the neighbourhood] to update itself’. ‘Finally the whole thing must be wiped out and a new cycle started’ (ibid., p. 198). This is discussed in the context of London in Chapter 8. Through her description of this process, she explained how, if the fundamental relationship between time, age diversity and incremental adaptation was not properly understood, locked-in cycles of demolition would continue to repeat themselves; and that diversity, once lost, would, even if the conditions were correct, take a very long period of time to recreate. In this way the potential of age diversity as an indicator of area vulnerability to demolition can begin to be seen.

Jacobs’ views in part reflected those of Charles Booth who had carried out his poverty mapping for London around seventy years before. Booth too questioned the efficiency of large-scale housing clearance programmes, and whether the improvement of slums could not be achieved through adaptation and small-scale adjustments at a slower, less disruptive pace. In 1889, he wrote of one of London’s most notorious slums in Shoreditch, ‘The place deserved destruction. A district of almost solid poverty... in which the houses were as broken down and deplorable as the unfortunate inhabitants’ (Yelling, 1986, p. 51), but of the replacement scheme he noted that, ‘Net benefit has undoubtedly resulted. But it is a question of whether an equal benefit might not have been gained in some gradual, less disturbing and less costly way’ (ibid.), ‘It is all a process of tinkering. Improvement is not coming structurally from a Haussmann...’ (ibid.).

Jacobs believed that the long-term socio-economic costs to society of these cycles of construction, failure and demolition needed to be taken into account when calculating the cost benefits of large-scale redevelopment, particularly in relation to housing. This issue has been raised more recently by Kate Crawford et al. and Anne Power, in the context of social housing demolition, and sustainable city goals, in the UK (Power, 2008, 2010; Crawford et al., 2014). Physically wiping out slums did not, Jacobs concluded, solve socio-economic problems, and in fact was likely to increase rather than decrease the vulnerability of the system as a whole: Wiping away slums and their populations ‘at best merely shifts slums from here to there… At worst, it destroys neighbourhoods where constructive and improving communities exist and where the situation calls for encouragement rather than destruction’ (Jacobs, 1961, p. 270–271).

Importantly, Jacobs considered that large-scale, homogenous new development was highly problematic if incremental adjustments as-and-when needed could not occur, (as further discussed in Chapter 4). Accommodation of change through ‘ingenious adaptations of old quarters to new uses’ was seen as crucial, with new build insertions also essential to improve and maintain vitality (ibid., p. 194). ‘The only harm of aged buildings to a city district or street is the harm that eventually comes of nothing but age – the harm that lies in everything being old and worn out. But a city is not a failure because of being all old. It is the other way round. The area is all old because it is a failure’ (ibid., pp. 188–189). Jacobs believed that if rules and conditions for generating diversity were applied, the tide of departing populations, and cycles of deprivation and disruption could be averted (ibid, p. 285). Her key point, as also described by Kohler and Yang, and Reyna Chester, (see Chapter 2), was that negative locked-in cycles within stocks needed to be understood and averted, and that for this to happen, access to historical information, and to statistical data on stocks, at building and city scale, were required.

1.3.5 Slowing rates of change using zoning/designation tools

Jacobs also described how specific planning tools could support cities in generating diversity and resilience. She described ‘zoning for diversity’ as a necessary method of consciously slowing development speed (ibid., p.252). She called for tools to apply ‘constraints on too rapid a replacement of too many buildings’ – so that stocks ‘cannot be overwhelmingly of one kind...so as to prevent the erosion of economic diversity’, and for incentives and regulation to limit the amount of a single type of land use in a given area’ (ibid., p.253). She also highlighted issues raised by area attractiveness (increased though the protection and generation of diversity) in relation to higher land prices and increased redevelopment pressure. This she explained could lead to diversity loss, and self-destruction being caused by an area’s success.
Area-based demolition controls had been pioneered in the US, in New Orleans, in the 1920s, and from 1969, in large part due to Jacobs’ efforts, conservation areas began to be designated in New York (Institute for Historic Building Conservation, 2014). In the UK today, justification of uniqueness is commonly required in Conservation Area designation, a characteristic closely associated with age diversity (Historic England, 2019b). The importance of uniqueness to the sustainability of urban areas is also discussed within the New Urban Agenda (United Nations, 2016), in UNESCO’s 2011 Historic Urban Landscape Recommendations (UNESCO, 2011), and in the World Bank’s report, The Economics of Uniqueness (Licciardi and Rana Amirtahmasebi, 2012). Johansson argues that unique areas, where designated, should be reclassified as non-renewable resources (Kohler and Hassler, 2002, p. 234). Globalisation, says Paul Knox, has created a fast world ‘of restless landscapes in which the more places change the more they seem to look alike, the less they are able to retain a distinctive sense of place, and the less they are able to sustain public social life’ (Knox, 2005, p. 3). Hassler argues that conservation policies should be central to future strateges designed to decelerate churn in the stock as a whole (Hassler, 2009). Open data platforms therefore also need to capture data on the protection and designation of buildings and areas, to illustrate where spatial constraints are being imposed on development speed, and to highlight buildings classified by the city as positively contributing to its operation. Jacobs work reinforces the importance of including precise age data within open building attribute data platforms, and of the potential value of this data type in forecasting area vulnerability and resilience.

4.4 Bottom-up, rule-based, spatiotemporal microsimulation models Jacobs’ theories on diversity and complexity were to influence the work of Mike Batty, and his research into organised complexity in cities, and hierarchies and codes’ of operation that determine patterns of urban growth (Batty, 2007). Batty realised that cities contained too many variables and interactions to use traditional methods of urban modelling (ibid.). From the 1990s he began to move away from top-down models using aggregated data on land use and urban form, and began to experiment with bottom-up, microspatial, visual, time-conscious microsimulation models. These focused on urban processes, feedback loops, complex pattern interpretation at ‘cell’ level, and on local interdependencies and their relationship with urban systems as a whole (Batty and Xie, 1997). Batty recognised, like Jacobs, that gathering microspatial information on dynamics was central to this process: ‘dynamics of course represent the key…. Although we know how incremental change feeds on past change to generate its own momentum through the realization of scale economies, we still have little clue as to why such change takes place in the locations it actually does’ (Batty, 2018, p.9). Batty quotes John Holland’s description of the city (from the 1990s), as a ‘pattern in time’, in which no single constituent element remains precisely in place, but the city persists and maintains coherence without it being understood how (ibid., p.8).

Batty began to test computer models, known as cellular automata, in which local ‘rules’ of operation are iterated, to generate unpredictable ‘emergent’ patterns at much wider geographic scales. First developed by John Conway, in 1960, for The Game of Life, these models allow the user to experiment with the survival and mortality of components within a system, using what Conway described as ‘genetic laws’ (Gardner, 1970, p. 120). Here the computer programme iterates rules at the local, or immediate neighbourhood level, to effect the ‘life’ and ‘death’ of cells on a grid.

Batty’s initial interest in these models stemmed from his work on computer graphics and pattern visualisation, and his research into fractals with Paul Longley (in which irregular urban structures were found to repeat themselves and to be ordered hierarchically, across many scales) (Batty and Longley, 1994). Batty and Longley showed that cities follow certain rules of growth, through agglomeration and economies of scale. They likened the process of urban growth to the generation of life itself and argued for the need to identify and reveal ‘generative codes’ driving the survival, adaptation and mortality of urban components: ‘Like all natural growth, they [cities] evolve through the cumulative addition and deletion of basic units, cells or particles and they thus grow through successive accumulation at these basic scales…. Within the growth processes are codes which determine how the organisations of these basic units of urban development might repeat their form and function across many scales (ibid., pp. 228–229). Batty and Longley viewed the organically growing city, developed slowly over time, as being optimal in countless ways. The ability of simulation models to mimic the ageing of cities was also raised, ‘One could assume a process of aging and renewal, varying according to simple rules and policies, which would then enable a pseudo-dynamic simulation to be developed. A series of images of the typical house types in London over the next 50 years could be generated in this way’ (ibid., p. 161). ‘But’, they said, ‘these are for the future, and in any case there are many lines of inquiry that have to be followed up before then’ (ibid.). This idea is explored in the context of open building attribute platform simulations in the following section. Batty saw, like Jacobs, that large quantities of microspatial data were essential to reveal wider patterns with complex systems, held together through modularity and hierarchy, where modules repeat themselves at different scales (Batty, 2007). To simulate organically growing cities, modular algorithms were proposed, allowing many variations of the city or urban area to be generated from a basic block, but from different sets of local rules. These models also tested the urban theorist Christopher Alexander’s idea (discussed further below) that it is the ‘genetic code’ of the block or module that allows different variations of it to grow or not grow. The aim of testing microsimulation models was also to develop ‘knowledge of the same phenomenon at different levels or scales’ to allow it to be ‘synthesised consistently into an ordered understanding’ (Batty and Xie, 1994, p. S31). In Batty and Xie’s 1997 study of Amherst, Buffalo, US (Batty and Xie, 1997), property tax age data were used in a cellular automaton model designed to explore patterns evolving from local rules, according to macro laws which were not built into the model (ibid.). Building age data were used to spawn growth from multiple sites for which different local rules of mortality and survival were iterated, resulting in entirely different patterns of growth. Though Batty and Xie concluded that the method was of value, longitudinal spatial data, at the microlevel, were identified as being necessary to advance the work (ibid).

In 2002, Kiril Stanilov published a study of suburban growth in Greater Seattle (Stanilov, 2002). Within it, aerial photography, for a series of time points from the 1960s onwards, was ‘carefully assembled together, piece by piece, to visualize the sequence of incremental transformations in the metropolitan fabric’ (ibid., p. 192). Through this process, the spatial distribution of land uses and change to them over time were explored. Like Jacobs, Stanilov argued that a more systematic approach to the structuring and classification of data was required. He also commented on the difficulty with access to, as well as the predictive power of, historical data ‘Documentation on historical development patterns is scarce and sporadic yet it is an invaluable tool in reconstructing our past and predicting our future…a detailed set of historical records of past development patterns can be used as a powerful educational and analytical tool for visualizing the dynamics of growth and the identification of development trends’ (ibid., p. 174).

In 2005, Batty and Stanilov began work together on a five-year visual study into the spatial determinants of the growth of London, using a 400km2 sample area. For this, a cellular automaton model was built to investigate macro rules of urban growth (Stanilov and Batty, 2011). The model addressed what the authors described as the ‘Achilles heel’ of physical models of urban form, namely ‘their reliance on hypothetical or highly aggregate data’ that smooth out complex patterns, meaning that fast and slow dynamics cannot be easily tracked (ibid., p. 254). London was chosen as the system of interest as it offered ‘the opportunity to trace the patterns of urban growth with the highest level of precision’, owing to the amount of historical information, and number of OS map intervals from the mid-19th century, available (ibid, p. 258).

To programme the computer model, rules of change at the local level first had to be identified, recorded and defined. For this, thousands of features from the OS historical maps were manually vectorised by Stanilov; historical patterns of development were then studied. These vectorised features included small parcel polygons, and street and railway network segments. Land use and building typology information was also recorded at parcel level. Local rules of change were generated by analysing changes to the parcel polygons between 1875 and 1896, and then between 1896 and 1915, and by assessing change in land use from one period to the next, as well as by studying change to stations, the Central Business District (CBD) and suburban cores. Using a grid of small (25m x 25m) land use ‘cells’, observations on consistent repetitive, spatial behaviour were made in relation to cell location, land use of the cell’s neighbours, distance from various elements of the transport network, and the application of zoning restrictions. To be able to test the accuracy of the local rules employed, the OS 1915 survey was used as a proxy for the present day. Spatiotemporal change was then simulated at approximately twenty-year increments from 1915 up to 2005. In Figure 3.2, the remarkable power of this method can be seen through the similarity between the computer-generated

predictions of change for post-1915 intervals, and actual development shown on post-1915 OS maps. Using this method, Stanilov and Batty identified ‘enduring spatial relationships’ within the built fabric (ibid., p. 255) and were able to show that dynamic behaviour was strongly influenced by a general inertia and resistance to change and by the pre-industrial urban structure (preurban being classified as pre-1800). They concluded that ‘the ordering of activities in urban space is conditioned by rules that transcend historical circumstance and reflect deeper structures of spatial organisation’ (Stanilov, 2012, p. 40), and that changes in economic circumstances and political and technological regimes only act as an ‘overlay’ to these fixed spatial relationships. For example, the extent and spatial distribution of new residential development, when analysed in relation to locations such as the Central Business District and to railway stations, over time, was found to be very similar for each time interval (ibid.).

Rules of growth were noted to operate on two levels. Firstly rules were found at the microlevel in relation to the attraction and repulsion of land uses over time (e.g. high-density residential and commercial use were attracted, and residential and industrial uses found unlikely to collocate) (ibid.). Here a ‘generative code of urban development’ that governed the shape and growth of an urban area over the course of its existence were identified. Secondly, rules of growth were found to exist at the regional level in relation to the accessibility of the main transportation network, and in particular to the preurban street network, the Central Business District and suburban activity clusters, which had ‘decisive influence’ on long-term spatial patterns of land use (Stanilov and Batty, 2011, p. 256). These elements were described as ‘largely exogenous to the code of development and, once determined, the code adapts to the form that is laid down from above’ (ibid.).

The same four components, identified by Stanilov and Batty, and Jacobs, as being necessary to study complexity in the city, and processes, patterns and relationships within the urban fabric, were also identified by Frank Brown (Brown,1985). These were: the development of hypotheses for local rules of change (including constraints); a computational approach; access to historical resources, and a ‘microscope, so to speak’ with which to view the city (ibid, p79). Brown used computers, in the 1980s, to test hypothetical rules for the way in which London’s urban fabric developed during the medieval period. Like Stanilov and Batty, he concluded that ‘the urban fabric was generated according to consistent local rules…, checked and limited by various physical constraints’, revealing a deep structural logic (Brown, 1985, p. 79). Only through the four step process described above did Brown believe that patterns and rules could be understood at the city scale, ‘Because the urban fabric was locally generated by an organic process of piecemeal development, one may infer that any order discernible on a large scale had its origins in decisions taken at local level. It is these local decisions concerning the way buildings were put up that we need to identify if we are to decipher the code that lay behind the generation of the medieval urban structure’ (ibid., p.78). Like Christopher Alexander, Brown argued that, ‘to be reasonably authentic, patterns of connections should unfold in a step-by-step fashion, rather than be laid out as a whole’ (ibid., 1985, p. 89). He also highlighted the problem of accessing data to generate models mimicking this process; a ‘major difficulty is encountered for these [models], for all important spatial decisions are seldom of the kind that find their way into documentary record’ (ibid., p. 78).

This group of studies provides evidence of the potential, and importance of longitudinal spatial data to increase accuracy in the long-term forecasting of stock behaviour. They also show how underlying relationships between urban components, and constraints to development, can be revealed when these large-scale time series datasets are combined with microsimulation methods. In Chapter 7, data from the Batty/Stanilov study are used to test semi-automated approaches to large-scale open age data generation. Longitudinal data are also used to forecast demolition patterns.

In Appendix 4, Jacobs’ findings are incorporated into a conceptual colour model exploring the value of integrating interactive age and lifespan microsimulations into open data platforms. Before describing this model, a brief introduction is given to microsimulation models which use longitudinal data to reveal (as Jacobs had done) underlying patterns and rules operating within the urban structure. The model raises many questions: What, for example, will become of ghost cities in the next twenty years? How can help prevent single-age cities being subject to ongoing cycles of demolition in future, where all buildings ‘fail’ simultaneously, and allow time for the process of incremental development to generate diversity of age and form, and resilience? Should, as Hassler and Jacobs have argued, existing designation/zoning tools be applied at city scale to effect a more precise calibration of rates of change within planning systems? And could interactive rule-based simulations such as this, allowing sliders to be moved to calibrate change and simulate scenarios, based on local rules of change, using historical and current data, be incorporated into open data platforms? Would this not help policy makers and citizens to test and visualise sustainable and unsustainable trajectories?

Colour simulations of this kind also have the potential to illustrate the ease and speed with which uniqueness, resilience and vitality can be lost in cities, the difficulty of its reconstruction, and the scale of energy and waste flows (physical, socio-cultural and economic) associated with such loss. Though the development of such simulations is beyond the initial scope of the prototype first features designed to collect data are developed for the Dynamics section as discussed in Chapter 9.