the quantitative science of urban form
What is Urban Morphometrics
Pretty much like human beings, and indeed any living creature, cities are all distinctively different from each other to our sight, and yet all share inner patterns of order. Such patterns are embedded in space and are unique to each any recognisable type of place. It is precisely because of this implicit spatial order that the endless diversity of cities’ explicit, visible traits makes sense to our eyes, and places are somehow distinctively unique and recognisable to us, rather than chaotic. There sits the “fingerprint” of urban areas, their intrinsic character collective wisdom evolved through generations of bricks and mortar.
Urban morphometrics (UMM) is an emerging method of urban morphology analysis designed to extract the inner spatial patterns that distinctively characterise urban places, in the form of numerical place codes. The method brings together a detailed description of the form of cities with an extra-large scale of coverage, to generate a completely unsupervised, systematic and hierarchical classification of urban form.
Measuring urban form
The patterns extracted from the 2 sets of inputs are built using 300+ indicators that comprehensively describe the basic components of urban form, inclusive of its dimension (e.g. building footprint, street segment length), connectivity (e.g. proportion of 3-way intersections, local closeness), spatial relations (e.g. building alignment to streets, mean inter-building distance), and intensity (e.g. buildings per meter of street, floor area ratio). The process uses advanced geo-data processing techniques and an analytical architecture purposefully designed to be scalable up to unprecedented coverage.
Being written in Python, Urban Morphometric is flexible, allowing a full characterisation of urban form, but also the exploration of single indicators to tackle specific research questions.
Classifying urban form
At an urban scale for example, urban form can now be correlated to the demographic, social, environmental or economic performance of places, e.g. health, equality and deprivation, community prosperity and real estate value, energy consumption and walkability, resilience or safety. UMM allows us to study the specific contribution of different urban forms to such crucial dynamics; we can do so at local scale, metropolitan scale with excellent precision.
UMM also allows us to perform comparisons between distinct urban settlements, once a profile is built for each of them. Again, here the questions which we can ask are many, to do with each settlement’s performance against social, economic or environmental measures, or with their physical form for example to establish commonalities and peculiarities.
Fundamentally, with a solid, precise and scalable knowledge base as the one provided by UMM, we can interrogate settlements as never before, and gain precious information for their design, assessment and management
Designing urban form
UMM has the potential to support design and facilitate the informed and place sensitive co-production of our environments, not to substitute them. We in fact believe that design should be guided by evidence.
By describing with precision how urban form is and how it varies in the city and even between cities, UMM allows us to understand how aspects of life unfold in our built environments, and thus make judgements, in terms of what is efficient, accessible, desirable etc. This evidence is important for the design and occupation of space, to make informed decisions and identify the suitable actors to deliver them at the many stages of development.
Design Codes. If we understand how urban forms perform according to certain expectations, we can guide the development of new ones, or the regeneration of existing ones, using physical traits that are similar to those we consider most successful, or desirable. Design codes can be built using ranges of particularly distinctive (discriminatory) characters identified when we profile Urban Types and can be used to instruct the development, over time, of urban fabric whilst maintaining a degree of control on it. This does not replace professional expertise nor local knowledge, but allows to include performance monitoring in the design and delivery process.
- See: Kochi, and in particular the distinctive characters that we produced of its urban types.
- See: Leaflet on UMM and Design Codes.
Generative design. Urban designers use a relatively finite set of physical parameters to design our environments. We can draw these from UMM for each urban type, and build a process to recreate these types. Whilst ‘re-creation’ of forms is never the goal of design, it is useful in understanding the character of place and making sure that our intervention is considerate to it.
As for Design Codes, the purpose of Generative Design is not to replace professional and local expertise, but to increase awareness of place, so that development can be calibrated to it, or measured against it.
- See: Kochi. and in particular 4 examples of generative design of one urban type. Caption: Produced completely independently, they show that the character of place can be delivered in a range of ways. Their essence is similar, but their overall form is not, ensuring the important variations that characterise our cities.
Tools for urban morphometrics
Urban Morphometrics is based on two main parts: the comprehensive characterisation of the built environment and the identification of patterns of urban form. The first is carried out by computing a comprehensive set of morphometrics through the momepy Python library, developed in the context of a research project at the Urban Design Studies Unit (UDSU), University of Strathclyde, and supported by the Axel and Margaret Ax:son Johnson Foundation. The second is achieved through the implementation of several code notebooks, performing clustering procedures on the morphometrics to identify urban types (homogeneous patterns of urban form). Being written in Python, one of the best known and easy to use programming languages, Urban Morphometrics is flexible and adaptable. You can use the full procedure (measuring + classifying urban form) or parts of it, based on your research needs.