Urban Complexity – How Complexity Science Helps Explain and Predict City Dynamics

Urban Complexity Science and Complex Systems in Cities

However, cities are considerably more than just the sum of the buildings, streets and spaces they contain and are actually the result of an iterative evolution of countless interactions that take place between various networks of society and infrastructure that overlay the physical, determining the urban reality we all experience.

The science of complexity and the way in which individual elements, processes and interactions come together, plays perhaps the most crucial role in explaining not only how cities function, but also how they evolve and grow organically over time.

In understanding the way in which these complex systems form and play out in the urban environment, it’s possible to gain insight into the advanced dynamics that operate within the city space.


The Science of Complex Systems

When thinking about complex systems, some of the best examples can be found in nature.

Take for example, the human brain.

Ultimately, our brains are made up of somewhere around 100 billion individual neurons and when taken on their own, the operation of each of these cells is quite simple.

Put them together however, and these collectively self-organising units combine to provide us with cognition and the most advanced intelligence that we currently know of.

Ants are another demonstration of this principle in nature. Again, when taken individually, each insect operates at an extremely simple level, yet at the scale of the larger colony, the group is able to successfully accomplish complex tasks, without the need for a central point of control.

The fundamental properties of complex systems are as follows:

  • They are typically composed of many (but usually quite simple) individual components
  • The components of the system interact with each other in a nonlinear way – a small change to one component can have a large knock-on effect on the whole system
  • There is no centralised point of control and complex systems will tend to organise themselves in a decentralised way
  • These systems exhibit emergent behaviour – changes in overall behaviour of the system cannot be predicted by behaviour changes of the individual components
  • This kind of system has the tendency to evolve over time and can usually adapt to changes in the environment

So where does complexity science fit into all of this?

The ultimate goal of complexity science is to uncover regularities in the overall behaviour of any complex systems and in doing so, to come up with simple mathematical rules which can explain and (over time) help predict this behaviour.

To achieve this, complexity science adopts a system-level or ‘big picture’ approach to reveal regularities in how complex systems operate and utilises a variety of methods ranging from scaling (relating a system’s performance to its size) to network theory and agent-based modeling.

Given the similarities in the way various complex systems operate in terms of common behaviours, complexity science also looks to learn from other scientific disciplines where patterns and methods of study can be applied across knowledge domains – for example, are there characteristics within the behaviour of an ant colony which could be applied to the study of complex systems operating at the city scale?


Complexity Science and the City

So with all of this in mind, how do complex systems operate in the built environment and how do they influence and shape the paths cities take in the way they organise, evolve, develop and grow over time?

Complex systems at the city level take many forms and you don’t need to look far for examples of these kinds of processes in action.

Imagine for example, a situation where a new transport hub is due to be developed in a relatively quiet neighbourhood of the city.

The opening of this site will bring with it new people who now visit the location and a subsequent increase in new social contacts and interactions which come off of the back of this. New business will likely follow this change and may in turn attract more people to the area which subsequently repeats the cycle over further iterations.


Scaling, Big Data and the Prediction of City Dynamics

One of the core elements of complexity science is the theory of scaling, and it’s with this principle in mind that the ever-growing availability of big data is increasingly being applied to the field of urban complexity science to better understand the dynamics of cities.

In areas such as biology, the concept of scaling is well known as a measure of complex systems in assessing for example, the way in which aspects or characteristics of a biological entity change as a function of body size.

It would therefore follow that if urban dynamics abide by the same underlying principles as biological bodies, it may be possible to observe similar outcomes when the idea of scaling is applied to cities.

This is exactly the question that was asked by scientists at the Santa Fe Institute when they applied the concept to cities, applying a range of different socio-economic data points (wages, new inventions, crime levels, rates of infectious disease etc) against the population of a number of cities around the world.

The results of the study showed that there is indeed a positive correlation between increasing population size in cities and almost all socioeconomic quantities that were measured against this growth metric.

The exponent of the power law that was observed in the study was larger than 1 (taking a value of around 1.5) meaning that not only were the measured socioeconomic markers increasing with city size, but that they were doing so at a rate that was actually faster than linear with population of urban areas – a phenomenon referred to as ‘superlinear scaling’.

If taken at face value, the results of this research would suggest that as complex systems, there are certainly elements of behaviour in cities that can be predicted given the availability of a certain amount of information.

In the era of big data, this of course means that the opportunities for collecting such information and using it to understand and predict the complexity of the city and its dynamics is now greater than ever.

With the ability to utilise once unimaginable levels of data from the level of the individual (gathered via mobile devices and other similar interfaces) through to metrics captured at the level of city infrastructure, the potential for uncovering detailed insight into the systems of, and interactions within the urban space are immense.

As our data footprint continues to expand, our machines and models get smarter, and the field of complexity science develops, it’s clear that the opportunities for creating an urban environment that’s more sustainable, resilient and efficient will inevitably increase exponentially.

Indeed, with the right combination of models and data in place, the role of planning and development at the citywide level in the near-future is likely to continue to evolve significantly as it becomes possible to predict (with an increasing degree of accuracy), the effects of local measures on the larger urban scale before a single change is ever made in the real world.