Fifty four per cent of the world’s population live in cities. Expansion of urban land during the first three decades of the 21st century is predicted to be greater than the cumulative urban expansion in all of human history. To better understand the societal and environmental implications of this growth, scientists have turned to data science to predict urbanisation.
‘Some of the biggest environmental changes we’re seeing nowadays are occurring at a very large scale, over long time periods. To guide urban planning policy, we need accurate models to study the potential impacts of urbanisation in the long term,’ says Jing Gao, data scientist at the University of Delaware.
Gao and her team have developed a computer model that uses 15 of the best global datasets, including accurate satellite data over 40 years, to predict how urbanisation will unfold up to 2100.
First, the team traced urban history to characterise the process. ‘Mining historical data revealed three different urbanisation styles: rapidly urbanising, steadily urbanising and urbanised,’ says Gao. ‘Over time, countries evolve through these stages.’ According to the data, India and China have experienced rapid urban development, but have now transitioned to steadily urbanising nations. More rapidly urbanising countries in the present day are primarily low-income developing countries, such as Kenya.
Next, the team identified drivers of urbanisation that propel countries through the stages. ‘Our data showed that urban land expansion in rapidly urbanising countries is highly linked to the rate of population growth, as countries bolster basic infrastructure such as roads, schools, and hospitals,’ says Gao.
To then create a predictive model, the team blended a set of hypothetical societal trends (known as Shared Socioeconomic Pathways (SSPs)) with their land data, to predict how urbanisation will change under different models of development.
The SSPs, which are often inputted into key climate models, offer a range of pathways that the world could take, ranging from sustainability: where population growth is low, resource-intensive lifestyles are reduced, and environmental wellbeing is championed; to fossil- fuelled development: where accelerated globalisation is driven by material-intensive economies, with low concern for environmental impacts.
The model showed that globally, urban land could expand from 1.1 million square kilometres (the size of Ethiopia) under a sustainable development plan, to 3.6 million square kilometres with fossil-fuelled development. The extent of urban expansion at the country level depends hugely on the stage each country is at. Urban land expansion in the United Kingdom (an urbanised country) may increase by 1.3–4.4 times from the present to 2100, depending on a societal route of sustainability or fossil-fuelled development. In Kenya (a rapidly urbanising country), this ratio rises to 5.8–16.7 times.
‘We hope our long-term model, driven by societal trends, will allow us to evaluate present decision making on urban planning. If we can model how the future may look under different scenarios, then we can design policy and orientate society to favourable outcomes,’ says Gao.