Modelling and simulating urban expansion in Africa

I will address an important gap in quantitative urban research by developing an openly accessible model for predicting patterns of spatial expansion in African cities. According to UN estimates, Africa's cities will house almost one billion additional people over the next 30 years (UN, 2018). Much of this increase will occur in informal settlements. Predicting geographic patterns of urban expansion is crucial for planners and policy makers seeking to improve urban living conditions while minimising the impacts of urban growth on energy consumption, pollution and ultimately climate change. Openly accessible global urban models that are sensitive to local conditions provide useful decision support tools for effective urban and regional planning.

In this project, I will pursue four main objectives:

(1) Develop an urban expansion model for African cities that can be parameterised to account for the diverse socioeconomic and regulatory conditions that underpin processes of urban development on the continent;

(2) Validate the model in a range of primate and secondary cities across the continent;

(3) Make predictions regarding the likely locations and legal status of new residential developments in a large sample of African cities; and

(4) Disseminate the model to academics and practitioners through open access publications, an interactive website for the model, and training local planners and African researchers on how to use and modify the model to suit their needs.

Achieving these objectives will make a significant contribution to the field of urban analytics and city science by addressing the persistent 'Western bias' in urban modelling. The most popular urban expansion models are not well-suited to contexts of 'informal urbanisation'. Further, existing urban models are poorly validated in informal contexts. For example, recent global projections of future urban expansion employ SLEUTH, a popular urban model based on a geophysical simulation approach. However, SLEUTH, as with other urban models that employ this simulation approach, does not model the social processes and institutional dynamics that generate the geographical patterns we observe. While SLEUTH's performance at a coarser regional scale is generally good, it does not perform particularly well in capturing local urban growth patterns.

My research has shown that SLEUTH explains less than a fifth of the built-up expansion observed in Accra - a fast-growing city in West Africa experiencing informal urbanization - from 2000 to 2010. By contrast, TI-City, a model I developed to predict patterns of urban expansion in Accra explains up to 65 percent of the observed built-up expansion in the city over a decade. Unlike SLEUTH, TI-City accounts for the institutional and socioeconomic factors that shape decisions of households and developers by combining an agent-based model with a geophysical simulation approach. While the model is promising, it was developed based on urban expansion processes in Accra, and it is not a yet general model of the African city.

Given that roughly 40 percent of future population growth will be absorbed by African cities, the shortcomings of the urban models matter. To improve current and future decision making about planning and resource allocation in Africa, there is the need for a new general urban expansion model, which accounts for the diverse socioeconomic conditions and regulatory contexts that shape urban development across the continent.

By developing a general urban expansion model for African cities, this project will provide an openly accessible decision support tool for promoting effective urban planning and resource allocations thereby improving urban living conditions across the continent. The development of an interactive website, training of local planners, and publication of academic papers will help maximize the benefits of this project.