A multi-scale integrated assessment model to support urban sustainability

Tools purposed towards supporting the transition to more sustainable urban futures typically focus on specific phenomena at the local level. Whilst such approaches remain valuable, there is a need to complement this micro approach with broader integrated methods which deal with the interaction between different urban components as well as their relation to processes and policies enacted at higher scales. Through the adaptation of the World3 global model of Meadows et al. (The limits to growth, Universe Books, New York, 1972; Limits to growth: the 30-year update. Earthscan, London 2005), integrating both an urban system layer, and a national data layer inputting new data, we develop a proof-of-concept multi-scale integrated assessment model. This model is used to explore the relationship between the sustainability of the urban system relative to higher-scale contexts. By emphasising feedback, cascading effects, and unintended consequences, such a modelling framework allows for deeper consideration of coupling mechanisms between subsystems both within the urban system and across broader scales. Following the description of our model, we take Meadows et al. (2005)’s ‘Scenario 3’ as a starting point to generate several scenarios exploring potential intervention taken at the level of the individual urban system to tackle food security and localised pollution. Our results demonstrate that the evolution of the urban system is sensitively dependent on wider global events, and that while concerted intervention may mitigate some effects, the future of an individual system is largely at the mercy of the evolution of the global system. We argue that the results of this exercise suggest an important role for multi-scale models for informing the wider context of policy measures taken across different hierarchical scales. In an extended discussion section, we outline barriers and potential routes for building our work beyond a proof-of-concept relating to data, boundaries, politicisation, and building confidence in model outputs.