Abstract Body

This research seeks to improve the theoretical foundations of sustainable urban design in relation to water. A systematic framework to organise water-related urban design research and practice could help develop effective ways to reduce the impacts of urbanisation on the water cycle. However, existing methods, such as 'water-sensitive urban design' conceptualisations and neighbourhood sustainability assessment tools, emphasise urban design practice, lacking a complete evidence-based model of the water cycle in cities.

The research objective of this study was therefore to construct a novel spatial model of the water cycle in cities. This task required synthesising and organising scientific evidence from the environmental sciences. A system dynamics methodology aided by a Driver-Pressure-State-Impact-Response (DPSIR) framework was applied to the comparative analysis of the natural water cycle and its altered version in cities. An undisturbed cool forested landscape was used as reference natural landscape as this kind of landscape presents the most complete and desirable version of the natural water cycle to be emulated by cities.

The resulting model brings together those urban spatial features involved in generating water cycle mechanisms and processes. While some of these features can be found scattered within the sustainable urban design literature, many have not been linked to the water cycle until now. Another contribution of the model is its structure including a new hierarchy of four three-dimensional layers of the water cycle proposed: the precipitation layer, the canopy layer, the surface layer, and the groundwater layer. These layers organise the variety of processes of the water cycle and help assign causal relationships between them and urban form elements within each layer. The model also points at analogies between structural elements in forests and cities. From these analogies criteria and indicators for urban landscapes to mimic the reference landscape can be deduced.