unit-code
The project envisions life in a flooded London within the River Lea area, addressing the critical need to prepare for and adapt to challenges of rising water levels. The project begins with a meticulous site selection process, emphasising the strategic importance of the area as a microcosm of urban resilience challenges.
Through comprehensive site analysis utilising supervised and unsupervised machine learning algorithms, the team identified suitable locations for interventions aimed at enhancing city resilience, including urban farming, cultural industry development, biodiversity restoration, and improving accessible mobility and public space. These interventions are informed by a diverse range of datasets, comprising population and building density, biodiversity scores, and integration values, to ensure their effectiveness in addressing the challenges posed by flooding.
Design strategies centred around the use of light, flexible, inflatable, and floatable structures are proposed. These offer responsiveness to changing environmental conditions. Ebb & Evolve envisions a more resilient and adaptive urban future in the face of the climate emergency and environmental uncertainty.
As the climate emergency accelerates, London, like many cities, faces increased flood risk. However, this challenge offers opportunities to reimagine urban design, creating cities that not only withstand water but thrive in harmony with it.
Macro-level flood analysis reveals the impact on transport, economy, and vulnerable groups. Unsupervised machine learning (ML) shows diverse clusters, from which the River Lea was chosen for its socio-cultural, ecological, and flooding dynamics.
A selection of datasets was used to find the vulnerable regions along the River Lea. Based on this, seven different zones were identified and studied. Unsupervised ML tools were also used for the meso-level analysis.
A flooding simulation at meso-level helps envisage potential conditions of inundation and flooding.
Combined data of industries, population density, integration values, and biodiversity scores were used to categorise areas into 'biodiversity functions,' 'commercial and communal functions,' and 'cultural industry functions'.
Generative adversarial networks (GAN) analysed building density, coordinates, and income data to assess flood resilience. Low-resilience areas were refined by filters, and clustering identified high-priority zones for targeted interventions.
Design strategies for varying site conditions use 'Light' and 'Flexible' structures for adaptability. Guided by 'protect' or 'connect' principles, these interventions either link people to safe zones or provide protection during floods.
The function that each point would serve was determined using a generative algorithm, tailoring interventions to the site’s needs.
Extensive form exploration ensures that the outcomes are versatile, adaptive, and ready to face the challenges posed by flooding, creating safe and resilient urban environments.
Based on micro-level flooding simulation and clustering analysis, the area was divided into three distinct zones.
Based on previous analysis, three primary axes emerge, pinpointing the core areas. The three main zones are integrated using these axes. This layout ensures efficient navigation and utilisation of space, setting the stage for development.
Preserving key buildings, pathways were overlaid along the axes: commercial, biodiversity, and community. Four types of Interventions align with these axes, protecting heritage, residents, biodiversity. and information exchange.