Soutenance de thèse : Vincent PONS

The future of Green Infrastructure: From climate data to informed hydrological performance

Doctorant : Vincent PONS

Laboratoire INSA : DEEP

Ecole doctorale : ED162 : Mécanique, Energétique, Génie Civil, Acoustique de Lyon

The 21st century presents numerous challenges to urban stormwater management, including the impacts of changes in both climate and city morphology. These challenges necessitate rethinking the stormwater management paradigm, particularly in the context of existing and ageing infrastructure. This thesis deals with green infrastructures (GI) considered as decentralized multifunctional infrastructures that utilize evapotranspiration and/or horizontal and vertical infiltration to achieve a hydrological function.
This study evaluates the potential of GI to manage day-to-day rainfall events, attenuate major events, and contribute to the management of extreme events in the context of climate change adaptation. It also aims to provide a framework and tools to realign current GI modelling and design methods with the principles of robust decision-making.
The 21st century presents numerous challenges to urban stormwater management, including the impacts of changes in both climate and city morphology. These challenges necessitate rethinking the stormwater management paradigm, particularly in the context of existing and ageing infrastructure. This thesis deals with green infrastructures (GI) considered as decentralized multifunctional infrastructures that utilize evapotranspiration and/or horizontal and vertical infiltration to achieve a hydrological function.
This study evaluates the potential of GI to manage day-to-day rainfall events, attenuate major events, and contribute to the management of extreme events in the context of climate change adaptation. It also aims to provide a framework and tools to realign current GI modelling and design methods with the principles of robust decision-making.
The thesis investigates how to use climate and hydrological present and future data with hydrological GI models to extract relevant information for decision-making under deep uncertainty. The results provide guidelines for i) designing experiments to calibrate reliable hydrological models and ii) using available climate projections together with weather generators for GI performance evaluation. The proposed framework HIDES demonstrates how future downscaled time series can be used to evaluate annual retention distribution and frequency of exceedance, while sampling extreme events allows for estimating both a probability of failure and an indication of the behaviour of GI under failure.
The thesis suggests rethinking the methods for implementing GI at the city scale. The study shows that system-based design outperforms site-scale design through modelling at the roof scale of a neighbourhood, and that lumping GI models at a neighbourhood scale may neglect interactions and fail to estimate performance. The thesis highlights the need to couple GI to achieve challenges in stormwater management.