Toward the integration of ensemble marine flooding forecasts for decision-making under uncertainty : a pathway through production, translation and visualization challenges – ORACLES

ORACLES will address the production, translation and visualization challenges related to probabilistic forecasts of marine flooding and induced consequences on territory. It aims at: (WP1) better defining users needs and expectations regarding marine flooding probabilistic forecasts and describing user-specific relevant indicators (WP2) developing methods to produce spatialized flood forecasts from Météo-France Ensemble Prediction System using promising metamodeling techniques and characterizing the cascading sources of uncertainty throughout the entire modeling chain (WP3) designing crisis managers and expert-oriented innovative geovisualization frameworks to better explore, analyze and map probabilistic marine flooding forecasts and their related impacts on the territory.

The project will be handled on two complementary French pilot sites (Arcachon lagoon in Gironde and the city of Gâvres in Morbihan) covering different types of territories and flood processes. In addition to the multidisciplinary project team, it will also rely upon users groups covering a wide range of national and local actors involved in marine flooding crisis management to guide and provide feedbacks on the developments : operational forecasters, territorial marine flooding experts, technical networks managers, emergency services and decision-makers of state services and collectivities.

Based on advanced technologies and transversality, ORACLES will contribute to improve the state of the art on marine flooding forecast production and communication while responding to societal needs by proactively involving stakeholders. If the project results will be site-dependent, the developed methods and tools will be generic. Results will be delivered as coherent technological building blocks to facilitate their dissemination and keep open the possibilities of application and integration into existing operational tools. Beside scientific publications, the project outputs will be deposited on open source repositories and recommandations will be provided for their transposition on other sites.

1. Rohmer, J., Sire, C., Lecacheux, S. et al. Improved metamodels for predicting high-dimensional outputs by accounting for the dependence structure of the latent variables: application to marine flooding. Stoch Environ Res Risk Assess (2023). doi.org/10.1007/s00477-023-02426-z
2. Idier et al., 2023 : Coastal flood at Gâvres : a simulated dataset to support risk management and metamodels development, Journal of Marine Science and Engineering (accepted).

The high rates of population growth and urbanization in coastal regions tend to constantly increase marine flooding risks in low-lying areas. The variety of processes and scales to consider make it very difficult to predict several hours or days in advance. If marine flooding early warning systems have gained a significant impulse in the last decades, they still mostly follow a deterministic approach or rely on deep simplifications. Today, there is a growing demand not only to shift towards ever more precise and local marine flooding forecasts, but also to better account for uncertainties and provide information on possible consequences on the territory stakes. Yet, the pathway to provide relevant information for a risk-based decision-making runs up against a series of issues (computation time, cascading uncertainty and communication to stakeholders), thus calling for an interdisciplinary thinking.
ORACLES will address the production, translation and visualization challenges related to probabilistic forecasts of marine flooding and induced consequences on territory. It aims at: (1) better defining users needs and expectations regarding marine flooding probabilistic forecasts, (2) developing methods to produce spatialized flood forecasts from Météo-France Ensemble Prediction System using promising metamodeling techniques, (3) characterizing the cascading sources of uncertainty throughout the entire modeling chain from meteorological scenarios to marine flooding assessment, (4) describing user-specific relevant indicators and possible domino-effects of consequences on critical networks resulting from marine flooding, (5) designing crisis managers and expert-oriented innovative geovisualization frameworks to better explore, analyze and map probabilistic marine flooding forecasts and their related impacts on the territory.
The project will be handled on two complementary French pilot sites (Arcachon lagoon in Gironde and the city of Gâvres in Morbihan) covering different types of territories and flood processes. In addition to the multidisciplinary project team, it will also rely upon users groups covering a wide range of national and local actors involved in marine flooding crisis management to guide and provide feedbacks on the developments : operational forecasters, territorial marine flooding experts, technical networks managers, emergency services and decision-makers of state services and collectivities.
Based on advanced technologies and transversality, ORACLES will contribute to improve the state of the art on marine flooding forecast production and communication while responding to societal needs by proactively involving stakeholders. If the project results will be site-dependent, the developed methods and tools will be generic. Results will be delivered as coherent technological building blocks to facilitate their dissemination and keep open the possibilities of application and integration into existing operational tools. Beside scientific publications, the project outputs will be deposited on open source repositories and recommandations will be provided for their transposition on other sites.