A5028955976- Oceanographic and Atmospheric Processes
- Air Quality and Health Impacts
- Soil Geostatistics and Mapping
- Meteorological Phenomena and Simulations
- Air Quality Monitoring and Forecasting
- Reservoir Engineering and Simulation Methods
- Climate variability and models
- Atmospheric chemistry and aerosols
- Ocean Waves and Remote Sensing
- Geophysics and Gravity Measurements
- Vehicle emissions and performance
- Atmospheric and Environmental Gas Dynamics
- Gaussian Processes and Bayesian Inference
- Seismic Imaging and Inversion Techniques
- Model Reduction and Neural Networks
- Remote Sensing in Agriculture
- Underwater Acoustics Research
- Health, Environment, Cognitive Aging
- Toxic Organic Pollutants Impact
- Environmental Justice and Health Disparities
- Neural Networks and Applications
- Geological Modeling and Analysis
- Wind and Air Flow Studies
- Spatial and Panel Data Analysis
- AI and HR Technologies
IMT Atlantique
2020-2024
Université de Bretagne Occidentale
2020-2024
Joint Research Centre
2024
Laboratoire des Sciences et Techniques de l’Information de la Communication et de la Connaissance
2020-2024
Danish Meteorological Institute
2024
Institut national de l'environnement industriel et des risques
2012-2022
Centre National de la Recherche Scientifique
2020-2021
Signal Processing (United States)
2021
Centre de Recherche en Sciences et Technologies de l'Information et de la Communication
2020
École Nationale Supérieure des Mines de Paris
2017-2018
Abstract. A modified version of CHIMERE 2009, including new methodologies in emissions modelling and an urban correction, is used to perform a simulation at high resolution (0.125° × 0.0625°) over Europe for the year 2009. The model reproduces temporal variability NO2, O3, PM10, PM2.5 better rural (RB) than (UB) background stations, with yearly correlation values different pollutants ranging between 0.62 0.77 RB sites 0.52 0.73 UB sites. Also, fractional biases (FBs) show that performs...
Abstract. The reconstruction of sea surface currents from satellite altimeter data is a key challenge in spatial oceanography, especially with the upcoming wide-swath SWOT (Surface Water and Ocean Topography) mission. Operational systems, however, generally fail to retrieve mesoscale dynamics for horizontal scales below 100 km timescales 10 d. Here, we address this through 4DVarnet framework, an end-to-end neural scheme backed on variational assimilation formulation. We introduce...
Over the last few years, a very active field of research has aimed at exploring new data-driven and learning-based methodologies to propose computationally efficient strategies able benefit from large amount observational remote sensing numerical simulations for reconstruction, interpolation prediction high-resolution derived products geophysical fields. In this paper, we investigate how they might help solve oversmoothing state-of-the-art optimal (OI) techniques in reconstruction sea...
Abstract. This paper addresses physics-informed deep learning schemes for satellite ocean remote sensing data. Such observation datasets are characterized by the irregular space-time sampling of surface due to sensors’ characteristics and orbits. With a focus on altimetry, we show that end-to-end based variational formulations provide new means explore exploit such datasets. Through Observing System Simulation Experiments (OSSE) using numerical simulations real nadir wide-swath altimeter...
As part of the OptimESM project, this work aims to prototype a framework for downscaling post-CMIP6 Earth System Models (ESMs) refine long-term projections up 2300. This effort focuses on understanding regional climate impacts and extreme events, including heatwaves, droughts, precipitation extremes, with goal supporting robust informing adaptation strategies across Europe. Within broader context, our study investigates application deep learning techniques downscale daily temperature fields,...
Earth observation satellite missions provide invaluable global observations of geophysical processes in play the atmosphere and oceans. Due to sensor technologies (e.g., infrared sensors), atmospheric conditions clouds heavy rains), orbits polar-orbiting satellites), satellite-derived often involve irregular space–time sampling patterns large missing data rates. Given current development learning-based schemes for earth observation, question naturally arises whether one might learn some...
Reducing health inequalities involves the identification and characterization of social exposure factors way they accumulate in a given area. The areas accumulation then allow for prioritization interventions. present study aims to build spatial composite indicators based on aggregation environmental, their inter-relationships.Preliminary work was carried out firstly homogenize coverage, secondly variation environmental (EI), socioeconomic (SI) (HI) indicators. different performed using...
Abstract Air quality modeling tools are largely used to assess air pollution mitigation and monitoring strategies. While neural networks (NN) were mostly developed based on observations derive statistical models at stations, the use of Eulerian chemistry transport (CTMs) was mainly devoted predictions over large areas evaluation emission reduction In this study, we investigate deep learning architectures create a metamodel process oriented CTM CHIMERE significantly reduce computing times...
The reconstruction of better-resolved sea surface currents is a key challenge in space oceanography. Besides the upcoming SWOT wide-swath altimeter mission, new algorithms are explore to produce improved gap-free gridded products. Based on recent development generic end-to-end deep learning scheme for inverse problems backed variational formulation, we investigate how this framework applies space-time interpolation satellite-derived SSH fields. We consider different parameterization proposed...
Reducing environmental health inequalities has become a major focus of public efforts in France, as evidenced by the French action plans for and environment. To evaluate inequalities, routine monitoring networks provide valuable source data on contamination, which can be used integrated assessments, to identify overexposed populations prioritize actions. However, available databases generally do not meet sufficient spatial representativeness characterize population exposure, they are usually...
Abstract Background At a regional or continental scale, the characterization of environmental health inequities (EHI) expresses idea that populations are not equal in face pollution. It implies an analysis be conducted order to identify and manage areas at risk overexposure where increasing human is suspected. The development methods prerequisite for implementing public activities aimed protecting populations. Methods This paper presents methodological framework developed by INERIS (French...
The spatio-temporal interpolation of large geophysical datasets has historically been adressed by Optimal Interpolation (OI) and more sophisticated model-based or data-driven DA techniques. In the last ten years, link established between Stochastic Partial Differential Equations (SPDE) Gaussian Markov Random Fields (GMRF) opened a new way handling both physically-induced covariance matrix in Interpolation. Recent advances deep learning community also enables to adress this problem as neural...
The SLICING (Sea Level Innovations and Collaborative Intercomparison for the Next-Generation products) project is a Copernicus Service Evolution that responds to evolving landscape of sea level processing. promotes novel approach centered on open collaborative data challenges altimetric product developments assessments. With focus fostering collaboration methodological advancement, aligns with objectives Marine (CMEMS) overarching spirit European Digital Twin Ocean (DTO).   growing...
(1) Background: Lower socioeconomic status increases psychiatric service use, exacerbated during the COVID-19 pandemic by environmental stressors like air pollution and limited green spaces. This study aims to assess influence of sociodemographic factors on mental health utilisation. (2) Methods: retrospective uses an administrative database focusing community services in Northeast Italy. Spatial temporal analyses were used address space–time dependencies. (3) Results: Findings showed that...