A5043426282- Oceanographic and Atmospheric Processes
- Marine and coastal ecosystems
- Ocean Waves and Remote Sensing
- Marine and fisheries research
- Tropical and Extratropical Cyclones Research
- Fluid Dynamics and Turbulent Flows
- Meteorological Phenomena and Simulations
- Solar and Space Plasma Dynamics
- Methane Hydrates and Related Phenomena
- Complex Systems and Time Series Analysis
- Time Series Analysis and Forecasting
- Aquatic and Environmental Studies
- Diffusion and Search Dynamics
- Experimental and Theoretical Physics Studies
- Underwater Acoustics Research
- Quantum, superfluid, helium dynamics
- Hydrological Forecasting Using AI
- Fluid Dynamics and Heat Transfer
- Quantum chaos and dynamical systems
- Global Trade and Competitiveness
- Coastal and Marine Dynamics
- Marine Bivalve and Aquaculture Studies
- Remote-Sensing Image Classification
- Oil Spill Detection and Mitigation
- Stock Market Forecasting Methods
Mercator Ocean (France)
2024
Centre National de la Recherche Scientifique
2021-2023
Institut polytechnique de Grenoble
2021-2023
Université Grenoble Alpes
2021-2023
Institut des Géosciences de l'Environnement
2023
Institut de Recherche pour le Développement
2023
Centre Inria de l'Université Grenoble Alpes
2021
NOAA Oceanic and Atmospheric Research
2020-2021
Laboratoire Jean Kuntzmann
2021
Centre de Recherche Inria Bordeaux - Sud-Ouest
2019-2020
Accurate ocean forecasting is essential for a range of critical applications, from maritime safety to climate adaptation strategies. Given the inherent uncertainties in dynamics, ability predict probable states key informed decision-making. Here, we present MerCast, probabilistic model designed redefine global-scale prediction by quantifying uncertainty state estimates. Trained on decades high-resolution reanalysis products, MerCast integrates diffusion models generate ensembles daily...
The European Commission launched the Digital Twin of Ocean (EDITO) at One Summit in Brest, France, February 2022. EU is building infrastructure backbone EDITO through two projects (EDITO-Model Lab and EDITO-Infra) finishing respectively beginning end 2025 with a continuity until 2028 EDITO2. This aligned Copernicus Marine Service where strong connection will be managed during new starting phase (2025-2028). presentation focus on main achievements demonstration global ocean model component...
The increasing adoption of AI-based approaches in Earth system sciences has led to breakthroughs modeling and forecasting, exemplified by state-of-the-art performance neural weather forecasting systems [Bi et al., 2023, Lam 2022]. In oceanography, Deep Learning techniques show significant promise for advancing ocean state combining both modeled observational datasets [Febvre Martin Wang 2024]. However, the context deployment faces challenges such as sparse data uncertainties existing...
Advanced ocean forecasting systems play an essential role in understanding and managing the ever-evolving dynamics, especially face of climate environmental challenges. These provide crucial information for maritime safety, marine resource management, ecosystem protection policies.This presentation will explore innovations forecasting, performance improvements, new perspectives simulation monitoring oceanographic biogeochemical processes.In framework Copernicus Marine Service, Mercator Ocean...
Analysis and study of coastal upwelling using sea surface temperature (SST) satellite images is a common procedure because its coast effectiveness (economic, time, frequency, manpower). Developing on the Ekman theory, we propose robust method to identify regions along north-west African margin. The proposed comes overcome issues encountered in recent devoted for same purpose system. Afterward, show how our can serve as framework monitor spatio-temporal variability phenomenon studied region.
Being a component of the Eastern Boundary Upwelling (EBU) ecosystem, Morocco’s Atlantic coast presents high biological production throughout year, with seasonal variations in upwelling dynamics. This characterization reflects inherent nature EBU’s ecosystems. In this work, we develop novel methodology to compute new index based on analysis sea surface temperature (SST) images. Our is not only simple calculate but also efficient. Indeed, it limited region, which has allowed improvement...
The region along the North-West African coast (20°N to 36°N and 4°W 19°W) is characterized by a persistent variable upwelling phenomenon almost all year round. In this article, features are investigated using an algorithm dedicated delimit area from thermal biological satellite observations. This method has been developed specifically for sea-surface temperature (SST) images, since they present high latitudinal variation, which not in chlorophyll-a concentration images. Developing on...
The Atlantic coast of Morocco, being part the Easter Boundaries Upwelling Ecosystem, is characterized by high biological productivity and seasonally variable upwelling all year around. In this work, we develop new deep learning tools to monitor Moroccan from physical satellite images. proposed method consists a convolutional neural network (CNN) based on an encoder-decoder built U-Net structure, localize regions. Furthermore, provide indices analysis sea surface temperature (SST)...
Near-shore water along the North-West African margin is one of world’s major upwelling regions. It associated with physical structures oceanic fronts which influence biological productivity. The study these coherent in connection chlorophyll concentration data fundamental importance for understanding spatial distributions plankton. In this work, we horizontal stirring and mixing different areas using Lagrangian (LCSs). These LCSs are calculated recent geodesic theory LCSs. We use to link...
We study the transport properties of mesoscale eddies (i.e., vortices 100–200 km in diameter) over a finite time duration. While these oceanic structures are well-known to stir and mix surrounding water, they can also carry water coherent manner. In this paper, we interested dynamic structures, despite their chaotic environment. Here, reveal that such be identified based simple decomposition Lagrangian trajectories. identify extract as material lines along which particles’ trajectories share...
In this paper we aim to present a new methodology derive rigorous SST-based coastal upwelling index for the purpose of conducting saisonal variablity area along Moroccan Atlantic coast. The method is based on scientific knowledge and its spatial distribution provided by expert oceanographers. latter consists in automatically identify extract region covered waters costal ocean Morocco using Fuzzy c-means algorithm finding regions homogeneous pixels. Then Region Growing process used filter out...
The aim of this work is to automatically identify and extract the upwelling area in coastal ocean Morocco using satellite observation chlorophyll concentration. algorithm starts by application FCM for purpose finding regions homogeneous concentration chlorophyll, resulting c-partitioned labeled images. A region-growing then used filter out noisy structures offshore waters not belonging regions. proposed methodology has been validated an oceanographer tested over a database 166 weekly Sea...
This research deals with the problem of identifying and extracting effectively main Moroccan upwelling front. The proposed methodology, based on image-fusion concept, comes to benefit from information available in both sea-surface temperature (SST) chlorophyll-a satellite images. Moreover, a new validation index is by computing simple gradient along extracted limit. developed procedure applied over database 366 SST images 2007 2014, covering Atlantic coast. final results are validated...
Automatic identification and tracking of mesoscale eddies are crucial in large oceanic observational numerical model data. This work proposes a fully automated method that identifies tracks from single Lagrangian advection, all eddies, without prior knowledge their lifespans. The eddies' detection use hybrid based on geometrical properties the evolving velocity along trajectories grid density-based clustering algorithm. high monitoring capacity proposed is demonstrated by automatically...
We study the transport properties of coherent vortices over a finite-time duration. Here, we reveal that such can be identified based on frequency-domain representation Lagrangian trajectories. use Fourier analysis to convert particles' trajectories from their time domain presentation in frequency domain. then identify and extract as material surfaces along which share similar frequencies. Our method identifies all an automatic manner, showing high vortices' monitoring capacity. illustrate...
Mesoscale eddies significantly influence ocean circulation, nutrient distribution, and climate patterns globally.  A thorough reconstruction of the eddy field is therefore important, yet classical detection algorithms based on sea level anomaly (SLA) suffer from low coverage current altimetry network. In this work, we evaluate efficacy deep learning techniques in enhancing oceanic an operational forecasting system. We use two models Observing System Simulation Experiments (OSSE): a...
Accurate ocean forecasting is crucial in different areas ranging from science to decision making. Recent advancements data-driven models have shown significant promise, particularly weather community, but yet no approaches matched the accuracy and scalability of traditional global systems that rely on physics-driven numerical can be very computationally expensive, depending their spatial resolution or complexity. Here, we introduce GLONET, a neural network-based system, developed by Mercator...
We present significant improvements to our previous work on noise reduction in {\sl Herschel} observation maps by defining sparse filtering tools capable of handling, a unified formalism, significantly improved as well deconvolution order reduce effects introduced the limited instrumental response (beam). implement greater flexibility allowing wider choice parsimonious priors noise-reduction process. More precisely, we introduce and approach type $l^2$-$l^p$, with $p > 0$ variable apply it...