A5022478491- Meteorological Phenomena and Simulations
- Flood Risk Assessment and Management
- Climate variability and models
- Hydrology and Watershed Management Studies
- Robotic Locomotion and Control
- Robot Manipulation and Learning
- Motor Control and Adaptation
- Reservoir Engineering and Simulation Methods
- Hemispheric Asymmetry in Neuroscience
- Robotic Mechanisms and Dynamics
- Geophysics and Gravity Measurements
- Spatial Neglect and Hemispheric Dysfunction
- Musculoskeletal pain and rehabilitation
- Hydrocarbon exploration and reservoir analysis
- Oceanographic and Atmospheric Processes
- Pesticide and Herbicide Environmental Studies
- Neural and Behavioral Psychology Studies
- Modular Robots and Swarm Intelligence
- Weed Control and Herbicide Applications
- Balance, Gait, and Falls Prevention
- Hydrology and Drought Analysis
- Fluid Dynamics and Turbulent Flows
- Atmospheric chemistry and aerosols
- Robotic Path Planning Algorithms
- Smoking Behavior and Cessation
Bjerknes Centre for Climate Research
2022-2024
University of Bergen
2022-2024
Nansen Environmental and Remote Sensing Center
2024
Fondation Lenval
2022
Centre National de la Recherche Scientifique
2008-2018
Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique
2012-2018
Climat, Environnement, Couplages et Incertitudes
2016
Géosciences Environnement Toulouse
2016
City University of New York
2014
SoftBank Robotics (France)
2014
Increasing the resolution of a model can improve performance data assimilation system: first because field are in better agreement with high observations, then corrections sustained and, ensemble assimilation, forecast error covariances improved. However, increase is associated cubical computational costs. Here we testing an approach inspired from images super-resolution techniques and called "Super-resolution assimilation" (SRDA). Starting low-resolution forecast, neural network (NN)...
A computational simplification of the Kalman filter (KF) is introduced – parametric (PKF). The full covariance matrix dynamics KF, which describes evolution along analysis and forecast cycle, replaced by error variance diffusion tensor, related to correlation length-scales. PKF developed here has been applied simplified framework advection–diffusion a passive tracer, for its use in chemical transport model assimilation. easy compute computationally cost-effective than an ensemble (EnKF) this...
Abstract Ensemble data assimilation methods, such as the Kalman Filter (EnKF), are well suited for climate reanalysis because they feature flow‐dependent covariance. However, Earth System Models heavy computationally, method uses a few tens of members. Sampling error in covariance matrix can introduce biases deep ocean, which may cause drift and predictions. Here, we assess potential hybrid approach (EnKF‐OI) to counteract sampling error. The EnKF‐OI combines computed from dynamical ensemble...
In this contribution, Metapod, a novel C++ library computing efficiently dynamic algorithms is presented. It uses template-programming techniques together with code-generation. The achieved performances shows some advantage over the state-of-the art RBDL mostly on ATOM processor and for inertia matrix computation, which are relevant robotics application. On recent desktop computer, ratio of gain not so obvious in general time by both significantly different inverse dynamics. that it...
Background ROHHAD syndrome (Rapid-onset Obesity with Hypothalamic dysfunction, Hypoventilation and Autonomic Dysregulation) is rare. Rapid-onset morbid obesity usually the first recognizable sign of this syndrome, however a subset patients develop without obesity. The prevalence entity currently unknown. Alteration respiratory control as well dysautonomic disorders often have fatal outcome, thus early recognition essential. Material methods A retrospective, observational, multicenter study...
Ensemble data assimilation methods, such as the Kalman Filter (EnKF), are well suited for climate reanalysis because they feature flow-dependent covariance.However, Earth System Models heavy computationally, method uses a few tens of members.Sampling error in covariance matrix can introduce biases deep ocean, which may cause drift and predictions.Here, we assess potential hybrid approach (EnKF-OI) to counteract sampling error.The EnKF-OI combines computed from dynamical ensemble with another...
Spray drift of glyphosate has the potential to affect non-target vegetation and surface waters close application area. To assess likelihood such impact along Swedish railways, four field experiments were conducted at three railway sites during 2019 2020. An herbicide spraying train applied Roundup Ultra (glyphosate) speeds 33 48 km/h. Quantitative filter papers placed 0.5, 1, 1.5, 2, 3 5 m distances capture spray droplets. Wind low (0-2 m/s), but found be representative normal operating...
Over the last few years, a collaborative work between CERFACS, LNHE (EDF R&D), SCHAPI and CE-REMA resulted in implementation of Data Assimilation (DA) method on top MASCARET framework real-time forecasting. This prototype was based simplified Kalman filter where description background error covariances is prescribed off-line climatology constant over time. approach showed promising results Adour Marne catchments as it improves forecast skills hydraulic model using water level discharge...
This paper deals with the biomimetic design of a humanoid head prototype. prototype is developed in order to offer some mechanical device for multimodality objectives, and demonstrate importance uncoupled eyes mechanism function. Indeed, development based on our understanding humans properties filed visual vestibular capabilities. The final will have 3 DOF each eye 2 neck. as one dof neck able show ocular reflex (VOR) target tracking (TT) real time. To carry out all these capabilities,...
<title>Abstract</title> The super-resolution data assimilation (SRDA) enhances a low-resolution (LR) model with Neural Network (NN) having learned the differences between high and models offline performs in high-resolution (HR). method accuracy of EnKF-LR system for minor computational overhead. However, performance quickly saturates when increasing ensemble size due to error introduced by NN. We therefore combine SRDA mixed-resolution (MRDA), into called ''Hybrid covariance assimilation"...
Abstract The super-resolution data assimilation (SRDA) enhances a low-resolution (LR) model with Neural Network (NN) that has learned the differences between high and models offline performs in high-resolution (HR). method accuracy of EnKF-LR system for minor computational overhead. However, performance quickly saturates when ensemble size is increased due to error introduced by NN. We therefore combine SRDA mixed-resolution (MRDA) into called “Hybrid covariance assimilation” (Hybrid SRDA)....
Sea surface temperature (SST) observations are a critical data set for long-term climate reconstruction. However, their assimilation with an ensemble-based method can degrade performance in the ocean interior due to spurious covariances. Assimilation isopycnal coordinates delay degradation, but it remains problematic long reanalysis. We introduce vertical localization SST coordinate. The tapering functions formulated empirically from large pre-industrial ensemble. propose three schemes: 1)...