Results 1  10
of
11
Data Assimilation for Convective Cells Tracking on Meteorological Image Sequences
, 2011
"... This paper focuses on the tracking and analysis of convective clouds systems from Meteosat Second Generation images. The highly deformable nature of convective clouds, the complexity of the physic processes involved, but also the partially hidden measurements available from image data make difficult ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
This paper focuses on the tracking and analysis of convective clouds systems from Meteosat Second Generation images. The highly deformable nature of convective clouds, the complexity of the physic processes involved, but also the partially hidden measurements available from image data make difficult a direct use of conventional image analysis techniques for tasks of detection, tracking and characterization. In this paper, we face these issues using variational data assimilation tools. Such techniques enable to perform the estimation of an unknown state function according to a given dynamical model and to noisy and incomplete measurements. The system state we are setting in this study for the clouds representation is composed of two nested curves corresponding to the exterior frontiers of the clouds and to the interior coldest parts (core) of the convective clouds. Since no reliable simple dynamical model exists for such phenomena at the image grid scale, the dynamics on which we are relying has been directly defined from image based motion measurements and takes into account an uncertainty modeling of the curves dynamics along time. In addition to this assimilation technique, we show in appendix how each cell of the recovered clouds system can be labeled and associated to characteristic parameters (birth or death time, mean temperature, velocity, growth, etc.) of great interest for meteorologists.
THEME Observation and Modeling for Environmental
"... IN PARTNERSHIP WITH: Institut national de recherche en sciences et technologies pour l’environnement et l’agriculture ..."
Abstract
 Add to MetaCart
(Show Context)
IN PARTNERSHIP WITH: Institut national de recherche en sciences et technologies pour l’environnement et l’agriculture
Author manuscript, published in "Geoscience and Remote Sensing Symposium (IGARSS) (2011)" ANALYSIS OF SST IMAGES BY WEIGHTED ENSEMBLE TRANSFORM KALMAN FILTER
"... This paper proposes a particle filter extension of the Ensemble Transform Kalman filter [5, 6]. This filter employs a nonlinear observation model that directly relies on SST images and allows us to directly extract the fluid velocity fields from a SST image sequence. Ensemble Transform Kalman filter ..."
Abstract
 Add to MetaCart
(Show Context)
This paper proposes a particle filter extension of the Ensemble Transform Kalman filter [5, 6]. This filter employs a nonlinear observation model that directly relies on SST images and allows us to directly extract the fluid velocity fields from a SST image sequence. Ensemble Transform Kalman filter (ETKF) is an ensemble implementation of the well known recursive Kalman filter (KF), similar to the Ensemble Kalman Filter [3], which minimizes the mean square error of the state estimates through a set of analytic equations, and contrary to the EnKF it doesn’t use perturbed observations. Data assimilation filters couple a dynamical model describing the state evolution and a measurement model that relates the state variables, vorticity in our case, to the observations. In the proposed approach we employ a stochastic filtered version of vorticityvelocity NavierStokes formulation as the dynamical model and rely directly on a nonlinear displaced frame reconstruction error of SST images for the measurement model. Vorticity maps are first sampled by combining the dynamics and the Kalman updates equations at the analysis step and then corrected through an importance sampling weighting provided by the observations likelihood, so the result of the ETKF analysis can viewed as the proposal sampling distribution of the particle filter. We further propose an uncertainty model which enables us to handle regions with missing data, while boundary conditions are imposed on coastal regions. This filter is referred as the Weighted Ensemble
SELFSIMILAR PRIOR AND WAVELET BASES FOR HIDDEN INCOMPRESSIBLE TURBULENT MOTION
, 2013
"... Abstract. This work is concerned with the illposed inverse problem of estimating turbulent flows from the observation of an image sequence. From a Bayesian perspective, a divergencefree isotropic fractional Brownian motion (fBm) is chosen as a prior model for instantaneous turbulent velocity field ..."
Abstract
 Add to MetaCart
(Show Context)
Abstract. This work is concerned with the illposed inverse problem of estimating turbulent flows from the observation of an image sequence. From a Bayesian perspective, a divergencefree isotropic fractional Brownian motion (fBm) is chosen as a prior model for instantaneous turbulent velocity fields. This selfsimilar prior characterizes accurately secondorder statistics of velocity fields in incompressible isotropic turbulence. Nevertheless, the associated maximum a posteriori involves a fractional Laplacian operator which is delicate to implement in practice. To deal with this issue, we propose to decompose the divergentfree fBm on wellchosen wavelet bases. As a first alternative, we propose to design wavelets as whitening filters. We show that these filters are fractional Laplacian wavelets composed with the Leray projector. As a second alternative, we use a divergencefree wavelet basis, which takes implicitly into account the incompressibility constraint arising from physics. Although the latter decomposition involves correlated wavelet coefficients, we are able to handle this dependence in practice. Based on these two wavelet decompositions, we finally provide effective and efficient algorithms to approach the maximum a posteriori. An intensive numerical evaluation proves the relevance of the proposed waveletbased selfsimilar priors. Key words. Bayesian estimation, fractional Brownian motion, divergencefree wavelets, fractional Laplacian, connection coefficients, fast transforms, opticflow, isotropic turbulence. AMS subject classifications. 60G18, 60G22, 60H05, 62F15, 65T50, 65T60 1. Introduction. This
SERIES A DYNAMIC METEOROLOGY AND OCEANOGRAPHY
, 2010
"... In the context of tackling the illposed inverse problem of motion estimation from image sequences, we propose to introduce prior knowledge on flow regularity given by turbulence statistical models. Prior regularity is formalised using turbulence power laws describing statistically selfsimilar stru ..."
Abstract
 Add to MetaCart
(Show Context)
In the context of tackling the illposed inverse problem of motion estimation from image sequences, we propose to introduce prior knowledge on flow regularity given by turbulence statistical models. Prior regularity is formalised using turbulence power laws describing statistically selfsimilar structure of motion increments across scales. The motion estimation method minimises the error of an image observation model while constraining secondorder structure function to behave as a power law within a prescribed range. Thanks to a Bayesian modelling framework, the motion estimation method is able to jointly infer the most likely power law directly from image data. The method is assessed on velocity fields of 2D or quasi2D flows. Estimation accuracy is first evaluated on a synthetic image sequence of homogeneous and isotropic 2D turbulence. Results obtained with the approach based on physics of fluids outperform stateoftheart. Then, the method analyses atmospheric turbulence using a real meteorological image sequence. Selecting the most likely power law model enables the recovery of physical quantities, which are of major interest for turbulence atmospheric characterisation. In particular, from meteorological images we are able to estimate energy and enstrophy fluxes of turbulent cascades, which are in agreement with previous in situ measurements.
Thèmes COG et NUM — Systèmes cognitifs et Systèmes numériques
, 2009
"... Th„emes COG et NUM apport de recherche ..."
(Show Context)
ProjectTeam FLUMINANCE Fluid flow Analysis, Description and Control from Image Sequences
"... c t i v it y e p o r t ..."
analysis Description of the project Context
"... Advanced turbulence models for the assimilation of flows from image sequences ..."
Abstract
 Add to MetaCart
(Show Context)
Advanced turbulence models for the assimilation of flows from image sequences
Sciences
"... Institut national de recherche en sciences et technologies pour l’environnement et l’agriculture ..."
Abstract
 Add to MetaCart
(Show Context)
Institut national de recherche en sciences et technologies pour l’environnement et l’agriculture