Skip to Main content Skip to Navigation
Reports

Sea Surface Flow Estimation via Ensemble-based Variational Data Assimilation*

Shengze Cai 1, 2 Etienne Mémin 1 Yin Yang 1 Chao Xu 2
1 FLUMINANCE - Fluid Flow Analysis, Description and Control from Image Sequences
IRMAR - Institut de Recherche Mathématique de Rennes, IRSTEA - Institut national de recherche en sciences et technologies pour l'environnement et l'agriculture, Inria Rennes – Bretagne Atlantique
Abstract : In this paper, we propose a data assimilation method for consistently estimating the velocity fields from a whole image sequence depicting the evolution of sea surface temperature transported by oceanic surface flow. The estima-tor is conducted through an ensemble-based variational data assimilation, which is designed by combining the advantages of two approaches: the ensemble Kalman filter and the variational data assimilation. This idea allows us to obtain the optimal initial condition as well as the full system trajectory. In order to extract the velocity fields from fluid images, a surface quasi-geostrophic model representing the generic evolution of the temperature field of the flow, and the optical flow constraint equation derived from the image intensity constancy assumption, are involved in the assimilation context. Numerical experimental evaluation is presented on a synthetic fluid image sequence. The results indicate good performance and efficiency of the proposed estimator.
Complete list of metadata

Cited literature [23 references]  Display  Hide  Download

https://hal.inria.fr/hal-01589637
Contributor : Etienne Memin <>
Submitted on : Monday, September 18, 2017 - 6:01:23 PM
Last modification on : Thursday, January 7, 2021 - 4:12:46 PM

File

EnVar_SST_Cai_20170918.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01589637, version 1

Citation

Shengze Cai, Etienne Mémin, Yin Yang, Chao Xu. Sea Surface Flow Estimation via Ensemble-based Variational Data Assimilation*. [Research Report] Inria Rennes - Bretagne Atlantique; IRMAR, University of Rennes 1. 2017. ⟨hal-01589637⟩

Share

Metrics

Record views

1112

Files downloads

340