Neural-Network-based Kalman Filters for the Spatio-Temporal Interpolation of Satellite-derived Sea Surface Temperature

Said Ouala; Ronan Fablet; Cédric Herzet; Bertrand Chapron; Ananda Pascual; Fabrice Collard; Lucile Gaultier

Neural-Network-based Kalman Filters for the Spatio-Temporal Interpolation of Satellite-derived Sea Surface Temperature

Said Ouala ^{1, 2} Ronan Fablet ^{1, 2} Cédric Herzet ^{1, 2, 3} Bertrand Chapron ⁴ Ananda Pascual ⁵ Fabrice Collard ⁶ Lucile Gaultier ⁶

1 Lab-STICC - Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance

2 IMT Atlantique - SC - Département Signal et Communications

3 SIMSMART - SIMulation pARTiculaire de Modèles Stochastiques

IRMAR - Institut de Recherche Mathématique de Rennes, Inria Rennes – Bretagne Atlantique

4 LOS - Laboratoire d’Oéanographie Spatiale [Plouzané]

5 IMEDEA - Institut Mediterrani d'Estudis Avancats

6 OceanDataLab

Abstract : In this work we address the reconstruction of gap-free Sea Surface Temperature (SST) fields 1 from irregularly-sampled satellite-derived observations. We develop novel Neural-Network-based 2 (NN-based) Kalman filters for spatio-temporal interpolation issues as an alternative to ensemble 3 Kalman filters (EnKF). The key features of the proposed approach are twofold: the learning of 4 a probabilistic NN-based representation of 2D geophysical dynamics, the associated parametric 5 Kalman-like filtering scheme for a computationally-efficient spatio-temporal interpolation of Sea 6 Surface Temperature (SST) fields. We illustrate the relevance of our contribution for an OSSE 7 (Observing System Simulation Experiment) in a case-study region off South Africa. Our numerical 8 experiments report significant improvements in terms of reconstruction performance compared with 9 operational and state-of-the-art schemes (e.g., optimal interpolation, Empirical Orthogonal Function 10 (EOF) based interpolation and analog data assimilation).

Keywords : Data-driven Neural networks Kalman filter Data assimilation Dynamical model

Document type :

Journal articles

Domain :

Environmental Sciences

Mathematics [math] / Statistics [math.ST]

Mathematics [math] / Dynamical Systems [math.DS]

Complete list of metadatas

Cited literature [3 references] Display Hide Download

https://hal-imt-atlantique.archives-ouvertes.fr/hal-01896654
Contributor : Said Ouala <>
Submitted on : Tuesday, October 16, 2018 - 12:31:27 PM
Last modification on : Tuesday, February 26, 2019 - 11:38:05 AM

S.OUALA_Paper.pdf

Files produced by the author(s)

Neural-Network-based Kalman Filters for the Spatio-Temporal Interpolation of Satellite-derived Sea Surface Temperature

File

Identifiers

Collections

Citation

Export

Metrics

518

186

Contact

Neural-Network-based Kalman Filters for the Spatio-Temporal Interpolation of Satellite-derived Sea Surface Temperature

File

Identifiers

Collections

Citation

Export

Share

Metrics

518

186

Contact