Significant wave heights from Sentinel-1 SAR: Validation and applications - INSU - Institut national des sciences de l'Univers Accéder directement au contenu
Article Dans Une Revue Journal of Geophysical Research. Oceans Année : 2017

Significant wave heights from Sentinel-1 SAR: Validation and applications

J. E. Stopa
  • Fonction : Auteur
A. Mouche
  • Fonction : Auteur

Résumé

Two empirical algorithms are developed for wave mode images measured from the synthetic aperture radar aboard Sentinel-1 A. The first method, called CWAVE_S1A, is an extension of previous efforts developed for ERS2 and the second method, called F, uses the azimuth cutoff among other parameters to estimate significant wave heights and average wave periods without using a modulation transfer function. Neural networks are trained using colocated data generated from WAVEWATCH III and independently verified with data from altimeters and in situ buoys. We use neural networks to relate the nonlinear relationships between the input SAR image parameters and output geophysical wave parameters. CWAVE_S1A performs well and has reduced precision compared to with root mean square errors within 0.5 and 0.6 m, respectively. The developed neural networks extend the SAR's ability to retrieve useful wave information under a large range of environmental conditions including extratropical and tropical cyclones in which estimation is traditionally challenging. Plain Language SummaryTwo empirical algorithms are developed to estimate integral wave parameters from high resolution synthetic aperture radar (SAR) ocean images measured from recently launched the Sentinel 1 satellite. These methods avoid the use of the complicated image to wave mapping typically used to estimate sea state parameters. In addition, we are able to estimate wave parameters that are not able to be measured using existing techniques for the Sentinel 1 satellite. We use a machine learning technique to create a model that relates the ocean image properties to geophysical wave parameters. The models are developed using data from a numerical model because of the sufficiently large sample of global ocean conditions. We then verify that our developed models perform well with respect to independently measured wave observations from other satellite sensors and buoys. We successfully created models that estimate integrated wave parameters, like the commonly used significant wave height, accurately in a large range of sea states (up to 13 m). This allows the data from the SAR technology to be applied under a large range of environmental conditions including extra-tropical and tropical cyclones.
Fichier principal
Vignette du fichier
47415.pdf (24.52 Mo) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

insu-03682753 , version 1 (22-06-2022)

Licence

Copyright (Tous droits réservés)

Identifiants

Citer

J. E. Stopa, A. Mouche. Significant wave heights from Sentinel-1 SAR: Validation and applications. Journal of Geophysical Research. Oceans, 2017, 122, pp.1827-1848. ⟨10.1002/2016JC012364⟩. ⟨insu-03682753⟩
48 Consultations
7 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More