Sunglint correction of the Multi-Spectral Instrument (MSI)-SENTINEL-2 imagery over inland and sea waters from SWIR bands

Abstract : Remote sensing of inland and sea waters depends on the quality of the retrieval of the water-leaving radiance from the top-of-atmosphere measurements. The water-leaving radiance can be difficult to observe due to the reflection of direct sunlight on the air-water interface (sunglint) in the direction of the satellite field of view. The viewing geometry of Sentinel-2 satellite (European Space Agency) makes it vulnerable to sunglint contamination. In this paper, an original method is proposed to correct Sentinel-2-like imagery for sunglint contamination. The sunglint contribution is first estimated from the shortwave-infrared (SWIR) part of the spectrum and then extrapolated toward the near-infrared and visible bands. The spectral variation of the sunglint signal is thus revisited for a wide spectral range (from 350 to 2500 nm). The bidirectional reflectance distribution function related to the sunglint is shown to vary by > 28% from the SWIR to the blue bands of Sentinel-2. The application of the proposed algorithm on actual Sentinel-2 data demonstrates that sunglint patterns are satisfactorily removed over the entire images whatever the altitude of the observed target. Comparison with in situ data of water-leaving radiances (AERONET-OC) showed that our proposed algorithm significantly improves the correlation between satellite and in situ data by 55% (i.e., from R2 = 0.56 to R2 = 0.87). In addition, the discrepancies between satellite and in situ measurements are reduced by 60%. It is also shown that the aerosol data provided by the Copernicus Atmosphere Monitoring Service (CAMS) can be safely used within the proposed algorithm to correct the Sentinel-2-like satellite data for both sunglint and atmospheric radiances. Improvements of the proposed method potentially rely on simultaneous retrievals of the aerosol optical properties. The proposed method is applicable to any satellite sensor which is able to measure in SWIR spectral bands over aquatic environments.
Type de document :
Article dans une revue
Remote Sensing of Environment, Elsevier, 2018, 204, pp.308-321. 〈10.1016/j.rse.2017.10.022〉
Liste complète des métadonnées

Littérature citée [67 références]  Voir  Masquer  Télécharger

https://hal-insu.archives-ouvertes.fr/insu-01628728
Contributeur : Catherine Cardon <>
Soumis le : vendredi 5 janvier 2018 - 10:18:37
Dernière modification le : jeudi 11 octobre 2018 - 08:25:16

Fichier

Harmel_etal_2018_sunglint_sans...
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Tristan Harmel, Malik Chami, Thierry Tormos, Nathalie Reynaud, Pierre-Alain Danis. Sunglint correction of the Multi-Spectral Instrument (MSI)-SENTINEL-2 imagery over inland and sea waters from SWIR bands. Remote Sensing of Environment, Elsevier, 2018, 204, pp.308-321. 〈10.1016/j.rse.2017.10.022〉. 〈insu-01628728〉

Partager

Métriques

Consultations de la notice

268

Téléchargements de fichiers

43