Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage
Mesoscale oceanic eddies have a visible signature on sea surface temperature (SST) satellite images, portraying diverse patterns of coherent vortices, temperature gradients, and swirling filaments. However, learning the regularities of such signatures defines a challenging pattern recognition task,...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2020-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9115298/ |
id |
doaj-199cf8c9a2914a329b68bcced4bd6581 |
---|---|
record_format |
Article |
spelling |
doaj-199cf8c9a2914a329b68bcced4bd65812021-06-03T23:01:29ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352020-01-01133437344710.1109/JSTARS.2020.30018309115298Classification of Eddy Sea Surface Temperature Signatures Under Cloud CoverageEvangelos Moschos0https://orcid.org/0000-0003-1134-3220Alexandre Stegner1https://orcid.org/0000-0001-7882-493XOlivier Schwander2https://orcid.org/0000-0003-0266-1585Patrick Gallinari3https://orcid.org/0000-0001-9060-9001Institut Polytechnique de Paris, ENS, PSL Université, LMD/IPSL, École Polytechnique, Sorbonne Université, CNRS, Palaiseau, FranceInstitut Polytechnique de Paris, ENS, PSL Université, LMD/IPSL, École Polytechnique, Sorbonne Université, CNRS, Palaiseau, FranceLIP6, Sorbonne Université, Paris, FranceLIP6, Sorbonne Université, Paris, FranceMesoscale oceanic eddies have a visible signature on sea surface temperature (SST) satellite images, portraying diverse patterns of coherent vortices, temperature gradients, and swirling filaments. However, learning the regularities of such signatures defines a challenging pattern recognition task, due to their complex structure but also to the cloud coverage which can corrupt a large fraction of the image. We introduce a novel deep learning approach to classify sea temperature eddy signatures, even if they are corrupted by strong cloud coverage. A large dataset of SST image patches is automatically retained and used to train a CNN-based classifier. Classification is performed with very high accuracy on coherent eddy signatures and is robust to a high level of cloud coverage, surpassing human expert efficiency on this task. This methodology can serve to validate and correct detections on satellite altimetry, the standard method used until now to track mesoscale eddies.https://ieeexplore.ieee.org/document/9115298/Computer visiondeep learningmesoscale eddiesoceanographypattern recognitionremote sensing |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Evangelos Moschos Alexandre Stegner Olivier Schwander Patrick Gallinari |
spellingShingle |
Evangelos Moschos Alexandre Stegner Olivier Schwander Patrick Gallinari Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Computer vision deep learning mesoscale eddies oceanography pattern recognition remote sensing |
author_facet |
Evangelos Moschos Alexandre Stegner Olivier Schwander Patrick Gallinari |
author_sort |
Evangelos Moschos |
title |
Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage |
title_short |
Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage |
title_full |
Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage |
title_fullStr |
Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage |
title_full_unstemmed |
Classification of Eddy Sea Surface Temperature Signatures Under Cloud Coverage |
title_sort |
classification of eddy sea surface temperature signatures under cloud coverage |
publisher |
IEEE |
series |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
issn |
2151-1535 |
publishDate |
2020-01-01 |
description |
Mesoscale oceanic eddies have a visible signature on sea surface temperature (SST) satellite images, portraying diverse patterns of coherent vortices, temperature gradients, and swirling filaments. However, learning the regularities of such signatures defines a challenging pattern recognition task, due to their complex structure but also to the cloud coverage which can corrupt a large fraction of the image. We introduce a novel deep learning approach to classify sea temperature eddy signatures, even if they are corrupted by strong cloud coverage. A large dataset of SST image patches is automatically retained and used to train a CNN-based classifier. Classification is performed with very high accuracy on coherent eddy signatures and is robust to a high level of cloud coverage, surpassing human expert efficiency on this task. This methodology can serve to validate and correct detections on satellite altimetry, the standard method used until now to track mesoscale eddies. |
topic |
Computer vision deep learning mesoscale eddies oceanography pattern recognition remote sensing |
url |
https://ieeexplore.ieee.org/document/9115298/ |
work_keys_str_mv |
AT evangelosmoschos classificationofeddyseasurfacetemperaturesignaturesundercloudcoverage AT alexandrestegner classificationofeddyseasurfacetemperaturesignaturesundercloudcoverage AT olivierschwander classificationofeddyseasurfacetemperaturesignaturesundercloudcoverage AT patrickgallinari classificationofeddyseasurfacetemperaturesignaturesundercloudcoverage |
_version_ |
1721398756451024896 |