A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements

An efficient combination of remotely-sensed data and in situ measurements is needed to obtain accurate 3D ocean state estimates, representing a fundamental step to describe ocean dynamics and its role in the Earth climate system and marine ecosystems. Observations can either be assimilated in ocean...

Full description

Bibliographic Details
Published in:Remote Sensing
Main Author: Bruno Buongiorno Nardelli
Format: Article in Journal/Newspaper
Language:English
Published: MDPI AG 2020
Subjects:
artificial intelligence
machine learning
deep learning
neural networks
Earth observations
ocean dynamics
Science
Q
North Atlantic
Online Access:https://doi.org/10.3390/rs12193151
https://doaj.org/article/07bad88d2a674907b5cd979db4286734
id ftdoajarticles:oai:doaj.org/article:07bad88d2a674907b5cd979db4286734
record_format openpolar
spelling ftdoajarticles:oai:doaj.org/article:07bad88d2a674907b5cd979db4286734 2023-05-15T17:32:48+02:00 A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements Bruno Buongiorno Nardelli 2020-09-01T00:00:00Z https://doi.org/10.3390/rs12193151 https://doaj.org/article/07bad88d2a674907b5cd979db4286734 EN eng MDPI AG https://www.mdpi.com/2072-4292/12/19/3151 https://doaj.org/toc/2072-4292 doi:10.3390/rs12193151 2072-4292 https://doaj.org/article/07bad88d2a674907b5cd979db4286734 Remote Sensing, Vol 12, Iss 3151, p 3151 (2020) artificial intelligence machine learning deep learning neural networks Earth observations ocean dynamics Science Q article 2020 ftdoajarticles https://doi.org/10.3390/rs12193151 2022-12-31T15:16:48Z An efficient combination of remotely-sensed data and in situ measurements is needed to obtain accurate 3D ocean state estimates, representing a fundamental step to describe ocean dynamics and its role in the Earth climate system and marine ecosystems. Observations can either be assimilated in ocean general circulation models or used to feed data-driven reconstructions and diagnostic models. Here we describe an innovative deep learning algorithm that projects sea surface satellite data at depth after training with sparse co-located in situ vertical profiles. The technique is based on a stacked Long Short-Term Memory neural network, coupled to a Monte-Carlo dropout approach, and is applied here to the measurements collected between 2010 and 2018 over the North Atlantic Ocean. The model provides hydrographic vertical profiles and associated uncertainties from corresponding remotely sensed surface estimates, outperforming similar reconstructions from simpler statistical algorithms and feed-forward networks. Article in Journal/Newspaper North Atlantic Directory of Open Access Journals: DOAJ Articles Remote Sensing 12 19 3151
institution Open Polar
collection Directory of Open Access Journals: DOAJ Articles
op_collection_id ftdoajarticles
language English
topic artificial intelligence
machine learning
deep learning
neural networks
Earth observations
ocean dynamics
Science
Q
spellingShingle artificial intelligence
machine learning
deep learning
neural networks
Earth observations
ocean dynamics
Science
Q
Bruno Buongiorno Nardelli
A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
topic_facet artificial intelligence
machine learning
deep learning
neural networks
Earth observations
ocean dynamics
Science
Q
description An efficient combination of remotely-sensed data and in situ measurements is needed to obtain accurate 3D ocean state estimates, representing a fundamental step to describe ocean dynamics and its role in the Earth climate system and marine ecosystems. Observations can either be assimilated in ocean general circulation models or used to feed data-driven reconstructions and diagnostic models. Here we describe an innovative deep learning algorithm that projects sea surface satellite data at depth after training with sparse co-located in situ vertical profiles. The technique is based on a stacked Long Short-Term Memory neural network, coupled to a Monte-Carlo dropout approach, and is applied here to the measurements collected between 2010 and 2018 over the North Atlantic Ocean. The model provides hydrographic vertical profiles and associated uncertainties from corresponding remotely sensed surface estimates, outperforming similar reconstructions from simpler statistical algorithms and feed-forward networks.
format Article in Journal/Newspaper
author Bruno Buongiorno Nardelli
author_facet Bruno Buongiorno Nardelli
author_sort Bruno Buongiorno Nardelli
title A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
title_short A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
title_full A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
title_fullStr A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
title_full_unstemmed A Deep Learning Network to Retrieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements
title_sort deep learning network to retrieve ocean hydrographic profiles from combined satellite and in situ measurements
publisher MDPI AG
publishDate 2020
url https://doi.org/10.3390/rs12193151
https://doaj.org/article/07bad88d2a674907b5cd979db4286734
genre North Atlantic
genre_facet North Atlantic
op_source Remote Sensing, Vol 12, Iss 3151, p 3151 (2020)
op_relation https://www.mdpi.com/2072-4292/12/19/3151
https://doaj.org/toc/2072-4292
doi:10.3390/rs12193151
2072-4292
https://doaj.org/article/07bad88d2a674907b5cd979db4286734
op_doi https://doi.org/10.3390/rs12193151
container_title Remote Sensing
container_volume 12
container_issue 19
container_start_page 3151
_version_ 1766131088697065472

Similar Items