Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea

<p>Sea level predictions, typically derived from 3D hydrodynamic models, are computationally intensive and subject to uncertainties stemming from physical representation and inaccuracies in initial or boundary conditions. As a complementary alternative, data-driven machine learning models prov...

Full description

Saved in:
Bibliographic Details
Main Authors: A. Barzandeh, M. Ličer, M. Rus, M. Kristan, I. Maljutenko, J. Elken, P. Lagemaa, R. Uiboupin
Format: Article
Language:English
Published: Copernicus Publications 2025-07-01
Series:Ocean Science
Online Access:https://os.copernicus.org/articles/21/1315/2025/os-21-1315-2025.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849319298132082688
author A. Barzandeh
M. Ličer
M. Ličer
M. Rus
M. Rus
M. Kristan
I. Maljutenko
J. Elken
P. Lagemaa
R. Uiboupin
author_facet A. Barzandeh
M. Ličer
M. Ličer
M. Rus
M. Rus
M. Kristan
I. Maljutenko
J. Elken
P. Lagemaa
R. Uiboupin
author_sort A. Barzandeh
collection DOAJ
description <p>Sea level predictions, typically derived from 3D hydrodynamic models, are computationally intensive and subject to uncertainties stemming from physical representation and inaccuracies in initial or boundary conditions. As a complementary alternative, data-driven machine learning models provide a computationally efficient solution with comparable accuracy. This study employs the deep-learning model HIDRA2 to forecast hourly sea levels at five coastal stations along the Estonian coastline of the Baltic Sea, evaluating its performance across various forecast lead times. Compared to the regional NEMO<span class="inline-formula"><sub>BAL</sub></span> and subregional NEMO<span class="inline-formula"><sub>EST</sub></span> hydrodynamic models, HIDRA2 frequently outperforms both, particularly in terms of overall forecast skill. While HIDRA2 shows limitations in resolving high-frequency sea level variability above (6 <span class="inline-formula">h</span>)<span class="inline-formula"><sup>−1</sup></span>, it effectively reproduces energy in lower-frequency bands below (18 <span class="inline-formula">h</span>)<span class="inline-formula"><sup>−1</sup></span>. Errors tend to average out over longer time windows encompassing multiple seiche periods, enabling HIDRA2 to surpass the overall performance of the NEMO models. These findings underscore HIDRA2's potential as a robust, efficient, and reliable tool for operational sea level forecasting and coastal management in the eastern Baltic Sea region.</p>
format Article
id doaj-art-322b45a3b1ce440ca26ae01aea1bd3aa
institution Kabale University
issn 1812-0784
1812-0792
language English
publishDate 2025-07-01
publisher Copernicus Publications
record_format Article
series Ocean Science
spelling doaj-art-322b45a3b1ce440ca26ae01aea1bd3aa2025-08-20T03:50:31ZengCopernicus PublicationsOcean Science1812-07841812-07922025-07-01211315132710.5194/os-21-1315-2025Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic SeaA. Barzandeh0M. Ličer1M. Ličer2M. Rus3M. Rus4M. Kristan5I. Maljutenko6J. Elken7P. Lagemaa8R. Uiboupin9Department of Marine Systems, Tallinn University of Technology, Tallinn, EstoniaSlovenian Environment Agency, Ljubljana, SloveniaNational Institute of Biology, Marine Biology Station, Piran, SloveniaSlovenian Environment Agency, Ljubljana, SloveniaFaculty of Computer and Information Science, Visual Cognitive Systems Lab, University of Ljubljana, Ljubljana, SloveniaFaculty of Computer and Information Science, Visual Cognitive Systems Lab, University of Ljubljana, Ljubljana, SloveniaDepartment of Marine Systems, Tallinn University of Technology, Tallinn, EstoniaDepartment of Marine Systems, Tallinn University of Technology, Tallinn, EstoniaDepartment of Marine Systems, Tallinn University of Technology, Tallinn, EstoniaDepartment of Marine Systems, Tallinn University of Technology, Tallinn, Estonia<p>Sea level predictions, typically derived from 3D hydrodynamic models, are computationally intensive and subject to uncertainties stemming from physical representation and inaccuracies in initial or boundary conditions. As a complementary alternative, data-driven machine learning models provide a computationally efficient solution with comparable accuracy. This study employs the deep-learning model HIDRA2 to forecast hourly sea levels at five coastal stations along the Estonian coastline of the Baltic Sea, evaluating its performance across various forecast lead times. Compared to the regional NEMO<span class="inline-formula"><sub>BAL</sub></span> and subregional NEMO<span class="inline-formula"><sub>EST</sub></span> hydrodynamic models, HIDRA2 frequently outperforms both, particularly in terms of overall forecast skill. While HIDRA2 shows limitations in resolving high-frequency sea level variability above (6 <span class="inline-formula">h</span>)<span class="inline-formula"><sup>−1</sup></span>, it effectively reproduces energy in lower-frequency bands below (18 <span class="inline-formula">h</span>)<span class="inline-formula"><sup>−1</sup></span>. Errors tend to average out over longer time windows encompassing multiple seiche periods, enabling HIDRA2 to surpass the overall performance of the NEMO models. These findings underscore HIDRA2's potential as a robust, efficient, and reliable tool for operational sea level forecasting and coastal management in the eastern Baltic Sea region.</p>https://os.copernicus.org/articles/21/1315/2025/os-21-1315-2025.pdf
spellingShingle A. Barzandeh
M. Ličer
M. Ličer
M. Rus
M. Rus
M. Kristan
I. Maljutenko
J. Elken
P. Lagemaa
R. Uiboupin
Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
Ocean Science
title Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
title_full Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
title_fullStr Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
title_full_unstemmed Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
title_short Application of the HIDRA2 deep-learning model for sea level forecasting along the Estonian coast of the Baltic Sea
title_sort application of the hidra2 deep learning model for sea level forecasting along the estonian coast of the baltic sea
url https://os.copernicus.org/articles/21/1315/2025/os-21-1315-2025.pdf
work_keys_str_mv AT abarzandeh applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT mlicer applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT mlicer applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT mrus applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT mrus applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT mkristan applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT imaljutenko applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT jelken applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT plagemaa applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea
AT ruiboupin applicationofthehidra2deeplearningmodelforsealevelforecastingalongtheestoniancoastofthebalticsea