Simple item page
 

Daily water‐level forecasting for multiple polish lakes using multiple data‐driven models

cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid0000-0002-1453-0374
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid917b05fe-6da6-4828-82f0-08b7c58485fd
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
dc.abstract.enWater level in lakes fluctuates frequently due to the impact of natural and anthropogenic forcing. Frequent fluctuations of water level will impact lake ecosystems, and it is thus of great significance to have a good knowledge of water-level dynamics in lakes. However, forecasting daily water-level fluctuation in lake systems remains a tough task due to its non-linearity and complexity. In this study, two deep data-driven models, including gated recurrent unit (GRU) and long short-term memory (LSTM), were coupled with attention mechanism for the forecasting of daily water level in lakes for the first time. Daily water-level times series in five lowland lakes in Poland were used to evaluate the models. Root mean squared error (RMSE) and mean average error (MAE) were used for the evaluation of model performance. The modelling results were compared with the traditional feed-forward neural networks (FFNN), GRU, LSTM, and zero-order forecast. The modelling results showed that sequential deep learning models do not outperform feed-forward models in all cases. In most cases, LSTM with attention mechanism (average RMSE = 0.88 cm, average MAE = 0.69 cm) outperforms GRU with attention mechanism (average RMSE = 1.00 cm, average MAE = 0.81 cm). However, attention mechanism did not help to improve the accuracy of the GRU and LSTM for most cases. Based on the average performance in different lakes, GRU performs the best among the deep learning models (average RMSE = 0.84 cm, average MAE = 0.66 cm). Zero-order forecast models perform better than deep learning models for predicting tomorrow (average RMSE = 0.71 cm, average MAE = 0.39 cm), while deep learning models perform better as the horizon of prediction increases.
dc.affiliationWydział Inżynierii Środowiska i Inżynierii Mechanicznej
dc.affiliation.instituteKatedra Melioracji, Kształtowania Środowiska i Gospodarki Przestrzennej
dc.contributor.authorZhu, Senlin
dc.contributor.authorJi, Qingfeng
dc.contributor.authorPtak, Mariusz
dc.contributor.authorSojka, Mariusz
dc.contributor.authorKeramatfar, Abdalsamad
dc.contributor.authorChau, Kwok Wing
dc.contributor.authorBand, Shahab S.
dc.date.accessioned2025-11-04T11:40:00Z
dc.date.available2025-11-04T11:40:00Z
dc.date.issued2022
dc.description.abstract<jats:title>Abstract</jats:title><jats:p>Water level in lakes fluctuates frequently due to the impact of natural and anthropogenic forcing. Frequent fluctuations of water level will impact lake ecosystems, and it is thus of great significance to have a good knowledge of water‐level dynamics in lakes. However, forecasting daily water‐level fluctuation in lake systems remains a tough task due to its non‐linearity and complexity. In this study, two deep data‐driven models, including gated recurrent unit (GRU) and long short‐term memory (LSTM), were coupled with attention mechanism for the forecasting of daily water level in lakes for the first time. Daily water‐level times series in five lowland lakes in Poland were used to evaluate the models. Root mean squared error (RMSE) and mean average error (MAE) were used for the evaluation of model performance. The modelling results were compared with the traditional feed‐forward neural networks (FFNN), GRU, LSTM, and zero‐order forecast. The modelling results showed that sequential deep learning models do not outperform feed‐forward models in all cases. In most cases, LSTM with attention mechanism (average RMSE = 0.88 cm, average MAE = 0.69 cm) outperforms GRU with attention mechanism (average RMSE = 1.00 cm, average MAE = 0.81 cm). However, attention mechanism did not help to improve the accuracy of the GRU and LSTM for most cases. Based on the average performance in different lakes, GRU performs the best among the deep learning models (average RMSE = 0.84 cm, average MAE = 0.66 cm). Zero‐order forecast models perform better than deep learning models for predicting tomorrow (average RMSE = 0.71 cm, average MAE = 0.39 cm), while deep learning models perform better as the horizon of prediction increases.</jats:p>
dc.description.bibliographyil., bibliogr.
dc.description.financepublication_nocost
dc.description.financecost0,00
dc.description.if3,0
dc.description.number2
dc.description.points100
dc.description.volume189
dc.identifier.doi10.1111/geoj.12488
dc.identifier.eissn1475-4959
dc.identifier.issn0016-7398
dc.identifier.urihttps://sciencerep.up.poznan.pl/handle/item/5720
dc.languageen
dc.relation.ispartofGeographical Journal
dc.relation.pages357-369
dc.rightsClosedAccess
dc.sciencecloudnosend
dc.subject.enattention mechanism
dc.subject.endeep learning
dc.subject.enlakes
dc.subject.enmachine learning
dc.subject.enwater level
dc.titleDaily water‐level forecasting for multiple polish lakes using multiple data‐driven models
dc.typeJournalArticle
dspace.entity.typePublication
oaire.citation.issue2
oaire.citation.volume189