the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jun 2024
GEST: Accurate global ocean surface current reconstruction withmulti-scale dynamics-informed neural network
Abstract. Exceptional precision and excellent resolution reconstruction of sea surface currents are beneficial for exploring complex oceanic dynamic processes. Normally, this required physical inversion models for global or regional oceans are constructed to reconstruct oceanic currents. These models are based on the analysis of sea surface geostrophic and Ekman currents derived from satellite observations of sea level and wind stress fields. Nevertheless, the presence of various typical dynamic processes in marine environments, such as mesoscale eddies and small-scale waves, continues to pose challenges in accurately reconstructing oceanic currents. Meanwhile, any product of surface current that neglects the contribution of wave motion would, at best, be incomplete. Therefore, in this paper, we introduce an accurate sea surface current product at a depth of 15 m, named GEST (Geostrophic-Ekman-Stokes-Tide). This product is produced by a multi-scale dynamics-informed neural network that learns the intricate representation of concealed characteristics in Ekman, geostrophic currents, wave-induced Stokes drift, and TPXO9 tidal currents. Its structure design is predicated upon the intricate coupling relationships between various ocean surface components and the veritable currents discerned by the deployment of drift buoys, with each ocean surface component correlating to discrete physical processes. Compared with the prevailing product, the GEST confers an elevation in precision by approximately 9.2 cm/s over the traditional multinomial fitting method, 10.4 cm/s beyond the OSCAR, and 8.81 cm/s surpassing GlobCurrent.
- Preprint
(14410 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2024-190', Anonymous Referee #1, 09 Aug 2024
General Comments: The manuscript (MS No.: essd-2024-190) introduces the GEST product, representing a significant advancement in reconstructing sea surface currents using a multi-scale dynamics-informed neural network. This innovative approach effectively integrates various ocean surface components, including geostrophic and Ekman currents, wave-induced Stokes drift, and TPXO9 tidal currents, to enhance the precision and resolution of current reconstructions. However, several critical areas need to be addressed to ensure the robustness, reliability, and practical applicability of the GEST model. Major revisions must resolve these issues before the manuscript can be considered for publication in ESSD.
Specific Comments：
- The manuscript lacks detailed information on how the stability and robustness of the GEST model are ensured. Stability and reliability are crucial for the practical application of deep learning models. The authors should include comprehensive validation and verification steps to demonstrate the model's consistency and dependability over time.
- There is an absence of a thorough sensitivity analysis of the parameters used in the neural network algorithm. Understanding how variations in these parameters affect the model's performance is essential for assessing its robustness. The manuscript should include a detailed sensitivity analysis to highlight the impact of different parameters on the GEST model's accuracy and reliability.
- To validate the robustness of the model under various conditions, the manuscript should include additional comparison experiments. These experiments should evaluate the model's performance across different regions and seasonal conditions, providing a clearer understanding of the GEST product's effectiveness in diverse scenarios.
- The manuscript should provide detailed steps to ensure the reproducibility of the results. This includes clear documentation of the data preprocessing, model training, and validation processes. Reproducibility is a fundamental aspect of scientific research, and the lack of detailed methodology undermines the credibility of the findings.
- The manuscript should explore the scalability of the GEST model for real-time ocean surface current prediction applications. Discussing the model's potential for real-time implementation and its computational requirements would enhance its practical value and relevance to the field of oceanography.
Citation: https://doi.org/10.5194/essd-2024-190-RC1 -
RC2: 'Comment on essd-2024-190', Anonymous Referee #2, 27 Sep 2024
This manuscript devised a novel multiscale-dynamics-informed model using machine learning and deep learning techniques. The authors produced a new product of the global ocean current, integrating Geostrophic, Ekman, Stokes, and Tidal multiscale current. However, despite the authors' efforts, it is questionable how innovative the new product is compared to previous products. It seems that the accuracy of the results by the model currently described in this manuscript and the comparison results with previous products alone are insufficient to emphasize the creativity of this study. The input data used to train the GESTNet model are not very characteristic. Above all, it is doubtful whether a daily product with 0.25-degree resolution is sufficient to reveal the multi-scale dynamic process emphasized by the author. To prove this, it is necessary to present a practical application of the new product. Due to this, this manuscript is not recommended to be published.
Citation: https://doi.org/10.5194/essd-2024-190-RC2
Status: closed
-
RC1: 'Comment on essd-2024-190', Anonymous Referee #1, 09 Aug 2024
General Comments: The manuscript (MS No.: essd-2024-190) introduces the GEST product, representing a significant advancement in reconstructing sea surface currents using a multi-scale dynamics-informed neural network. This innovative approach effectively integrates various ocean surface components, including geostrophic and Ekman currents, wave-induced Stokes drift, and TPXO9 tidal currents, to enhance the precision and resolution of current reconstructions. However, several critical areas need to be addressed to ensure the robustness, reliability, and practical applicability of the GEST model. Major revisions must resolve these issues before the manuscript can be considered for publication in ESSD.
Specific Comments：
- The manuscript lacks detailed information on how the stability and robustness of the GEST model are ensured. Stability and reliability are crucial for the practical application of deep learning models. The authors should include comprehensive validation and verification steps to demonstrate the model's consistency and dependability over time.
- There is an absence of a thorough sensitivity analysis of the parameters used in the neural network algorithm. Understanding how variations in these parameters affect the model's performance is essential for assessing its robustness. The manuscript should include a detailed sensitivity analysis to highlight the impact of different parameters on the GEST model's accuracy and reliability.
- To validate the robustness of the model under various conditions, the manuscript should include additional comparison experiments. These experiments should evaluate the model's performance across different regions and seasonal conditions, providing a clearer understanding of the GEST product's effectiveness in diverse scenarios.
- The manuscript should provide detailed steps to ensure the reproducibility of the results. This includes clear documentation of the data preprocessing, model training, and validation processes. Reproducibility is a fundamental aspect of scientific research, and the lack of detailed methodology undermines the credibility of the findings.
- The manuscript should explore the scalability of the GEST model for real-time ocean surface current prediction applications. Discussing the model's potential for real-time implementation and its computational requirements would enhance its practical value and relevance to the field of oceanography.
Citation: https://doi.org/10.5194/essd-2024-190-RC1 -
RC2: 'Comment on essd-2024-190', Anonymous Referee #2, 27 Sep 2024
This manuscript devised a novel multiscale-dynamics-informed model using machine learning and deep learning techniques. The authors produced a new product of the global ocean current, integrating Geostrophic, Ekman, Stokes, and Tidal multiscale current. However, despite the authors' efforts, it is questionable how innovative the new product is compared to previous products. It seems that the accuracy of the results by the model currently described in this manuscript and the comparison results with previous products alone are insufficient to emphasize the creativity of this study. The input data used to train the GESTNet model are not very characteristic. Above all, it is doubtful whether a daily product with 0.25-degree resolution is sufficient to reveal the multi-scale dynamic process emphasized by the author. To prove this, it is necessary to present a practical application of the new product. Due to this, this manuscript is not recommended to be published.
Citation: https://doi.org/10.5194/essd-2024-190-RC2
Data sets
GEST Ocean Surface Current 2.0 Linyao Ge and Guiyu Wang https://doi.org/10.5281/zenodo.11142408
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 448 | 103 | 80 | 631 | 18 | 19 |
- HTML: 448
- PDF: 103
- XML: 80
- Total: 631
- BibTeX: 18
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Linyao Ge
Guiyu Wang
Baoxiang Huang
Chuanchuan Cao
Xiaoyan Chen
Ge Chen
High precision in reconstructing sea surface currents is vital for understanding ocean dynamics. Our paper introduces GEST (Geostrophic-Ekman-Stokes-Tide), a 15 m depth sea current product. GEST, generated by a neural network, captures Ekman, geostrophic currents, Stokes drift, and TPXO9 tidal currents. Its design accounts for complex ocean surface dynamics, surpassing OSCAR and GlobCurrent by 10.4 cm/s and 8.81 cm/s, respectively.
High precision in reconstructing sea surface currents is vital for understanding ocean...