the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Feb 2024
A daily reconstructed chlorophyll-a dataset in South China Sea from MODIS using OI-SwinUnet
Abstract. Satellite remote sensing of sea surface chlorophyll products sometimes yields a significant amount of sporadic missing data due to various variables, such as weather conditions and operational failures of satellite sensors. The limited nature of satellite observation data impedes the utilization of satellite data in the domain of marine research. Hence, it is highly important to investigate techniques for reconstructing satellite remote sensing data to obtain spatially and temporally uninterrupted and comprehensive data within the desired area. This approach will expand the potential applications of remote sensing data and enhance the efficiency of data usage. To address this series of problems, based on the demand for research on the ecological effects of multiscale dynamic processes in the South China Sea, this paper combines the advantages of the optimal interpolation (OI) method and SwinUnet and successfully develops a deep learning model based on the expected variance in data anomalies, called OI-SwinUnet. The OI-SwinUnet method was used to reconstruct the MODIS chlorophyll-a concentration products of the South China Sea from 2013 to 2017. When comparing the performances of the DINEOF, OI, and Unet approaches, it is evident that the OI-SwinUnet algorithm outperforms the other algorithms in terms of reconstruction. We conduct a reconstruction experiment using different artificial missing patterns to assess the resilience of OI-SwinUnet. Ultimately, the reconstructed dataset was utilized to examine the seasonal variations and geographical distribution of chlorophyll-a concentrations in various regions of the South China Sea. Additionally, the impact of the plume front on the dispersion of phytoplankton in upwelling areas was assessed. The potential use of reconstructed products to investigate the process by which individual mesoscale eddies affect sea surface chlorophyll is also examined.
- Preprint
(8407 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2024-6', Chen Jun, 08 Mar 2024
In this manuscript, a deep learning model, OI-SwinUnet, is proposed for the reconstruction of remotely sensed chlorophyll products, and the model is used to generate MODIS chlorophyll-a concentration products in the South China Sea (SCS) from 2013 to 2017. The reconstructed products can be used to obtain comprehensive spatiotemporal continuum data in the SCS. Research on deep learning processing techniques for remote sensing data is currently quite popular. The proposed deep learning framework does a very creative job of addressing the crucial issue of missing data. The reconstructed data can be applied in ecological monitoring of small- and mesoscale processes in the ocean, indicating that the authors have accomplished excellent results. I think the paper satisfies the goals and specifications of this journal. Naturally, I have some particular comments that the authors should clarify or revise before the paper is officially accepted.
Â
The first comments is about the input. Why did the author select anomalies for SwinUnet's inputs from the first and last 15 days, respectively? (To put it another way, could this last for three days or a week?)
Secondly, the author's method of demonstrating the model's reconstruction performance under various mask percent settings is commendable, but it appears that the graph's performance findings are not sufficiently clear (see from Fig. 12). Here, two recommendations are made: first, select an alternative reconstruction product at a time when there will be a sufficient difference to support the author's position; and second, include graphs with mask percentages of 30% and 70%, i.e., set the plot step size to 20% to reflect more specific information about the changes.
Thirdly, in my opinion, one of the best parts of this research is the use of reconstructed data in specific instances of mesoscale eddies. A useful database for researching the ecological effects of small- and mesoscale ocean phenomena may be produced if the chlorophyll data reconstructed using OI-SwinUnet, as suggested by the authors, are able to accurately restore the chlorophyll information of the missing regions.
Fourthly, the authors link upwelling to the high chlorophyll values seen along the Vietnamese coast throughout the summer. Have other studies verified that upwelling at this location results in changes in chlorophyll, and can relevant literature be shown to bolster the authors' claims?
Other minor comments：
- Is the "satellite-derived" in the x-axis of Fig.7 extracted from aqua or terra, or is the data merged from two sensors? please clarify this.
- The better background color of Fig. 1, 2, 3, 4, 16, and 17 is white.
- It is recommended that Fig. 10's colormap be changed to something other to make it more clear, such as jet.
- 16a is rough and it should be improved. The below figure in Fig. 16a shows too little information.
-
AC1: 'Reply on RC1', Haibin Ye, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-AC1-supplement.pdf
-
RC2: 'Comment on essd-2024-6', Anonymous Referee #2, 30 Mar 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-RC2-supplement.pdf
-
AC2: 'Reply on RC2', Haibin Ye, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Haibin Ye, 30 Apr 2024
Status: closed
-
RC1: 'Comment on essd-2024-6', Chen Jun, 08 Mar 2024
In this manuscript, a deep learning model, OI-SwinUnet, is proposed for the reconstruction of remotely sensed chlorophyll products, and the model is used to generate MODIS chlorophyll-a concentration products in the South China Sea (SCS) from 2013 to 2017. The reconstructed products can be used to obtain comprehensive spatiotemporal continuum data in the SCS. Research on deep learning processing techniques for remote sensing data is currently quite popular. The proposed deep learning framework does a very creative job of addressing the crucial issue of missing data. The reconstructed data can be applied in ecological monitoring of small- and mesoscale processes in the ocean, indicating that the authors have accomplished excellent results. I think the paper satisfies the goals and specifications of this journal. Naturally, I have some particular comments that the authors should clarify or revise before the paper is officially accepted.
Â
The first comments is about the input. Why did the author select anomalies for SwinUnet's inputs from the first and last 15 days, respectively? (To put it another way, could this last for three days or a week?)
Secondly, the author's method of demonstrating the model's reconstruction performance under various mask percent settings is commendable, but it appears that the graph's performance findings are not sufficiently clear (see from Fig. 12). Here, two recommendations are made: first, select an alternative reconstruction product at a time when there will be a sufficient difference to support the author's position; and second, include graphs with mask percentages of 30% and 70%, i.e., set the plot step size to 20% to reflect more specific information about the changes.
Thirdly, in my opinion, one of the best parts of this research is the use of reconstructed data in specific instances of mesoscale eddies. A useful database for researching the ecological effects of small- and mesoscale ocean phenomena may be produced if the chlorophyll data reconstructed using OI-SwinUnet, as suggested by the authors, are able to accurately restore the chlorophyll information of the missing regions.
Fourthly, the authors link upwelling to the high chlorophyll values seen along the Vietnamese coast throughout the summer. Have other studies verified that upwelling at this location results in changes in chlorophyll, and can relevant literature be shown to bolster the authors' claims?
Other minor comments：
- Is the "satellite-derived" in the x-axis of Fig.7 extracted from aqua or terra, or is the data merged from two sensors? please clarify this.
- The better background color of Fig. 1, 2, 3, 4, 16, and 17 is white.
- It is recommended that Fig. 10's colormap be changed to something other to make it more clear, such as jet.
- 16a is rough and it should be improved. The below figure in Fig. 16a shows too little information.
-
AC1: 'Reply on RC1', Haibin Ye, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-AC1-supplement.pdf
-
RC2: 'Comment on essd-2024-6', Anonymous Referee #2, 30 Mar 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-RC2-supplement.pdf
-
AC2: 'Reply on RC2', Haibin Ye, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://essd.copernicus.org/preprints/essd-2024-6/essd-2024-6-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Haibin Ye, 30 Apr 2024
Data sets
OI-SwinUnet reconstructed daily Chlorophyll-a products in the South China Sea Haibin Ye, Chaoyu Yang, Yuan Dong, Shilin Tang, and Chuqun Chen https://doi.org/10.5281/zenodo.10478524
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 476 | 85 | 39 | 600 | 27 | 32 |
- HTML: 476
- PDF: 85
- XML: 39
- Total: 600
- BibTeX: 27
- EndNote: 32
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1