the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jul 2026
A new High-Resolution daily Mean South China Sea Ocean Reanalysis dataset: SCSORA
Abstract. High-resolution three-dimensional (3D) ocean reanalysis data are essential for investigating multiscale ocean dynamics in the regional ocean and their impacts on energy transport, marine ecosystems, and climate change. The South China Sea (SCS), as one of the most eddy-active marginal seas in the global ocean, is characterized by complex 3D dynamical processes and frequent extreme ocean events, imposing urgent data demands on the scientific community. This paper presents a new South China Sea Ocean Reanalysis (SCSORA) dataset, with high resolution (1/30°) covering the period 2001–2024, which is generated by the second version of the South China Sea Operational Oceanography Forecast System (SCSOFSv2). SCSORA provides daily 3D fields of temperature, salinity, and current velocity, together with sea surface height (SSH).Three categories of observational data are assimilated into SCSOFSv2, including satellite-derived optimum interpolation sea surface temperature (OISST), along-track sea level anomaly (SLA) from AVISO, and in-situ temperature–salinity Argo profiles. Systematic validations against multisource satellite retrievals, in-situ observations, and existing reanalysis products demonstrates that SCSORA achieves satisfactory accuracy and reliability in reproducing sea surface temperature (RMSE: 0.34 °C), SLA (RMSE: 5.9 cm), and subsurface thermohaline structure. Kinetic energy spectral analysis reveals that SCSORA is capable of resolving ocean dynamical processes spanning from mesoscale to part of the submesoscale range. Representative applications demonstrate the potential of SCSORA in characterizing the spatiotemporal features of marine heatwaves (MHWs) in the SCS and in examining the 3D structural modulation of MHWs by mesoscale eddies, revealing distinct modulation mechanisms of different eddy polarities on the vertical structure of MHWs. SCSORA provides a critical 3D data foundation for the physical oceanography and extreme event research communities focusing on the SCS. It is publicly available at https://doi.org/10.12378/geodb.2026.2.005.V1 (Zhu et al., 2026).
- Preprint
(3682 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 04 Oct 2026)
-
CC1: 'Comment on essd-2026-440', Ranran Zhang, 14 Aug 2026
reply
-
CC2: 'Reply on CC1', Xueming Zhu, 19 Aug 2026
reply
Dear Dr. Zhang,
Thanks for your interesting to our SCSORA dataset. I sincerely apologize that the data URL can not be accessed a few days ago. It is because that the website had been checked for some regular security issues. Right now, it can be accessed.
Please do not hesitate to let us know if you have any question about the dataset.
Regards,
Xueming Zhu
Citation: https://doi.org/10.5194/essd-2026-440-CC2
-
CC2: 'Reply on CC1', Xueming Zhu, 19 Aug 2026
reply
-
RC1: 'Comment on essd-2026-440', Anonymous Referee #1, 26 Aug 2026
reply
A high-resolution ocean reanalysis dataset in the South China Sea named SCSORA was created, covering the period from 2001 to 2024. The horizontal resolution is 1/30 degree, which permits submesoscale circulations. Comparisons with the observation dataset, which are not used to assimilate SCSORA, show how well this high-resolution reanalysis dataset represents the small-scale oceanic state. The dataset appears to be very valuable for investigating spatiotemporal variations of oceanic circulations ranging from submesoscale to large scale in the South Cina Sea. I recommend several minor modifications before publication.
Minor comments
L100-102: What global reanalysis data is used for open boundary conditions?
L125-126: The number of Argo floats should be small in the SCS in the early 2000s. Also, the number of Argo floats should be limited in the southern SCS because the depth is shallow in the region. Please comment more about how the assimilation using Argo data works in the SCS.
L155: Please check the URL for the SLA product.
Sec. 3.1: Could you plot or check the SST anomaly of OISST (Fig. 2top)? Do the time series mostly overlap SCSORA due to the assimilation? There are annual variations in the SST RMSE: large in winter and small in summer. Is this why spatial variations are larger in winter than summer?
Sec. 3.2. If possible, please comment on why the SLA RMSE is relatively large in 2017-2022 (Fig. 3bottom).
Fig. 6: Near the coast and southern SCS south to the equator, the MLD in GLORYS is not estimated. Did you use the strictly same method to estimate the MLD in both dataset?
Sec. 4.2: Could you plot or check the time series of the MHW using OISST in Fig. 10 and Fig. 11? Do these time series overlap with those using SCSORA? How the assimilated data constrains the SST in SCSORA should be checked.
Sec. 4.3: Why is the number of eddies extremely large in 2022, 2023, and 2024 in SCSORA, which is not observed in AVISO (Fig 12 top panels)? Do the small-scale (submesoscale) eddies increase, which cannot be detected in AVISO?
Citation: https://doi.org/10.5194/essd-2026-440-RC1 -
CC3: 'Comment on essd-2026-440', Gyundo Pak, 21 Sep 2026
reply
General comments
This manuscript presents SCSORA, a 1/30 deg daily ocean reanalysis dataset for the SCS covering 2001-2024. The long-term, high-resolution 3-D fields have the potential to provide a useful resource for studies of regional ocean dynamics and extreme events. However, the current manuscript places relatively greater emphasis on scientific applications than on comprehensive validation of the reanalysis itself. Several important aspects, particularly the spatial error characteristics, subsurface structure, and circulation would benefit from further evaluation. In addition, some validation datasets are not fully independent from the observations assimilated into SCSORA, which should be carefully considered when interpreting the results. Given the scope of ESSD, I suggest placing greater emphasis on the validation, limitations, and added value of SCSORA, while reducing some of the detailed process-oriented interpretations in Section 4.3. My specific comments are provided below.1. Description of SCSOFSv2
- The description of SCSOFSv2 in Section 2.1 is currently too brief. Although the system has been documented previously in Zhu et al. (2022), readers should be able to understand the essential configuration and generation procedure of SCSORA without relying heavily on an external paper. I suggest that the authors provide a concise but sufficiently complete summary of the model configuration, SBC/OBCs, initialization or spin-up processes, major mixing parameterizations, and particularly the data assimilation configuration (ensemble construction and size, assimilation cycle and window, localization scheme, observation errors, and assimilation frequency...). In addition, if the SCSOFSv2 configuration used to produce SCSORA differs in any respect from that described in Zhu et al. (2022), these modifications should also be clearly documented.2. Independent observation data sets
- The validation of SST and SSH/SLA would benefit from a three-way comparison among SCSORA, the observations directly assimilated into the system, and the independent or independently processed observational products used for validation. For SST, for example, OISST, SCSORA, and OSTIA could be compared together. Similarly, the assimilated along-track altimeter, SCSORA, and gridded AVISO could be examined for SSH/SLA. Such comparisons would help clarify (1) how closely the reanalysis follows the assimilated observations, (2) whether the resulting variability is also consistently reproduced in the independent validation product, (3) whether discrepancies between the two observational products themselves may be contributing to what is currently interpreted as model bias or error. This distinction is particularly important because the validation products are not fully independent from the observational systems used in the assimilation.3. Spatial pattern validation
- The current SST (Fig. 2) and SSH/SLA (Fig. 3) evaluations are largely focused on temporal variations and domain-averaged statistics. Given the high spatial resolution and regional nature of SCSORA, I suggest that the authors also provide spatial evaluations of the reanalysis performance, for example, maps of bias, RMSE, and correlations. For the subsurface fields, comparisons of temperature and salinity distributions along representative vertical sections would also provide a more direct assessment of how well SCSORA reproduces the observed 3-D ocean structure.4. Comparison with previous regional reanalysis data
- The manuscript would benefit from a clearer discussion of how SCSORA differs from previously available regional reanalysis products for the South China Sea (ex. https://www.nature.com/articles/sdata201452) in Introduction.5. Validation of circulation
- Although SCSORA provides 3-D current velocity fields, the validation of the circulation is limited. I suggest including additional validation metrics for the regional circulation and/or volume transport through key straits, where appropriate estimates are available.6. Use of glider data
- Since the Argo profiles are assimilated into SCSORA, their comparison mainly represents analysis-observation consistency rather than independent validation. I suggest making greater use of the independent glider observations, for example by comparing vertical sections of temperature and salinity. The glider data could also be used for MLD validation, while retaining the current GLORYS comparison to assess consistency between reanalysis products.7. Validation of high-wavenumber variability
- The spectral analysis (Fig.7 & 8) shows that SCSORA contains substantially more energy at high wavenumbers than the coarser reanalysis and observation products. However, the presence of high-wavenumber energy alone does not necessarily show that these small-scale signals are realistically reproduced. The authors are encouraged to provide further validation against available high-resolution observations, where feasible (e.g., SWOT), or otherwise discuss the effective spatial resolution and limitations of interpreting the high-wavenumber spectral energy.8. MHW analysis
- The MHW analysis in Section 4.2 is useful as an application example of SCSORA. However, the differences between SCSORA and OSTIA may partly reflect multiple factors, including the influence of assimilated OISST and the model dynamics. Therefore, statements suggesting superior temporal continuity or event stability of SCSORA (ex. L401) should be interpreted more cautiously, as these differences do not necessarily indicate better representation of MHWs. Without additional independent observations, it is difficult to determine which representation is more realistic.9. Scientific applications
- Section 4.3 goes considerably beyond demonstrating the potential applications of SCSORA and attempts to provide detailed interpretations of eddy dynamics and their modulation of MHWs. However, some of these process-oriented interpretations are not sufficiently supported by the analyses presented. Given that this is primarily a data paper, I suggest reducing the scope and level of interpretation in Section 4.3 and placing greater emphasis on the validation and characterization of SCSORA in Section 3. Section 4.3 could instead be more concisely presented as an example demonstrating the potential utility of the 3-D reanalysis for studying mesoscale eddies and their relationship with MHWs.Citation: https://doi.org/10.5194/essd-2026-440-CC3 -
RC2: 'Comment on essd-2026-440', Gyundo Pak, 23 Sep 2026
reply
Due to an account login error, I inadvertently submitted my review as a Community Comment under the name Gyundo Pak. As the Community Comment cannot be removed or converted to a Referee Comment, please consider the Community Comment I submitted as my official review of this manuscript.
I apologize for any confusion caused. Attached is the same review.
Data sets
South China Sea Ocean Reanalysis (SCSORA) dataset Xueming Zhu https://doi.org/10.12378/geodb.2026.2.005.V1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 264 | 79 | 63 | 406 | 43 | 72 |
- HTML: 264
- PDF: 79
- XML: 63
- Total: 406
- BibTeX: 43
- EndNote: 72
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This appears to be a very valuable dataset. However, the provided download links are non-standard. The institutional data URL cannot be accessed, which hinders data acquisition. It is suggested to release the dataset via commonly used international data repositories.