Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Partial-convolution-implemented generative adversarial network for global oceanic data assimilation

Nature Machine Intelligence volume 6, pages 834–843 (2024)Cite this article

Subjects

A preprint version of the article is available at Research Square.

Abstract

The oceanic data assimilation (DA) system has been developed to optimally combine numerical-model predictions with actual measurements from the ocean to create the best estimates of current ocean conditions and their uncertainties, improving our ability to forecast and understand the global climate variations. We developed DeepDA, a global oceanic DA system using deep learning, by integrating a partial convolutional neural network and a generative adversarial network. Partial convolution serves as an observation operator, mapping irregular observational data onto gridded fields, while generative adversarial network incorporates observational information from previous time frames. Our observing system simulation experiments, using simulated observations for the DA, revealed that DeepDA markedly reduces analysis error of the oceanic temperature, outperforming both background and observed values. DeepDA’s real-case global temperature reanalysis spanning from 1981 to 2020 accurately reconstructs observed global climatological temperature fields, along with their seasonal cycles, major oceanic temperature variabilities and global warming trend. Developed solely with a long-term control simulation, DeepDA lowers technical hurdles in creating global ocean reanalysis datasets using multiple numerical models’ physical constraints, thereby diminishing systematic uncertainties in estimating global oceanic states over decades with these models.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Description of the DeepDA system and the OSSEs.
Fig. 2: Comparison of the DeepDA reanalysis to other oceanic reanalysis datasets.
Fig. 3: Climatological fields of the DeepDA reanalysis.
Fig. 4: Interannual variations simulated in the DeepDA.

Similar content being viewed by others

Data availability

The following data related to this study can be downloaded from: HadIOD v.1.2.0.0, https://www.metoffice.gov.uk/hadobs/hadiod/download-hadiod1-2-0-0.html; OISST v.2, https://downloads.psl.noaa.gov/Datasets/noaa.oisst.v2.highres; ERSST v.5, https://psl.noaa.gov/data/gridded/data.noaa.ersst.v5.html, EN4.2.2, https://www.metoffice.gov.uk/hadobs/en4/download-en4-2-2.html; ORAS5, https://doi.org/10.24381/cds.67e8eeb7; COBE, https://downloads.psl.noaa.gov/Datasets/COBE; ECCO v.4r4, https://podaac.jpl.nasa.gov/announcements/2021-04-27-ECCO-Version-4-Datasets-Release; ECDA, https://data1.gfdl.noaa.gov/dods-data/gfdl_cm2_1/Fv_NetCDF_test/pp/ocean_interp/ts/monthly (alternative https://apdrc.soest.hawaii.edu/dods/public_data/GFDL/ecda_v2.0); GECCO3, https://icdc.cen.uni-hamburg.de/thredds/catalog/ftpthredds/EASYInit/GECCO3/regular_1x1_grid/catalog.html GODAS, https://psl.noaa.gov/data/gridded/data.godas.html; MERRA2, https://doi.org/10.5067/4IASLIDL8EEC; SODA v.3.4.2, http://www.atmos.umd.edu/~ocean/index_files/soda3_readme.htm; ARMOR3D L4, https://doi.org/10.48670/moi-00052 and Roemmich–Gilson Argo climatology, https://sio-argo.ucsd.edu/RG_Climatology.html and GHRSST MW-OI, https://doi.org/10.5067/GHMWO-4FR51. Source data are provided with this paper.

Code availability

TensorFlow (https://www.tensorflow.org) libraries were used to formulate the deep-learning model for the global oceanic DA. The developed model and the sample dataset are available via Code Ocean at https://doi.org/10.24433/CO.7269173.v2 and via Zenodo at https://doi.org/10.5281/zenodo.11255094 (ref. 61).

References

  1. Ghil, M. & Malanotte-Rizzoli, P. in Advances in Geophysics (eds Dmowska, R. & Saltzman, B.) Vol. 33, 141–266 (Elsevier, 1991).

  2. Derber, J. & Rosati, A. A global oceanic data assimilation system. J. Phys. Oceanogr. 19, 1333–1347 (1989).

    Article  Google Scholar 

  3. Keppenne, C. L. & Rienecker, M. M. Assimilation of temperature into an isopycnal ocean general circulation model using a parallel ensemble Kalman filter. J. Mar. Syst. 40, 363–380 (2003).

    Article  Google Scholar 

  4. Evensen, G. Using the extended Kalman filter with a multilayer quasi‐geostrophic ocean model. J. Geophys. Res. Oceans 97, 17905–17924 (1992).

    Article  Google Scholar 

  5. Zhang, S., Harrison, M. J., Rosati, A. & Wittenberg, A. System design and evaluation of coupled ensemble data assimilation for global oceanic climate studies. Mon. Weather Rev. 135, 3541–3564 (2007).

    Article  Google Scholar 

  6. Penny, S. G. et al. The local ensemble transform Kalman filter and the running-in-place algorithm applied to a global ocean general circulation model. Nonlinear Process. Geophys. 20, 1031–1046 (2013).

    Article  Google Scholar 

  7. Sugiura, N. et al. Development of a four-dimensional variational coupled data assimilation system for enhanced analysis and prediction of seasonal to interannual climate variations. J. Geophys. Res. Oceans 113, C10017 (2008).

  8. Kalnay, E. et al. The NCEP/NCAR reanalysis project. Bull. Am. Meteorol. Soc. 77, 437–471 (1996).

    Article  Google Scholar 

  9. Carton, J. A., Chepurin, G. A. & Chen, L. SODA3: a new ocean vlimate reanalysis. J. Clim. 31, 6967–1983 (2018).

    Article  Google Scholar 

  10. Zuo, H., Balmaseda, M. A., Tietsche, S., Mogensen, K. & Mayer, M. The ECMWF operational ensemble reanalysis-analysis system for ocean and sea ice: a description of the system and assessment. Ocean Sci. 15, 779–808 (2019).

    Article  Google Scholar 

  11. Saha, S. et al. The NCEP climate forecast system version 2. J. Clim. 27, 2185–2208 (2014).

    Article  Google Scholar 

  12. Takaya, Y. et al. Japan meteorological agency/meteorological research institute-coupled prediction system version 2 (JMA/MRI-CPS2): atmosphere–land–ocean–sea ice coupled prediction system for operational seasonal forecasting. Clim. Dyn. 50, 751–765 (2018).

    Article  Google Scholar 

  13. Waters, J. et al. Implementing a variational data assimilation system in an operational 1/4 degree global ocean model. Q. J. R. Meteorol. Soc. 141, 333–349 (2015).

    Article  Google Scholar 

  14. Hersbach, H. et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 146, 1999–2049 (2020).

    Article  Google Scholar 

  15. Casas, C. Q., Arcucci, R., Wu, P., Pain, C. & Guo, Y. K. A reduced order deep data assimilation model. Phys. D Nonlinear Phenom. 412, 132615 (2020).

    Article  MathSciNet  Google Scholar 

  16. Arcucci, R., Zhu, J., Hu, S. & Guo, Y. K. Deep data assimilation: integrating deep learning with data assimilation. Appl. Sci. 11, 1114 (2021).

    Article  Google Scholar 

  17. Tang, M., Liu, Y. & Durlofsky, L. J. A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems. J. Comput. Phys. 413, 109456 (2020).

    Article  MathSciNet  Google Scholar 

  18. Buizza, C. et al. Data learning: integrating data assimilation and machine learning. J. Comput. Sci. 58, 101525 (2022).

    Article  Google Scholar 

  19. Kim, H., Ham, Y. G., Joo, Y. S. & Son, S. W. Deep learning for bias correction of MJO prediction. Nat. Commun. 12, 3087 (2021).

    Article  Google Scholar 

  20. Grönquist, P. et al. Deep learning for post-processing ensemble weather forecasts. Philos. Trans. A Math. Phys. Eng. Sci. 379, 20200092 (2021).

    MathSciNet  Google Scholar 

  21. Liu, G. et al. Image inpainting for irregular holes using partial convolutions. In Proc. 15th European Conference on Computer Vision 89–105 (Springer, 2018).

  22. Kadow, C., Hall, D. M. & Ulbrich, U. Artificial intelligence reconstructs missing climate information. Nat. Geosci. 13, 408–413 (2020).

    Article  Google Scholar 

  23. Goodfellow, I. J. et al. Generative adversarial nets. Proc. Adv. Neural Inf. Process. Syst. 3, 2672–2680 (2014).

    Google Scholar 

  24. Pathak, D., Krähenbühl, P., Donahue, J., Darrell, T. & Efros, A. A. Context encoders: feature learning by inpainting. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 2536–2544 (IEEE, 2016).

  25. Lam, R. et al. Learning skillful medium-range global weather forecasting. Science 382, 1416–1421 (2023).

    Article  MathSciNet  Google Scholar 

  26. Iten, R., Metger, T., Wilming, H., Del Rio, L. & Renner, R. Discovering physical concepts with neural networks. Phys. Rev. Lett. 124, 010508 (2020).

    Article  Google Scholar 

  27. Bertino, L., Evensen, G. & Wackernagel, H. Sequential data assimilation techniques in oceanography. Int. Stat. Rev. 71, 223–241 (2003).

    Article  Google Scholar 

  28. Rodgers, K. B. et al. Ubiquity of human-induced changes in climate variability. Earth Syst. Dyn. 12, 1393–1411 (2021).

    Article  Google Scholar 

  29. Arnold, C. P. & Dey, C. H. Observing-systems simulation experiments: past, present, and future. Bull. Am. Meteorol. Soc. 67, 687–695 (1986).

    Article  Google Scholar 

  30. Atkinson, C. P., Rayner, N. A., Kennedy, J. J. & Good, S. A. An integrated database of ocean temperature and salinity observations. J. Geophys. Res.: Oceans 119, 7139–7163 (2014).

    Article  Google Scholar 

  31. Kalnay, E. Atmospheric Modeling, Data Assimilation and Predictability (Cambridge Univ. Press, 2003).

  32. Huang, B. et al. Improvements of the daily optimum interpolation sea surface temperature (DOISST) version 2.1. J. Clim. 34, 2923–2939 (2021).

    Article  Google Scholar 

  33. Kim, Y. H., Hwang, C. & Choi, B. J. An assessment of ocean climate reanalysis by the data assimilation system of KIOST from 1947 to 2012. Ocean Model. 91, 1–22 (2015).

    Article  Google Scholar 

  34. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. & Salakhutdinov, R. Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15, 1929–1958 (2014).

    MathSciNet  Google Scholar 

  35. Huang, B. et al. Extended Reconstructed Sea Surface Temperature, version 5 (ERSST v5): upgrades, validations, and intercomparisons. J. Clim. 30, 8179–8205 (2017).

    Article  Google Scholar 

  36. Good, S. A., Martin, M. J. & Rayner, N. A. EN4: quality controlled ocean temperature and salinity profiles and monthly objective analyses with uncertainty estimates. J. Geophys. Res. Oceans 118, 6704–6716 (2013).

    Article  Google Scholar 

  37. Dash, P. et al. Group for High Resolution Sea Surface Temperature (GHRSST) analysis fields inter-comparisons—part 2: near real time web-based level 4 SST Quality Monitor (L4-SQUAM). Deep Sea Res. II: Top. Stud. Oceanogr. 77, 31–43 (2012).

    Article  Google Scholar 

  38. Roemmich, D. & Gilson, J. The 2004–2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the Argo Program. Prog. Oceanogr. 82, 81–100 (2009).

    Article  Google Scholar 

  39. Castruccio, F. et al. An EnOI‐based data assimilation system with DART for a high‐resolution version of the CESM2 ocean component. J. Adv. Model. Earth Syst. 12, e2020MS002176 (2020).

    Article  Google Scholar 

  40. Freeman, E. et al. ICOADS release 3.0: a major update to the historical marine climate record. Int. J. Climatol. 37, 2211–2232 (2017).

    Article  Google Scholar 

  41. Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9, 1937–1958 (2016).

    Article  Google Scholar 

  42. DelSole, T., Nattala, J. & Tippett, M. K. Skill improvement from increased ensemble size and model diversity. Geophys. Res. Lett. 41, 7331–7342 (2014).

    Article  Google Scholar 

  43. Ronneberger, O., Fischer, P. & Brox, T. U-net: convolutional networks for biomedical image segmentation. In Proc. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 Vol. 9351 (eds Navab, N. et al.) 234–241 (Springer, 2015).

  44. Isola, P., Zhu, J.-Y., Zhou, T., Efros, A. A. & Research, B. A. Image-to-Image translation with conditional adversarial networks. In Proc. IEEE Conference on Computer Vision and Pattern Recognition 1125–1134 (2017).

  45. Laves, M. H., Ihler, S., Kortmann, K. P., & Ortmaier, T. Calibration of model uncertainty for dropout variational inference. Preprint at https://arxiv.org/abs/2006.11584 (2020).

  46. Leutbecher, M. et al. Stochastic representations of model uncertainties at ECMWF: state of the art and future vision. Q. J. R. Meteorol. Soc. 143, 2315–2339 (2017).

    Article  Google Scholar 

  47. He, K., Zhang, X., Ren, S. & Sun, J. Delving deep into rectifiers: surpassing human-level performance on imagenet classification. In Proc. IEEE International Conference on Computer Vision 1026–1034 (2015).

  48. Glorot, X. & Bengio, Y. Understanding the difficulty of training deep feedforward neural networks. In Proc. Thirteenth International Conference on Artificial Intelligence and Statistics 249–256 (2010).

  49. Ishii, M., Shouji, A., Sugimoto, S. & Matsumoto, T. Objective analyses of sea‐surface temperature and marine meteorological variables for the 20th century using ICOADS and the Kobe collection. Int. J. Climatol.: J. R. Meteorol. Soc. 25, 865–879 (2005).

    Article  Google Scholar 

  50. Gouretski, V. & Reseghetti, F. On depth and temperature biases in bathythermograph data: development of a new correction scheme based on analysis of a global ocean database. Deep Sea Res. I: Oceanogr. Res. Papers 57, 812–833 (2010).

    Article  Google Scholar 

  51. Gouretski, V. & Cheng, L. Correction for systematic errors in the global dataset of temperature profiles from mechanical bathythermographs. J. Atmos. Oceanic Technol. 37, 841–855 (2020).

    Article  Google Scholar 

  52. Guinehut, S., Dhomps, A. L., Larnicol, G. & Le Traon, P. Y. High resolution 3-D temperature and salinity fields derived from in situ and satellite observations. Ocean Sci. 8, 845–857 (2012).

    Article  Google Scholar 

  53. Fukumori, I. et al. Synopsis of the ECCO Central Production Global Ocean and Sea-Ice State Estimate, version 4 release 4 (4 release 4). Zenodo https://doi.org/10.5281/zenodo.3765929 (2020).

  54. Chang, Y. S., Zhang, S., Rosati, A., Delworth, T. L. & Stern, W. F. An assessment of oceanic variability for 1960–2010 from the GFDL ensemble coupled data assimilation. Clim. Dyn. 40, 775–803 (2013).

    Article  Google Scholar 

  55. Köhl, A. Evaluating the GECCO3 1948–2018 ocean synthesis–a configuration for initializing the MPI-ESM climate model. Q. J. R. Meteorol. Soc. 146, 2250–2273 (2020).

    Article  Google Scholar 

  56. Behringer, D. & Xue, Y. Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean. In Proc. Eighth Symp. on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface. 11–15 (AMS, 2004).

  57. Gelaro, R. et al. The modern-era retrospective analysis for research and applications, version 2 (MERRA-2). J. Clim. 30, 5419–5454 (2017).

    Article  Google Scholar 

  58. Woodruff, S. D. et al. ICOADS Release 2.5: extensions and enhancements to the surface marine meteorological archive. Int. J. Climatol. 31, 951–967 (2011).

    Article  Google Scholar 

  59. McPhaden, M. J., Busalacchi, A. J. & Anderson, D. L. A toga retrospective. Oceanography 23, 86–103 (2010).

    Article  Google Scholar 

  60. Argo. Argo float data and metadata from Global Data Assembly Centre (Argo GDAC). SEANOE https://doi.org/10.17882/42182 (2022).

  61. Kim, J.-H. The developed DeepDA model and the sample dataset. Zenodo https://doi.org/10.5281/zenodo.11255094 (2020).

Download references

Acknowledgements

This study is supported by Korea Environment Industry and Technology Institute through ‘Climate Change R&D Project for New Climate Regime’, funded by Korea Ministry of Environment (grant no. 2022003560006). Y.-G.H. was supported by the Ministry of Science and ICT through the National Research Foundation of Korea (NRF-2022M3K3A1094114).

Author information

Authors and Affiliations

  1. Department of Environmental Planning, Graduate School of Environmental Studies, Seoul National University, Seoul, South Korea

    Yoo-Geun Ham & Jeong-Gil Lee

  2. Department of Oceanography, Chonnam National University, Gwangju, South Korea

    Yong-Sik Joo

  3. Center for Sustainable Environment Research, Korea Institute of Science and Technology, Seoul, South Korea

    Jeong-Hwan Kim

Authors
  1. Yoo-Geun Ham
    View author publications

    Search author on:PubMed Google Scholar

  2. Yong-Sik Joo
    View author publications

    Search author on:PubMed Google Scholar

  3. Jeong-Hwan Kim
    View author publications

    Search author on:PubMed Google Scholar

  4. Jeong-Gil Lee
    View author publications

    Search author on:PubMed Google Scholar

Contributions

Y.-G.H. designed the study. J.-H.K., Y.-S.J. and Y.-G.H. formulated the deep-learning model and performed the experiments. J.-G.L. provided the observational data. Y.-S.J. analysed the results. Y.-G.H. wrote the paper. All the authors discussed the study results.

Corresponding author

Correspondence to Yoo-Geun Ham.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Machine Intelligence thanks James Carton, Peter van Leeuwen, Christopher Kadow and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Initial adjustment of DeepDA-produced subsurface temperature anomalies in the OSSEs.

The temperature anomalies of the (a) ground truth, (b) observations, (c) analysis, and (d) background states at 5 m at first data assimilation cycle in the Observing System Simulation Experiment (OSSEs). (e)-(h) same as (a)-(d), but for 105 m. (i)-(l) same as (a)-(d), but for 198 m. (m)-(p) same as (a)-(d), but for 707 m.

Extended Data Fig. 2 Horizontal distribution of the analysis increment in the OSSEs using the DeepDA.

The analysis increment (that is, analysis states minus background states) with a single observation at (a) 0.5oS, 101oW, (b) 0.5oS, 159oE, (c) 35.5oN, 51oW, (d) 19.5oS, 99oE, (e) 19.5oS, 59oE, and (f) 0.5oS, 6oW at January 1st 1997 in the OSSEs.

Extended Data Fig. 3 Reduction of the loss values in time and DeepDA-produced subsurface temperature anomalies in the OSSEs after 3-months spin-up.

(a) Time-series of the globally- and vertically- (from 0 to 700 m) averaged root-mean-squared-errors (RMSEs) of the temperature anomalies in the analysis fields from the ground truth during January to December 1974 in the OSSEs. Black dots denote each assimilation cycle. The temperature anomalies of the (b) ground truth, (c) observations, (d) analysis states, and (e) background states at the surface layer at 15th data assimilation cycle in the Observing System Simulation Experiments (OSSEs). (f)-(i) same as (b)-(e), but for 105 m. (j)-(m) same as (b)-(e), but for 198 m. (n)-(q) same as (b)-(e), but for 707 m.

Source data

Extended Data Fig. 4 Ensemble spread in DeepDA reanalysis.

Ensemble spread of the monthly temperature from 1981 to 2020 at each vertical level using 100 ensemble members in real-case DeepDA reanalysis (panels in left columns), spread between 6 different oceanic reanalysis products (that is, ORAS5, ECCO v4r4, ECDA, GECCO2, GODAS, SODA v3.4.2) (panels in mid-columns) and the ensemble spread in 5 ensemble members in ORAS5 (panels in right columns). Oceanic surface (a–c), 100 m (d–f), 200 m (g–i), 300 m (j–l) and 500 m depth (m,o).

Extended Data Fig. 5 RMSE of the DeepDA reanalysis in the observed locations.

(a) Tropical Pacific (TP, 14.5°S-14.5°N, 119.5°E-239.5°E) (b) North Pacific (NP, 20–55°N, 119.5°E-239.5°E), (c) Northern Atlantic (NA, 1.5°N-64.5°N, 279.5°E-359.5°E), and (d) Indian Ocean (IO, 19.5°S-19°N, 39.5°E-119°E) average of the monthly-averaged temperature RMSE from surface to 750 m in the DeepDA and other oceanic reanalysis products. The RMSE is calculated using the in situ ARGO and TAO observations.

Source data

Extended Data Fig. 6 Annual-mean SST and its biases in various oceanic reanalysis products.

Annual-mean SST (contour), and the difference from the reference data (that is, ERSST V5) (shading) during 1981–2020 in (a) ORAS5, (b) COBE, (c) ECCO v4r4, (d) ECDA v2, (e) GECCO2, (f) OSTIA, (g) GODAS, (h) MERRA2, and (i) SODA v3.4.2.

Extended Data Fig. 7 Annual-mean T300 and its biases in various oceanic reanalysis products.

Annual-mean temperature averaged from the surface to 300 m (T300) in DeepDA (contour), and the difference from (a) EN4.2.2, and (b) ARMOR3D L4 reference dataset. Annual-mean T300 (contour), and the difference from the Roemmich-Gilson Argo climatology (shading) during 1994–2018 in (c) ORAS5, (d) ECCO v4r4, (e) ECDA v2, (f) GECCO2, (g) GODAS, and (h) SODA v3.4.2.

Extended Data Fig. 8 Correlation skill of SST and T300 anomalies in various oceanic reanalysis products.

Temporal anomaly correlation coefficients of the monthly SST anomalies between the ERSST V5 and (a) ORAS5, (b) COBE, (c) ECCO v4r4, (d) ECDA v2, (e) GECCO2, (f) OSTIA, (g) GODAS, (h) MERRA2, or (i) SODA v3.4.2 during 1981–2020. Temporal anomaly correlation coefficients of the monthly T300 anomalies between the EN4.2.2 and (j) ORAS5, (k) ECCO v4r4, (l) ECDA v2, (m) GECCO2, (n) GODAS, or (o) SODA v3.4.2 during 1981–2020.

Extended Data Fig. 9 Equatorial subsurface temperature anomalies during El Nino peak season.

Equatorial temperature anomalies from surface to 300 m at January 1998 in (a) EN4.2.2, (b) DeepDA, (c) ORAS5, (d) ECCO v4r4, (e) ECDA v2, (f) GECCO2, (g) GODAS, and (h) SODA v3.4.2.

Extended Data Table 1 Oceanic temperature & SST database
Full size table

Source data

Source Data Fig. 1

r.m.s.e. of the temperature for testing period in the OSSEs.

Source Data Fig. 2

Taylor diagram of the global-mean monthly OHC500.

Source Data Fig. 3

Taylor diagram of the seasonal climatology difference T300.

Source Data Extended Data Fig. 3

r.m.s.e. between the analysis fields and the true states in the OSSEs.

Source Data Extended Data Fig. 5

Vertical profiles of monthly averaged temperature r.m.s.e.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ham, YG., Joo, YS., Kim, JH. et al. Partial-convolution-implemented generative adversarial network for global oceanic data assimilation. Nat Mach Intell 6, 834–843 (2024). https://doi.org/10.1038/s42256-024-00867-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • Issue date:

  • DOI: https://doi.org/10.1038/s42256-024-00867-x

This article is cited by

Search

Advanced search

Quick links

Nature Briefing Anthropocene

Sign up for the Nature Briefing: Anthropocene newsletter — what matters in anthropocene research, free to your inbox weekly.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing: Anthropocene