Skip to main content

Thank you for visiting 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.

Solar farside magnetograms from deep learning analysis of STEREO/EUVI data

Matters Arising to this article was published on 12 February 2021


Solar magnetograms are important for studying solar activity and predicting space weather disturbances1. Farside magnetograms can be constructed from local helioseismology without any farside data2,3,4, but their quality is lower than that of typical frontside magnetograms. Here we generate farside solar magnetograms from STEREO/Extreme UltraViolet Imager (EUVI) 304-Å images using a deep learning model based on conditional generative adversarial networks (cGANs). We train the model using pairs of Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA) 304-Å images and SDO/Helioseismic and Magnetic Imager (HMI) magnetograms taken from 2011 to 2017 except for September and October each year. We evaluate the model by comparing pairs of SDO/HMI magnetograms and cGAN-generated magnetograms in September and October. Our method successfully generates frontside solar magnetograms from SDO/AIA 304-Å images and these are similar to those of the SDO/HMI, with Hale-patterned active regions being well replicated. Thus we can monitor the temporal evolution of magnetic fields from the farside to the frontside of the Sun using SDO/HMI and farside magnetograms generated by our model when farside extreme-ultraviolet data are available. This study presents an application of image-to-image translation based on cGANs to scientific data.

This is a preview of subscription content

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Comparison between HMI magnetograms and AI-generated ones from SDO/AIA 304-Å images.
Fig. 2: A series of 304-Å images and magnetograms.
Fig. 3: A temporal evolution of total unsigned magnetic flux of the NOAA active region 12087 from 3 to 19 June 2014.

Data availability

The code is available at In the readme file, we explain the architecture and selected hyperparameters. The SDO data are available from the SDO data centre (, the Joint Science Operations Center ( and the Korean Data Center for SDO ( The STEREO data are available from the STEREO Science Center ( We used the following open-source packages: NumPy ( and Keras (


  1. 1.

    Solanki, S. K., Inhester, B. & Schussler, M. The solar magnetic field. Rep. Prog. Phys. 69, 563–668 (2006).

    ADS  Article  Google Scholar 

  2. 2.

    Lindsey, C. & Braun, D. C. Seismic images of the far side of the Sun. Science 287, 1799–1801 (2000).

    ADS  Article  Google Scholar 

  3. 3.

    Braun, D. C. & Lindsey, C. Seismic imaging of the far hemisphere of the Sun. Astrophys. J. Lett. 560, 189–192 (2001).

    ADS  Article  Google Scholar 

  4. 4.

    Lindsey, C. & Braun, D. C. Seismic imaging of the Sun’s far hemisphere and its applications in space weather forecasting. Space Weather 15, 761–781 (2017).

    ADS  Article  Google Scholar 

  5. 5.

    Goodfellow, I. et al. Generative adversarial nets. Adv. Neur. 2014, 2672–2680 (2014).

    Google Scholar 

  6. 6.

    Isola, P., Zhu, J.-Y., Zhou, T. & Efros, A. A. Image-to-image translation with conditional adversarial networks. IEEE Proc. Comput. Vision Pattern Recog. 2017, 1125–1134 (2017).

    Google Scholar 

  7. 7.

    Pesnell, W. D., Thompson, B. J. & Chamberlin, P. C. The Solar Dynamics Observatory (SDO). Sol. Phys. 275, 3–15 (2012).

    ADS  Article  Google Scholar 

  8. 8.

    Lemen, J. R. et al. The Atmospheric Imaging Assembly (AIA) on the Solar Dynamics Observatory (SDO). Sol. Phys. 275, 17–40 (2012).

    ADS  Article  Google Scholar 

  9. 9.

    Schou, J. et al. Design and ground calibration of the Helioseismic and Magnetic Imager (HMI) instrument on the Solar Dynamics Observatory (SDO). Sol. Phys. 275, 229–259 (2012).

    ADS  Article  Google Scholar 

  10. 10.

    Howard, R. A. et al. Sun Earth Connection Coronal and Heliospheric Investigation (SECCHI). Space Sci. Rev. 136, 67–115 (2008).

    ADS  Article  Google Scholar 

  11. 11.

    Kaiser, M. L. et al. The STEREO mission: an introduction. Space Sci. Rev. 136, 5–16 (2008).

    ADS  Article  Google Scholar 

  12. 12.

    Dere, K. P. et al. CHIANTIan atomic database for emission lines. Astron. Astrophys. Suppl. 125, 149–173 (1997).

    ADS  Article  Google Scholar 

  13. 13.

    Landi, E., Zanna, G. D., Young, P. R., Dere, K. P. & Mason, H. E. CHIANTI—an atomic database for emission lines. XII. Version 7 of the database. Astrophys. J. 744, 99–107 (2012).

    ADS  Article  Google Scholar 

  14. 14.

    Nitta, N. V. et al. Soft X-ray fluxes of major flares far behind the limb as estimated using STEREO EUV images. Sol. Phys. 288, 241–254 (2013).

    ADS  Article  Google Scholar 

  15. 15.

    Liu, Y. et al. Comparison of line-of-sight magnetograms taken by the Solar Dynamics Observatory/Helioseismic and Magnetic Imager and Solar and Heliospheric Observatory/Michelson Doppler Imager. Sol. Phys. 279, 295–316 (2012).

    ADS  Article  Google Scholar 

Download references


We thank the numerous team members who contributed to the success of the SDO mission, as well as the STEREO mission. We acknowledge the community efforts devoted to developing the open-source packages that were used in this work (NumPy and Keras). This work was supported by the BK21+ Program through the National Research Foundation (NRF) funded by the Ministry of Education of Korea, the Basic Science Research Program through the NRF funded by the Ministry of Education (NRF-2016R1A2B4013131), a grant from the NRF funded by the Korean government (number NRF-2013M1A3A3A02042232), the Korea Astronomy and Space Science Institute under the R&D programme supervised by the Ministry of Science and ICT, the Korea Astronomy and Space Science Institute under the R&D programme ‘Development of a Solar Coronagraph on International Space Station (Project No. 2019-1-850-02)’ supervised by the Ministry of Science and ICT, a grant from the Institute for Information & Communications Technology Promotion (IITP) funded by the Korea government (MSIP) (number 2018-0-01422, ‘Study on Analysis and Prediction Technique of Solar Flares’), and the Artificial Intelligence Laboratory at the InSpace Co., Ltd. The SDO data were (partly) provided by the Korea Data Center (KDC) for SDO in cooperation with NASA and the SDO/HMI Team, which is operated by the KASI.

Author information




T.K., E.P. and H.L. contributed equally to this study. T.K., E.P., H.L. and Y.-J.M. devised the method, analysed data and wrote the manuscript. S.-H.B. participated in discussing the results and contributed to improving the methodology. D.L., S.J., I.-H.C., L.K., M.C. and K.-S.C. participated in discussing the results.

Corresponding author

Correspondence to Yong-Jae Moon.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

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

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kim, T., Park, E., Lee, H. et al. Solar farside magnetograms from deep learning analysis of STEREO/EUVI data. Nat Astron 3, 397–400 (2019).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

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