pith. sign in

arxiv: 2606.08652 · v1 · pith:FM77PBPRnew · submitted 2026-06-07 · 🌌 astro-ph.SR · cs.AI· cs.CV

Reconstructing Synthetic SDO/AIA 193 A EUV Images from He I 10830 A Observations with Diffusion Model Translator

Marco Marena , Qin Li , Haimin Wang , Haodi Jiang , Prajwal Shah , Bo Shen This is my paper

Pith reviewed 2026-06-27 17:53 UTC · model grok-4.3

classification 🌌 astro-ph.SR cs.AIcs.CV
keywords solar coronaimage translationdiffusion modelsHe I 10830EUV reconstructioncoronal holeshistorical solar data
0
0 comments X

The pith

A diffusion model trained on recent solar images translates decades of helium observations into synthetic extreme-ultraviolet coronal images.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a conditional diffusion model called CH-aware DMT that takes full-disk He I 10830 Å images as input and outputs synthetic AIA 193 Å images. It trains the model on co-aligned pairs from 2011-2015 and tests reconstruction quality on held-out data plus comparisons with earlier SOHO/EIT, Yohkoh/SXT, and long-term activity proxies. The resulting synthetic images preserve large-scale coronal structure and coronal-hole locations well enough to serve as a proxy for periods before routine EUV imaging existed. This opens the possibility of studying multi-decade changes in coronal evolution using the continuous He I record that begins in 1974.

Core claim

The CH-aware Diffusion Model Translator reconstructs synthetic SDO/AIA 193 Å images from He I inputs such that dominant full-disk EUV morphology is preserved with correlation 0.92 and coronal-hole low-intensity structure is recovered with correlation 0.84 on 2015 test data; the same model applied to earlier KPVT He I data yields images whose morphology and disk-integrated emission track SOHO/EIT 195 Å, Yohkoh/SXT, sunspot number, and F10.7 flux over 1974-2015.

What carries the argument

The CH-aware Diffusion Model Translator, a diffusion-based conditional image translation network that accepts He I 10830 Å full-disk images and produces AIA 193 Å outputs while remaining sensitive to coronal-hole boundaries.

If this is right

  • Reconstructed images can be used for multi-decade analyses of large-scale coronal evolution before direct EUV imaging existed.
  • Morphology of the synthetic images remains consistent with independent SOHO/EIT 195 Å observations over 2005-2015.
  • Disk-integrated emission from the reconstructions tracks independent solar activity proxies across 1974-2015.
  • The same translation can be applied to KPVT He I data to generate proxies that align with Yohkoh/SXT soft X-ray observations.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same diffusion translation approach could be tested on other wavelength pairs to generate additional historical solar proxies.
  • Long-term reconstructed series might reveal systematic changes in coronal structure across solar cycles that are not captured in the short training window.
  • The method could be combined with existing He I-based open-field maps to produce joint historical datasets for space-weather or dynamo studies.

Load-bearing premise

The statistical mapping learned from 2011-2015 image pairs continues to produce accurate results when the input He I data come from different instruments and earlier parts of the solar cycle.

What would settle it

A quantitative comparison showing that the correlation between reconstructed 193 Å images and actual overlapping EIT or AIA observations drops below 0.7 for any multi-year interval would falsify the claim of historical applicability.

Figures

Figures reproduced from arXiv: 2606.08652 by Bo Shen, Haimin Wang, Haodi Jiang, Marco Marena, Prajwal Shah, Qin Li.

Figure 1
Figure 1. Figure 1: Diffusion Model Translator (DMT) for conditional EUV reconstruction. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: CH reconstruction on the AIA 193 ˚A test set (2011–2015). For three representative test dates, we show: (left) the conditioning He i 10830 ˚A input; (middle-left) the GT AIA 193 ˚A image; (middle-right) the synthetic AIA 193 ˚A image generated by CH-aware DMT; and (right) the CH mask accuracy map. CH masks are derived independently from GT and synthetic AIA 193 ˚A using Otsu thresholding (N. Otsu 1979). Gr… view at source ↗
Figure 3
Figure 3. Figure 3: Evaluation over 2005–2015 using SOHO/EIT 195 [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Qualitative comparison of Carrington synoptic intensity maps of ground truth EIT 195 [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Synthetic AIA 193 ˚A Carrington synoptic maps reconstructed from KPVT He i 10830 ˚A observations. Each panel shows a Carrington synoptic map constructed from the synthetic AIA 193 ˚A full-disk sequence generated by applying the trained translator to KPVT He i 10830 ˚A inputs from 1974 to 1993. The maps are assembled at a 1-day cadence using central-meridian strips, so gaps indicate undersampled longitudes … view at source ↗
Figure 6
Figure 6. Figure 6: Morphological comparison of synthetic AIA 193 [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Long-term normalized comparison of disk-integrated EUV statistics and solar-activity proxies. [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
read the original abstract

Routine full-disk EUV imaging has been available only since the modern era, such as SOHO and SDO. To extend EUV coronal context into earlier periods, we leverage the multi-decade availability of full-disk \HeI{} observations, whose absorption is modulated by coronal irradiance and magnetic topology and is widely used as a proxy for open-field regions. We present a diffusion-based conditional image translation framework, Coronal Hole-aware Diffusion Model Translator (CH-aware DMT), to reconstruct synthetic SDO/AIA 193 \AA{} EUV images from \HeI{} inputs. The model is trained on temporally co-aligned SOLIS \HeI{} and AIA 193 \AA{} pairs spanning 2011--2015 using a month-based split, where January--October are used for training, November is used for validation, and December for testing. On the held-out test set, the reconstructions preserve dominant full-disk EUV morphology (CC=0.92) and recover CH-related low-intensity structure (CC=0.84). We further assess historical applicability by (1) comparing reconstructed AIA 193 \AA{} morphology with SOHO/EIT 195 \AA{} over 2005--2015; (2) comparing reconstructed AIA 193 \AA{} images generated from KPVT \HeI{} inputs against Yohkoh/SXT soft X-ray observations; and (3) evaluating long-term reconstructed disk-integrated emission statistics against observational EUV series and independent solar activity proxies (sunspot number and F10.7 radio flux over 1974--2015). These results indicate that CH-aware DMT conditioned on \HeI{} can provide a physically plausible synthetic AIA 193 \AA{} coronal proxy for historical studies, supporting multi-decade analyses of large-scale coronal evolution before the direct EUV imaging was available.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 3 minor

Summary. The manuscript introduces a Coronal Hole-aware Diffusion Model Translator (CH-aware DMT) that uses a conditional diffusion model to reconstruct synthetic SDO/AIA 193 Å EUV images from He I 10830 Å full-disk observations. The model is trained on temporally co-aligned SOLIS HeI and AIA pairs from 2011–2015 with a month-based split (Jan–Oct training, Nov validation, Dec test), yielding test-set correlations of CC=0.92 (full disk) and CC=0.84 (CH regions). The trained model is then applied to KPVT HeI data spanning 1974–2015; historical applicability is assessed via morphological comparison to SOHO/EIT 195 Å (2005–2015), comparison to Yohkoh/SXT soft X-ray images, and disk-integrated emission statistics versus sunspot number and F10.7 flux. The central claim is that the resulting synthetic EUV images constitute a physically plausible proxy enabling multi-decade analyses of large-scale coronal evolution before routine EUV imaging.

Significance. If the learned conditional mapping generalizes reliably across instrument changes and solar-cycle phases, the work would provide a valuable extension of the EUV coronal record back to 1974, supporting long-term studies of coronal-hole evolution and global coronal structure that are otherwise limited to the post-SOHO era. The diffusion-model approach to conditional image translation is novel in this domain and the reported test-set metrics plus multi-proxy historical checks constitute concrete, falsifiable evidence.

major comments (2)
  1. [Abstract] Abstract (and § on historical validation): the headline claim that CH-aware DMT supplies a usable proxy for 1974–2015 KPVT data rests on the unquantified assumption that the 2011–2015 SOLIS/AIA mapping transfers without systematic degradation; the only pixel-level CC values are from the 2015 test month, while the three historical checks (EIT morphology, SXT comparison, disk-integrated proxies) supply no equivalent quantitative fidelity metric against AIA-like ground truth, leaving the generalization error unbounded.
  2. [Abstract] Training and evaluation description: no ablation studies, uncertainty quantification, or sensitivity tests to post-hoc choices (e.g., conditioning strength, month split, or CH mask definition) are reported; without these, it is impossible to assess how robust the reported CC=0.92/0.84 figures are to the specific experimental setup that underpins the historical applicability claim.
minor comments (3)
  1. The manuscript should report error bars or confidence intervals on the correlation coefficients and on the disk-integrated time series.
  2. Clarify the precise definition and construction of the “CH-aware” conditioning mask and whether it is derived from the HeI input or from an auxiliary map.
  3. Add a short discussion of potential calibration offsets between SOLIS and KPVT HeI data and how they might propagate into the synthetic EUV output.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback highlighting important limitations in our validation approach. We address each major comment below and indicate where revisions will be made to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract (and § on historical validation): the headline claim that CH-aware DMT supplies a usable proxy for 1974–2015 KPVT data rests on the unquantified assumption that the 2011–2015 SOLIS/AIA mapping transfers without systematic degradation; the only pixel-level CC values are from the 2015 test month, while the three historical checks (EIT morphology, SXT comparison, disk-integrated proxies) supply no equivalent quantitative fidelity metric against AIA-like ground truth, leaving the generalization error unbounded.

    Authors: We agree that no direct pixel-level quantitative metric (such as CC) can be computed for the 1974–2015 KPVT period because contemporaneous AIA 193 Å observations do not exist. The historical checks are necessarily indirect: morphological comparison to EIT 195 Å (2005–2015 overlap), visual comparison to Yohkoh/SXT, and disk-integrated emission trends versus sunspot number and F10.7. These provide supporting evidence of physical plausibility but do not bound generalization error in the same way as the 2015 test set. In revision we will (1) explicitly state this limitation in the abstract and historical-validation section, (2) add quantitative correlation statistics between the reconstructed disk-integrated EUV and the independent proxies over the full 1974–2015 interval, and (3) include a dedicated paragraph discussing the transferability assumptions and associated uncertainties. revision: partial

  2. Referee: [Abstract] Training and evaluation description: no ablation studies, uncertainty quantification, or sensitivity tests to post-hoc choices (e.g., conditioning strength, month split, or CH mask definition) are reported; without these, it is impossible to assess how robust the reported CC=0.92/0.84 figures are to the specific experimental setup that underpins the historical applicability claim.

    Authors: The original submission did not include ablation or sensitivity experiments. We will add these in revision: (a) retraining with alternative month-based splits and reporting CC variation, (b) sensitivity to CH-mask threshold and conditioning strength (guidance scale), and (c) a simple uncertainty estimate via ensemble of models with different random seeds. These results will be summarized in a new subsection and referenced from the abstract to demonstrate robustness of the reported metrics. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper trains a diffusion-based conditional image translation model (CH-aware DMT) on co-aligned SOLIS HeI 10830 Å and AIA 193 Å pairs from 2011-2015 using a month-based split, then applies the trained model to generate synthetic AIA images from historical KPVT HeI inputs (1974-2015). Model outputs are statistical predictions from the learned mapping, not algebraic reductions or self-definitions of the inputs. Historical assessments compare reconstructions to independent external data (SOHO/EIT 195 Å, Yohkoh/SXT, sunspot number, F10.7 flux) without any reported equations or fits that force the outputs by construction. No self-citations, uniqueness theorems, ansatzes, or renamings of known results are invoked as load-bearing steps in the abstract or described chain. The claim of physical plausibility for historical proxy use is an empirical generalization claim, externally checkable against the cited independent observations, and therefore self-contained against the enumerated circularity patterns.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on a trained neural network whose internal parameters are fitted to 2011-2015 data plus the domain assumption that the learned mapping generalizes across decades and instruments.

free parameters (1)
  • diffusion model weights and conditioning parameters
    All neural-network parameters are fitted during training on the 2011-2015 pairs; their specific values are not reported.
axioms (1)
  • domain assumption The statistical relationship between He I absorption and EUV emission remains sufficiently stationary to permit translation across different solar cycles and instruments.
    Invoked when the model trained on 2011-2015 data is applied to 1974-2015 historical inputs without contemporaneous EUV labels.

pith-pipeline@v0.9.1-grok · 5905 in / 1379 out tokens · 23133 ms · 2026-06-27T17:53:03.018613+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

76 extracted references · 10 canonical work pages

  1. [1]

    Dean and Thompson, B

    Pesnell, W. Dean and Thompson, B. J. and Chamberlin, P. C. , title =. Solar Physics , year =

  2. [2]

    Lemen, J. R. and Title, A. M. and Akin, D. J. and others , title =. Solar Physics , year =

  3. [3]

    Solar Physics , year =

    Boerner, Paul and Edwards, Christopher and Lemen, James and others , title =. Solar Physics , year =

  4. [4]

    and Del Zanna, G

    O'Dwyer, B. and Del Zanna, G. and Mason, H. E. and Weber, M. A. and Tripathi, D. , title =. Astronomy & Astrophysics , year =

  5. [5]

    Krieger, A. S. and Timothy, A. F. and Roelof, E. C. , title =. Solar Physics , year =

  6. [6]

    Nolte, J. T. and Krieger, A. S. and Timothy, A. F. and others , title =. Solar Physics , year =

  7. [7]

    , title =

    Cranmer, Steven R. , title =. Living Reviews in Solar Physics , year =

  8. [8]

    and Jones, H

    Andretta, V. and Jones, H. P. , title =. The Astrophysical Journal , year =

  9. [9]

    Harvey, J. W. , title =. Solar Physics , year =

  10. [10]

    Harvey, K. L. and Recely, F. , title =. Solar Physics , year =

  11. [11]

    EIT: Extreme-Ultraviolet Imaging Telescope for the

    Delaboudini. EIT: Extreme-Ultraviolet Imaging Telescope for the. Solar Physics , year =

  12. [12]

    and Acton, L

    Tsuneta, S. and Acton, L. W. and Bruner, M. E. and others , title =. Solar Physics , year =

  13. [13]

    Acton, L. W. and Weston, D. and Bruner, M. and others , title =. Solar Physics , year =

  14. [14]

    Kahler, S. W. and Davis, J. M. and Harvey, J. W. , title =. Solar Physics , year =

  15. [15]

    Caplan, R. M. and Downs, C. and Linker, J. A. , title =. The Astrophysical Journal , year =

  16. [16]

    and Asikainen, T

    Hamada, A. and Asikainen, T. and Mursula, K. , title =. Solar Physics , year =

  17. [17]

    Scholl, I. F. and Habbal, S. R. , title =. Solar Physics , year =

  18. [18]

    2020 , volume =

    The SunPy Project: Open Source Development and Status of the Version 1.0 Core Package , journal =. 2020 , volume =

    2020

  19. [19]

    Thompson, W. T. , title =. Astronomy & Astrophysics , year =

  20. [20]

    and Millman, K

    Harris, Charles R. and Millman, K. Jarrod and van der Walt, St. Array programming with. Nature , year =

  21. [21]

    IEEE Transactions on Systems, Man, and Cybernetics , year =

    Otsu, Nobuyuki , title =. IEEE Transactions on Systems, Man, and Cybernetics , year =

  22. [22]

    and Sheikh, Hamid R

    Wang, Zhou and Bovik, Alan C. and Sheikh, Hamid R. and Simoncelli, Eero P. , title =. IEEE Transactions on Image Processing , year =

  23. [23]

    New Phytologist , year =

    Jaccard, Paul , title =. New Phytologist , year =

  24. [24]

    , title =

    Dice, Lee R. , title =. Ecology , year =

  25. [25]

    A Method of Establishing Groups of Equal Amplitude in Plant Sociology Based on Similarity of Species Content , journal =

    S. A Method of Establishing Groups of Equal Amplitude in Plant Sociology Based on Similarity of Species Content , journal =. 1948 , volume =

    1948

  26. [26]

    1982 , isbn =

    Serra, Jean , title =. 1982 , isbn =

    1982

  27. [27]

    Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015 , series =

    Ronneberger, Olaf and Fischer, Philipp and Brox, Thomas , title =. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015 , series =. 2015 , doi =

    2015

  28. [28]

    2016 Fourth International Conference on 3D Vision (3DV) , year =

    Milletari, Fausto and Navab, Nassir and Ahmadi, Seyed-Ahmad , title =. 2016 Fourth International Conference on 3D Vision (3DV) , year =

    2016

  29. [29]

    and Li, Wenqi and Vercauteren, Tom and Ourselin, Sebastien and Cardoso, M

    Sudre, Carole H. and Li, Wenqi and Vercauteren, Tom and Ourselin, Sebastien and Cardoso, M. Jorge , title =. Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support , year =

  30. [30]

    Medical Image Analysis , year =

    Kervadec, Hoel and Bouchtiba, Jeremy and Desrosiers, Christian and Granger, Eric and Dolz, Jose and Ayed, Ismail Ben , title =. Medical Image Analysis , year =

  31. [31]

    Advances in Neural Information Processing Systems , year =

    Goodfellow, Ian and Pouget-Abadie, Jean and Mirza, Mehdi and Xu, Bing and Warde-Farley, David and Ozair, Sherjil and Courville, Aaron and Bengio, Yoshua , title =. Advances in Neural Information Processing Systems , year =

  32. [32]

    , title =

    Isola, Phillip and Zhu, Jun-Yan and Zhou, Tinghui and Efros, Alexei A. , title =. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , year =

  33. [33]

    Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , year =

    Wang, Ting-Chun and Liu, Ming-Yu and Zhu, Jun-Yan and Tao, Andrew and Kautz, Jan and Catanzaro, Bryan , title =. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , year =

  34. [34]

    and Maheswaranathan, Niru and Ganguli, Surya , title =

    Sohl-Dickstein, Jascha and Weiss, Eric A. and Maheswaranathan, Niru and Ganguli, Surya , title =. Proceedings of the 32nd International Conference on Machine Learning (ICML) , year =

  35. [35]

    2020 , eprint =

    Ho, Jonathan and Jain, Ajay and Abbeel, Pieter , title =. 2020 , eprint =

    2020

  36. [36]

    and Kumar, Abhishek and Ermon, Stefano and Poole, Ben , title =

    Song, Yang and Sohl-Dickstein, Jascha and Kingma, Diederik P. and Kumar, Abhishek and Ermon, Stefano and Poole, Ben , title =. International Conference on Learning Representations (ICLR) , year =

  37. [37]

    International Conference on Learning Representations (ICLR) , year =

    Song, Jiaming and Meng, Chenlin and Ermon, Stefano , title =. International Conference on Learning Representations (ICLR) , year =

  38. [38]

    Proceedings of the 38th International Conference on Machine Learning (ICML) , year =

    Nichol, Alex and Dhariwal, Prafulla , title =. Proceedings of the 38th International Conference on Machine Learning (ICML) , year =

  39. [39]

    2021 , eprint =

    Dhariwal, Prafulla and Nichol, Alex , title =. 2021 , eprint =

    2021

  40. [40]

    and Ho, Jonathan and Salimans, Tim and Fleet, David J

    Saharia, Chitwan and Chan, William and Chang, Haozhi and Lee, Chris A. and Ho, Jonathan and Salimans, Tim and Fleet, David J. and Norouzi, Mohammad , title =. 2022 , eprint =

    2022

  41. [41]

    A Diffusion Model Translator for Efficient Image-to-Image Translation , year=

    Xia, Mengfei and Zhou, Yu and Yi, Ran and Liu, Yong-Jin and Wang, Wenping , journal=. A Diffusion Model Translator for Efficient Image-to-Image Translation , year=

  42. [42]

    and McIntosh, P

    Zhang, J. and McIntosh, P. S. and others , title =. Geophysical Research Letters , year =

  43. [43]

    Bilenko, I. A. , title =. Astronomy & Astrophysics , year =

  44. [44]

    Reconstructing

    Marena, Marco and Li, Qin and Wang, Haimin and Shen, Bo , journal =. Reconstructing. 2025 , volume =

    2025

  45. [45]

    The Astrophysical Journal Supplement Series , volume =

    Jeong, Hyunjin and Moon, Yong-Jae and Park, Eunsu and Lee, Harim and Baek, Ji-Hye , title =. The Astrophysical Journal Supplement Series , volume =. 2022 , month = oct, doi =

    2022

  46. [46]

    Generation of

    Son, Jihyeon and Cha, Junghun and Moon, Yong-Jae and Lee, Harim and Park, Eunsu and Shin, Gyungin and Jeong, Hyun-Jin , journal =. Generation of. 2021 , volume =

    2021

  47. [47]

    Generation of Solar

    Park, Eunsu and Moon, Yong-Jae and Lee, Jin-Yi and Kim, Rok-Soon and Lee, Harim and Lim, Daye and Shin, Gyungin and Kim, Taeyoung , journal =. Generation of Solar. 2019 , volume =

    2019

  48. [48]

    and Cho, Il-Hyun and Choi, M

    Kim, Taeyoung and Park, Eunsu and Lee, Harim and Moon, Yong-Jae and Bae, Sung-Ho and Lim, Daye and Jang, Sungjun and Kim, L. and Cho, Il-Hyun and Choi, M. and others , journal =. Solar farside magnetograms from deep learning analysis of. 2019 , volume =

    2019

  49. [49]

    Solar Coronal Magnetic Field Extrapolation from Synchronic Data with

    Jeong, Hyun-Jin and Moon, Yong-Jae and Park, Eunsu and Lee, Harim , journal =. Solar Coronal Magnetic Field Extrapolation from Synchronic Data with. 2020 , volume =

    2020

  50. [50]

    Generation of High-resolution Solar Pseudo-magnetograms from

    Shin, Gyungin and Moon, Yong-Jae and Park, Eunsu and Jeong, Hyun-Jin and Lee, Harim and Bae, Sung-Ho , journal =. Generation of High-resolution Solar Pseudo-magnetograms from. 2020 , volume =

    2020

  51. [51]

    , title =

    Hunter, John D. , title =. Computing in Science & Engineering , year =

  52. [52]

    2000 , note =

    Bradski, Gary , title =. 2000 , note =

    2000

  53. [53]

    2026 , howpublished =

    Pillow: The friendly. 2026 , howpublished =

    2026

  54. [54]

    and Gibson, Sarah E

    Hewins, Ian M. and Gibson, Sarah E. and Webb, David F. and McFadden, Robert H. and Kuchar, Thomas A. and Emery, Barbara A. and McIntosh, Scott W. , title =. Solar Physics , year =

  55. [55]

    Annals of Data Science , volume =

    Dash, Ankan and Ye, Junyi and Wang, Guiling and Jin, Huiran , title =. Annals of Data Science , volume =. 2024 , doi =

    2024

  56. [56]

    and Veronig, A

    Jarolim, R. and Veronig, A. M. and Pötzi, W. and Podladchikova, T. , year=. A deep learning framework for instrument-to-instrument translation of solar observation data , volume=. Nature Communications , publisher=. doi:10.1038/s41467-025-58391-4 , number=

    work page doi:10.1038/s41467-025-58391-4

  57. [57]

    2015 , note =

    SILSO Sunspot Number V2.0 , howpublished =. 2015 , note =

    2015

  58. [58]

    Exploring the Limits of Synthetic Creation of Solar

    Salvatelli, Valentina and Guedes dos Santos, Luiz Fernando and Bose, Souvik and Neuberg, Brad and Cheung, Mark and Janvier, Miho and Jin, Meng and Gal, Yarin and Baydin, At. Exploring the Limits of Synthetic Creation of Solar. 2022 , journal =. doi:10.3847/1538-4357/ac867b , url =

    work page doi:10.3847/1538-4357/ac867b 2022

  59. [59]

    2021 , journal =

    Multi-Channel Auto-Calibration for the Atmospheric Imaging Assembly using Machine Learning , author =. 2021 , journal =. doi:10.1051/0004-6361/202040051 , url =

    work page doi:10.1051/0004-6361/202040051 2021

  60. [60]

    2025 , journal =

    Mitigating hallucination with non-adversarial strategies for image-to-image translation in solar physics , author =. 2025 , journal =. doi:10.1051/0004-6361/202555324 , url =

    work page doi:10.1051/0004-6361/202555324 2025

  61. [61]

    and Hackstein, S

    Ramunno, Francesco P. and Hackstein, S. and Kinakh, V. and Drozdova, M. and Quetant, G. and Csillaghy, A. and Voloshynovskiy, S. , year =. Solar synthetic imaging: Introducing denoising diffusion probabilistic models on. Astronomy & Astrophysics , volume =. doi:10.1051/0004-6361/202347860 , url =

    work page doi:10.1051/0004-6361/202347860

  62. [62]

    A Dynamic Deep-learning Model for Generating a Magnetogram Sequence from an

    Sun, Wenqing and Xu, Long and Ma, Suli and Yan, Yihua and Liu, Tie and Zhang, Weiqiang , year =. A Dynamic Deep-learning Model for Generating a Magnetogram Sequence from an. The Astrophysical Journal Supplement Series , volume =. doi:10.3847/1538-4365/ac85c0 , url =

    work page doi:10.3847/1538-4365/ac85c0

  63. [63]

    2022 , journal =

    Super-resolution of Solar Magnetograms Using Deep Learning , author =. 2022 , journal =. doi:10.1088/1674-4527/ac78ce , url =

    work page doi:10.1088/1674-4527/ac78ce 2022

  64. [64]

    2021 , journal =

    A deep learning method to estimate magnetic fields in solar active regions from photospheric continuum images , author =. 2021 , journal =. doi:10.1051/0004-6361/202140374 , url =

    work page doi:10.1051/0004-6361/202140374 2021

  65. [65]

    Can we properly determine differential emission measures from

    Youn, Junmu and Lee, Harim and Jeong, Hyun-Jin and Lee, Jin-Yi and Park, Eunsu and Moon, Yong-Jae , year =. Can we properly determine differential emission measures from. Astronomy & Astrophysics , volume =. doi:10.1051/0004-6361/202452304 , url =

    work page doi:10.1051/0004-6361/202452304

  66. [66]

    An experimental study on

    Dannehl, Markus and Delouille, V. An experimental study on. 2024 , journal =. doi:10.1029/2023EA002974 , url =

    work page doi:10.1029/2023ea002974 2024

  67. [67]

    Advances in Neural Information Processing Systems (NeurIPS) , year=

    Diffusion Models Beat GANs on Image Synthesis , author=. Advances in Neural Information Processing Systems (NeurIPS) , year=

  68. [68]

    2026 , howpublished =

    2026

  69. [69]

    The New Sunspot Number: Assembling All Corrections , journal =

    Clette, Fr. The New Sunspot Number: Assembling All Corrections , journal =. 2016 , volume =

    2016

  70. [70]

    Advances in Neural Information Processing Systems , year =

    Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and Desmaison, Alban and K. Advances in Neural Information Processing Systems , year =

  71. [71]

    Clark, Alex and Contributors , title =

  72. [72]

    Livingston, W. C. and Harvey, J. , title =. Applied Optics , volume =. 1976 , doi =

    1976

  73. [73]

    and Keller, C

    Henney, Carl J. and Keller, C. U. and Harvey, J. W. SOLIS-VSM Solar Vector Magnetograms. ASP Conf. Ser. 2006. arXiv:astro-ph/0612584

    Pith/arXiv arXiv 2006

  74. [74]

    Design Science , volume=

    DCG-GAN: design concept generation with generative adversarial networks , author=. Design Science , volume=. 2024 , publisher=

    2024

  75. [75]

    Journal of Mechanical Design , volume=

    Dde-gan: Integrating a data-driven design evaluator into generative adversarial networks for desirable and diverse concept generation , author=. Journal of Mechanical Design , volume=. 2023 , publisher=

    2023

  76. [76]

    Ieee Access , volume=

    Attribute-aware generative design with generative adversarial networks , author=. Ieee Access , volume=. 2020 , publisher=

    2020