Abstract
Research on the ionospheric Spread-F (SF) phenomenon holds significant value in both fields as ionospheric electrodynamic research and enhanced operational applications in radio-based technologies (e.g., communication and navigation). To date, the classification of Spread-F remains largely reliant on the manual interpretation of ionograms by experts, suffering from inefficiency (~10 seconds per figure) and subjectivity. There has been no publicly available ionogram dataset classifying Frequency/Range/Mix/Strong Range SF (FSF/RSF/MSF/SSF) by either human labor or machine processing. To address this problem, we introduce the first open, expert-guided ionogram dataset that is simultaneously the most comprehensive in terms of class coverage, the largest in volume, and the most extensive in temporal span. This collection encompasses 150,000 ionograms (30,000 per class, including a “non-SF” group) spanning 14 years from 2002 to 2016, thereby capturing a diverse range of solar and geomagnetic conditions. The attached classification SA-ResNet50 model based on this dataset could be applied to further data.
Similar content being viewed by others
Data availability
All ionograms analyzed in this study were generated by the digital ionosonde deployed at the Hainan station of the Chinese Meridian Project. The dataset, HA419-IonogramSet-202505, has been publicly released and is available for download from the Science Data Bank at https://doi.org/10.57760/sciencedb.26826.
Code availability
All codes for this project were implemented in Python 3.8.13 using the PyTorch framework (version 2.1.0). The complete codes have been made publicly available at: “https://github.com/CCaijh/Hainan_classification”. The repository includes:
• Training and inference scripts for the baseline model;
• Classification results of the baseline model on the 2016 test set (saved in TXT format);
• All necessary Python scripts for reproducing our experiments.
• The shuffle sequence we used during the training process
To ensure bit-wise reproducibility, we fixed every random generator to the same value (seed = 2026) before dataset shuffling and model initialization. In addition, we recorded the exact per-epoch shuffle permutations generated by the DataLoader; these index sequences (“.\train_order.txt”) are released with our code repository, enabling readers to replay the identical mini-batch order used in all reported runs. However, it is worth noting that differences in the operating system version or language settings may affect the generation of random numbers, potentially leading to final training outcomes that differ from those we have provided.
It should be noted that the pretrained weights provided were derived from ionogram data collected by the digital ionosonde at the Hainan station. Previous statistical studies19,20,21,22,23,24,25,26 have demonstrated that, at low geomagnetic latitudes (±20°), the occurrence types of Spread-F and their corresponding ionogram characteristics are highly consistent across Asia, Africa, and the Americas. Consequently, the pretrained weights are applicable to regions within ±20° geomagnetic latitude, corresponding to the Equatorial Ionization Anomaly zone. For Spread-F classification tasks based on ionograms collected at stations outside this latitude range, we recommend retraining or fine-tuning the shared model weights on a reasonably sized set of manually labeled ionograms, following a workflow similar to that proposed in this study.
References
Fejer, B. G., Scherliess, L. & De Paula, E. R. Effects of the vertical plasma drift velocity on the generation and evolution of equatorial spread F. J. Geophys. Res. Space Phys. 104, 19859–19869 (1999).
Abdu, M. A. Day-to-day and short-term variabilities in the equatorial plasma bubble/spread F irregularity seeding and development. Prog. Earth Planet. Sci. 6 (2019).
Olugbon, B. et al. Daytime Equatorial Spread F-Like Irregularities Detected by HF Doppler Receiver and Digisonde. Space Weather 19, e2020SW002676 (2021).
Piggott, W. R. & Rawer, K. URSI Handbook of Ionogram Interpretation and Reduction. 68–73 (1972).
Yasyukevich, Y. et al. Small-Scale Ionospheric Irregularities of Auroral Origin at Mid-latitudes during the 22 June 2015 Magnetic Storm and Their Effect on GPS Positioning. Remote Sens. 12, 1579 (2020).
Booker, H. G. & Wells, H. W. Scattering of radio waves by the F-region of the ionosphere. Terr. Magn. Atmospheric Electr. 43, 249–256 (1938).
Carter, B. A. et al. An analysis of the quiet time day‐to‐day variability in the formation of postsunset equatorial plasma bubbles in the Southeast Asian region. J. Geophys. Res. Space Phys. 119, 3206–3223 (2014).
Klenzing, J. et al. A system science perspective of the drivers of equatorial plasma bubbles. Front. Astron. Space Sci. 9 (2023).
Mendillo, M., Meriwether, J. & Biondi, M. Testing the thermospheric neutral wind suppression mechanism for day‐to‐day variability of equatorial spread. F. J. Geophys. Res. Space Phys. 106, 3655–3663 (2001).
S Saito & T Maruyama. Large-Scale Zonal Structure of the Equatorial Ionosphere and Equatorial Spread F Occur- rence. J. Natl. Inst. Inf. Commun. Technol. 56, Nos. 1-4 (Special Issue: Space Weather Forecast), 267–276 (2009).
Maruyama, T. & Matuura, N. Longitudinal variability of annual changes in activity of equatorial spread F and plasma bubbles. J. Geophys. Res. Space Phys. 89, 10903–10912 (1984).
Basu, S. et al. Day‐to‐day variability of the equatorial ionization anomaly and scintillations at dusk observed by GUVI and modeling by SAMI3. J. Geophys. Res. Space Phys. 114 (2009).
Tsunoda, R. T. On the enigma of day-to-day variability in equatorial spread F. Geophys. Res. Lett. 32 (2005).
Takahashi, H. et al. Multi-instrument study of longitudinal wave structures for plasma bubble seeding in the equatorial ionosphere. Earth Planet. Phys. 5, 1–10 (2021).
Kelley, M. C., Larsen, M. F., LaHoz, C. & McClure, J. P. Gravity wave initiation of equatorial spread F: A case study. J. Geophys. Res. Space Phys. 86, 9087–9100 (1981).
Abdu, M. A. et al. Gravity wave initiation of equatorial spread F/plasma bubble irregularities based on observational data from the SpreadFEx campaign. Ann. Geophys. 27, 2607–2622 (2009).
Abdu, M. A. Electrodynamics of ionospheric weather over low latitudes. Geosci. Lett. 3 (2016).
Chandra, H. & Rastogi, R. G. Spread-F at magnetic equatorial station Thumba. Ann Geophys 28, 37–44 (1972).
Shi, J. K. et al. Relationship between strong range spread F and ionospheric scintillations observed in Hainan from 2003 to 2007. J. Geophys. Res. Space Phys. 116 (2011).
Wang, G. J., Shi, J. K., Wang, X. & Shang, S. P. Seasonal variation of spread-F observed in Hainan. Adv. Space Res. 41, 639–644 (2008).
Wang, G. J. et al. The statistical properties of spread F observed at Hainan station during the declining period of the 23rd solar cycle. Ann. Geophys. 28, 1263–1271 (2010).
Jiang, C. et al. Ionosonde observations of daytime spread F at low latitudes. J. Geophys. Res. Space Phys. 121 (2016).
Alfonsi, L. et al. Comparative analysis of spread‐F signature and GPS scintillation occurrences at Tucumán, Argentina. J. Geophys. Res. Space Phys. 118, 4483–4502 (2013).
González, G. D. L. Spread-F characteristics over Tucumán near the southern anomaly crest in South America during the descending phase of solar cycle 24. Adv. Space Res. 69, 1281–1300 (2022).
Sahai, Y. et al. Generation of large-scale equatorial F-region plasma depletions during lowrange spread-F season. Ann. Geophys. 22, 15–23 (2004).
Wang, Z. et al. Strong range SF observed in low latitude ionosphere over acsension IS in Atlantic ocean. in 2017 Progress In Electromagnetics Research Symposium - Spring (PIERS) 1879–1883, https://doi.org/10.1109/piers.2017.8262056 (IEEE, St. Petersburg, 2017).
Wang, Z., Shi, J. K., Torkar, K., Wang, G. J. & Wang, X. Correlation between ionospheric strong range spread F and scintillations observed in Vanimo station. J. Geophys. Res. Space Phys. 119, 8578–8585 (2014).
Ye, H. et al. Multi-Instrumental Observations of Midlatitude Plasma Irregularities over Eastern Asia during a Moderate Magnetic Storm on 16 July 2003. Remote Sens. 15, 1160 (2023).
Wang, N. et al. Longitudinal differences in the statistical characteristics of ionospheric spread-F occurrences at midlatitude in Eastern Asia. Earth Planets Space 71, 47 (2019).
Xie, H. et al. Statistical Study on the Occurrences of Postsunset Ionospheric E, Valley, and F Region Irregularities and Their Correlations Over Low-Latitude Sanya. J. Geophys. Res. Space Phys. 123, 9873–9880 (2018).
Cueva, R. Y. C., De Paula, E. R. & Kherani, A. E. Statistical analysis of radar observed F region irregularities from three longitudinal sectors. Ann. Geophys. 31, 2137–2146 (2013).
Thammavongsy, P., Phakphisut, W. & Supnithi, P. A Preliminary Neural Network Model for Range Spread-F Events at Chumphon Station, Thailand. in 2018 18th International Symposium on Communications and Information Technologies (ISCIT) 395–398, https://doi.org/10.1109/iscit.2018.8588013 (IEEE, Bangkok, 2018).
Thammavongsy, P., Supnithi, P., Phakphisut, W., Hozumi, K. & Tsugawa, T. Spread-F prediction model for the equatorial Chumphon station, Thailand. Adv. Space Res. 65, 152–162 (2020).
Feng, J., Wu, A. & Zheng, W. S. Shape-erased feature learning for visible-infrared person re-identification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition 22752–22761 (2023).
Bhaneja, P., Klenzing, J., Pacheco, E. E., Earle, G. D. & Bullett, T. W. Statistical analysis of low latitude spread F at the American, Atlantic, and Pacific sectors using digisonde observations. Front. Astron. Space Sci. 11 (2024).
Lan, T., Zhang, Y., Jiang, C., Yang, G. & Zhao, Z. Automatic identification of Spread F using decision trees. J. Atmospheric Sol.-Terr. Phys. 179, 389–395 (2018).
Lan, T., Hu, H., Jiang, C., Yang, G. & Zhao, Z. A comparative study of decision tree, random forest, and convolutional neural network for spread-F identification. Adv. Space Res. 65, 2052–2061 (2020).
Luwanga, C., Fang, T., Chandran, A. & Lee, Y. Automatic Spread‐F Detection Using Deep Learning. Radio Sci. 57 (2022).
Venkateswara Rao, T., Sridhar, M. & Venkata Ratnam, D. Auto-detection of sporadic E and spread F events from the digital ionograms. Adv. Space Res. 70, 1142–1152 (2022).
Wang, Z. et al. Automatic Detection and Classification of Spread‐F From Ionosonde at Hainan With Image‐Based Deep Learning Method. Space Weather 21 (2023).
Benchawattananon, P. et al. Automatic classification of spread‐F types in ionogram images using support vector machine and convolutional neural network. Earth Planets Space 76, 56 (2024).
He, K., Zhang, X., Ren, S. & Sun, J. Deep Residual Learning for Image Recognition. in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 770–778, https://doi.org/10.1109/cvpr.2016.90 (IEEE, Las Vegas, NV, USA, 2016).
Tan, M. & Le, Q. V. EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.
Dosovitskiy, A. et al. An image is worth 16 × 16 words: transformers for image recognition at scale. In International Conference on Learning Representations (2021).
Gao, P. et al. HA419-IonogramSet-202505. Science Data Bank (ScienceDB) https://doi.org/10.57760/sciencedb.26826.
CBAM: Convolutional Block Attention Module. in Lecture Notes in Computer Science 3–19, https://doi.org/10.1007/978-3-030-01234-2_1 (Springer International Publishing, Cham, 2018).
Zhao, X. et al. The Prediction of Day‐to‐Day Occurrence of Low Latitude Ionospheric Strong Scintillation Using Gradient Boosting Algorithm. Space Weather 19 (2021).
Acknowledgements
This research was supported by the Public Computing Cloud of CUC and the following grants: Project of Stable Support for Youth Team in Basic Research Field, CAS (YSBR-018); Youth Innovation Promotion Association CAS (2023000116); The Strategic Priority Research Program of the Chinese Academy of Sciences, Grant No. XDA0470301; Specialized Research Fund for State Key Laboratories; Pandeng Program of National Space Science Center, Chinese Academy of Sciences; the High-quality and Cutting-edge Disciplines Construction Project for Universities in Beijing (Internet Information, Communication University of China); the Fundamental Research Funds for the Central Universities. Mostly, we would like to acknowledge the use of data from the Chinese Meridian Project.
Author information
Authors and Affiliations
Contributions
Conceptualization: P.G., Z.W., Data Collection: M.Z.H., X.W., K.D., Data curation: Q.Z., Q.Q., Z.W., G.W., Y.Z.H., J.S., Data proofreading: Q.Z., M.Z.H., Z.W., G.W., Y.Z.H., Software: P.G., J.C., Q.Q., C.Q., B.W., Experiments: Q.Z., J.C., Z.W., G.W., Visualization: Q.Z., C.Q., B.W., Manuscript writing: P.G., J.C., Z.W.
Corresponding author
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
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Gao, P., Zhu, Q., Cai, J. et al. A Deep Learning-Enabled Ionogram Dataset for Detection and Classification of Low-latitude Spread-F Phenomena. Sci Data (2026). https://doi.org/10.1038/s41597-025-06493-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41597-025-06493-5
