EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Synthetic‐Data‐Driven Supervised Learning for Pixel‐Wise Recognition of Ionogram Layer Traces

Ruslan Sherstyukov, Alexander Kozlovsky, Thomas Ulich, Samson Moges, Thomas Leyser

Journal of Geophysical Research: Machine Learning and Computation · 2026-06 · 3 (3)

DOI: 10.1029/2025jh001021 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Geophysical Research: Machine Learning and Computation, online erschienen am 23.06.2026. Verfasst von Ruslan Sherstyukov und 4 weiteren Autor:innen. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Research Council of Finland.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The significance of automatic ionogram data interpretation is growing with the increasing number of ionosondes operating at minute‐level cadence. In recent years, deep‐learning models have substantially advanced ionogram pattern recognition. However, they typically require extensive labeled ionospheric data sets for training. A key challenge is developing models for pixel‐wise identification of ionogram layer traces with sufficient generalization to maintain consistent performance across independent test sets from different years and phases of the solar cycle. Traditionally, entire trace labeling is not routine, as human resources are limited to determining key ionospheric parameters. We propose a method to generate ground truth without manual labeling and train a U‐Net with attention for robust pixel‐wise segmentation of ionogram layers at high latitudes. The model recognizes E, F1, and F2 traces and handles complex F‐spread patterns and overlapping higher‐order reflections. This capability was demonstrated using an independent test data set of ionograms. Following layer segmentation, we evaluated the mean absolute errors of key parameters: minimum frequency ( f min ), maximum frequency (fmax), and minimum virtual height ( h min ). The mean absolute errors for the E layer are f min = 0.79 pixels (0.047 MHz), fmax = 0.85 pixels (0.051 MHz), and h min = 0.54 pixels (3.10 km); for the F1 layer, f min = 0.66 pixels (0.040 MHz), f max = 0.91 pixels (0.055 MHz), and h min = 0.58 pixels (3.33 km); and for the F2 layer, f min = 1.19 pixels (0.071 MHz), f max = 0.75 pixels (0.045 MHz), and h min = 1.24 pixels (7.12 km).

Förderung

  • Research Council of Finland

Zitieren

APA

Sherstyukov, R., Kozlovsky, A., Ulich, T., Moges, S., & Leyser, T. (2026). Synthetic‐Data‐Driven Supervised Learning for Pixel‐Wise Recognition of Ionogram Layer Traces. Journal of Geophysical Research: Machine Learning and Computation, 3(3). https://doi.org/10.1029/2025jh001021

Vancouver

Sherstyukov R, Kozlovsky A, Ulich T, Moges S, Leyser T. Synthetic‐Data‐Driven Supervised Learning for Pixel‐Wise Recognition of Ionogram Layer Traces. Journal of Geophysical Research: Machine Learning and Computation. 2026;3(3). doi:10.1029/2025jh001021

BibTeX
@article{Sherstyukov2026,
  author = {Sherstyukov, Ruslan and Kozlovsky, Alexander and Ulich, Thomas and Moges, Samson and Leyser, Thomas},
  title = {Synthetic‐Data‐Driven Supervised Learning for Pixel‐Wise Recognition of Ionogram Layer Traces},
  journal = {Journal of Geophysical Research: Machine Learning and Computation},
  year = {2026},
  volume = {3},
  number = {3},
  doi = {10.1029/2025jh001021}
}
Kostenlos starten →