Ruslan O. Sherstyukov

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).

This study investigates the characteristics of medium scale traveling ionospheric disturbances (MSTIDs) relative amplitudes () in the ionospheric E to lower F regions, using the EISCAT VHF radar data measured from 41 daytime runs in years 2008–2024 carried out mostly between UT. Spectral analysis is applied to identify the dominant wave amplitudes and periods, while a cross‐correlation method is used to determine the vertical wavelength, . It is observed that the dominant reveal both seasonal and height variations. The amplitudes of MSTIDs observed during winter () are approximately times larger compared to other seasons (2.5). Additionally, exhibit a general increase with height during winter, whereas during other seasons, they show mixed trends. The seasonal dependence of MSTID amplitudes may depend on the pattern of meridional wind direction and the prevailing conditions such as absence or presence of strong mesospheric wind shear for upward propagating gravity waves. The seasonal variation for the altitude behaviors may be subject to the competition between how the neutral density, ion‐neutral friction, kinematic viscosity and thermal conductivity varies with altitude. Our estimates of show a gradual/faster increase during winter/summer, corresponding to colder/hotter ionosphere. This suggests that although atmospheric gravity wave‐MSTIDs can reach higher altitudes in summer due to faster growth of , they can be less detectable due to lower amplitudes. Furthermore, waves with shorter that would reach the F2‐region during winter are more likely to decay at lower altitudes in summer.

We investigated the amplitude of medium scale traveling ionospheric disturbances (MSTIDs, with periods 25–100 min) and their dependence on the solar activity using 16 years data of the rapid run‐ionosonde operating at high latitudes (N, Sodankylä, Finland). A deep learning neural network was applied to ionograms to extract critical frequency of the F2 region (foF2) with a 1 min time resolution. Then, we analyzed the relative amplitude of MSTIDs (i.e., foF2/foF2), which corresponds to the amplitude of atmospheric gravity waves (AGWs) causing MSTIDs. The amplitude of AGWs propagating upward increases with height due to the decreasing density of the air, and hmF2 varies depending on local time, seasonal and solar activity conditions. To account for this effect, we calculated a corrected MSTID amplitude by normalizing the relative amplitude for the air density at the hmF2. The corrected amplitudes show no clear dependence on F10.7 during winter (0–12 UT), equinox (20‐01 UT) and summer (19‐01 UT), while a positive dependence of corrected amplitudes on F10.7 was observed during winter and equinox, in 14–22 UT and 15–19 UT, respectively. Corresponding to the dependence behaviors of corrected and relative amplitudes, two likely mechanisms of MSTIDs, AGWs from the lower atmosphere and auroral sources, are inferred. Their subsequent roles in the solar activity dependence of MSTID amplitudes were separately discussed, although in reality, the observed dependence is complex and often involves several mechanisms together.

Typical ionosondes operate with >5 min time intervals, which is enough to obtain regular parameters of the ionosphere, but insufficient to observe short‐term processes in the Earth's ionosphere. The key point for this study is to increase the ionosondes data time resolution by automatization of ionogram scaling routine. In this study we show the results of implementation of deep learning approach for ionogram parameters scaling. We trained the model on 13 years ionogram data set of Sodankyla ionosonde at high latitude region (67°N). We tested our autoscaling program tool on 2021 years data set and evaluate errors between operator and automatic parameters scaling. The root mean square errors for critical frequencies foF2, foF1, foE, foEs, fmin, fbEs and virtual heights h′F, h′E, h′Es are estimated as 0.12 MHz (2 pixels), 0.07 MHz (1.16 pixels), 0.15 MHz (2.5 pixels), 0.33 MHz (5.5 pixels), 0.15 MHz (2.5 pixels), 0.17 MHz (2.83 pixels), 7.7 km (1.34 pixels), 7.0 km (1.22 pixels), 7.1 km (1.24 pixels), respectively.

The potential of deep learning for the investigation of medium scale traveling ionospheric disturbances (MSTIDs) has been exploited through the Sodankylä rapid‐run ionosonde in this statistical study. The complementing observations of the Sodankylä ionosonde with those of the Sodankylä meteor radar reveals the diurnal and seasonal occurrence rate of high‐latitude MSTIDs in the recent low solar activity period, 2018–2020. In our results, the daytime, nighttime and dusk MSTIDs are predominantly identified during winter, summer, and equinoctial months, respectively. The winter daytime higher (lower) occurrence rate is well correlated with the lower (higher) altitude of the height of the F2‐layer peak (hmF2), and the low occurrence rate of the summer daytime is well correlated with the mesosphere‐lower‐thermosphere wind shear and higher gradient of temperature. Relatively high occurrence rate (>0.4) of summer nighttime MSTIDs has a general—but not one‐to‐one agreement—with post‐noon to evening IU (eastward auroral current index) inferred ionospheric conductivity. Rather, we see a one‐to‐one relationship between the summer nighttime MSTIDs and zonal wind shear suggesting that the wind shear‐induced electrodynamic processes could play significant roles for higher occurrence rate of MSTIDs. Furthermore, significant MSTIDs with ∼0.4 occurrence rate are so far revealed during spring and autumn transition periods. The enhanced nighttime MSTID amplitudes during the equinox are observed to be well correlated with IL index (westward auroral current indicator) suggesting that the particle precipitation during substorms could be the primary cause.

Powerful HF radio emission to the Earth’s ionosphere can be a reason of complex phenomena, which leads to the generation of artificial ionospheric turbulence. One of the most important components of turbulence is artificial ionospheric irregularities with a transverse dimension from a fraction of a meter to ten or more kilometers. The investigations of their spectral and dynamic characteristics makes possible to study the instabilities properties caused by the powerful radio wave and to study the features of dynamic processes in the ionosphere, and also to determine the possible influence of artificial irregularities to the different ranges radio waves propagation. Therefore, in August 5 2010, a heating experiment was carried out at the SURA heating facilities. The experiment was carried out during evening hours. Observation of generated irregularities occurred at a distance 170 km from the SURA facility at the observatory of Kazan Federal University (near Kazan) with ionosonde Cyclone. The control system of the ionosonde was set to the fast mode of ionograms recording (1 ionogram per minute), which made it possible to observe variations in the amplitude of reflections from the sporadic E-layer (Es). These variations look like synchronous with the heating time. A possible mechanism of the observed effect is discussed.

The paper deals with the problems of the mid-latitude MSTIDs detection with low intensity of electron density, ΔN/N, using the GNSS transionospheric sounding method. To collect data, a dense network of GNSS receivers located on the territory of the Russian Federation (more than 150 pcs.) has been used. On the basis of the obtained experimental data, the selectivity of the method of transionospheric sounding towards the MSTIDs, depending on the satellite-receiver line-of-sight orientation, is shown. Thus, the use of two-dimensional maps of TEC perturbations allowed to simultaneously observe the MSTIDs signatures using the satellites R03, G18 (which have similar geometric conditions of observation) and not observe them using the satellite R18. Accounting the mutual layout of the satellite-receiver lines-of-sight and phase front of the wave made it possible to form a method for observing the low-intensity midlatitude MSTID.