Abstract Equatorial plasma bubbles (EPBs) are typically nighttime ionospheric irregularities, and their persistence into daytime sectors remains insufficiently understood. Here, we investigate long‐lived EPB‐related plasma depletion structures over the Pacific sector during the recovery phase of the 23–25 April 2023 geomagnetic storm using Formosa Satellite‐7/Constellation Observing System for Meteorology, Ionosphere, and Climate‐2 (FORMOSAT‐7/COSMIC‐2) in situ ion density, Swarm‐B electron density, ground‐based Global Navigation Satellite System (GNSS) rate of total electron content index (ROTI), COSMIC‐2 Global Ionospheric Specification (GIS) electron density, and Thermosphere Ionosphere Mesosphere Energetics and Dynamics/Global Ultraviolet Imager (TIMED/GUVI) atomic oxygen‐to‐molecular nitrogen column density ratio (O/N 2 ) observations. COSMIC‐2 observations show that depletion signatures increased on 24 April, persisted from post‐sunrise through local noon, and remained identifiable in the afternoon during 0–6 Universal Time (UT) on 25 April, corresponding to an observable persistence interval of about 20 hr. Consecutive along‐track density profiles reveal that the structures gradually weakened from clear depletions into faint residual signatures while remaining within similar longitude ranges. Ground‐based ROTI observations provide independent evidence for sustained EPB‐related irregularities from the postmidnight sector to local noon at the Northern Hemisphere stations and a weaker but longer‐lasting response extending into the afternoon at the American Samoa Power Authority (ASPA) station in the Southern Hemisphere. Swarm‐B independently detected a morning super plasma depletion extending to ∼35° magnetic latitude in the Southern Hemisphere, corresponding to an apex altitude exceeding ∼3,500 km. COSMIC‐2 GIS data show enhanced postmidnight low‐latitude topside electron density followed by reduced low‐to‐middle latitude background density after sunrise, while GUVI O/N 2 observations indicate a hemispherically asymmetric thermospheric composition response. These observations suggest that the daytime signatures were weak residual traces of storm‐time nighttime/postmidnight EPBs rather than newly generated daytime irregularities. High‐apex field‐aligned depletion structure and reduced post‐sunrise background plasma density likely slowed refilling, enabling residual EPB‐related structures to persist into unusually late local‐time sectors.
Abstract Global Ionospheric Specification (GIS) three‐dimensional electron density profiles constructed by assimilating the line‐of‐sight total electron content (TEC) from global navigation satellite system (GNSS) radio occultation (RO) measurements by FORMOSAT‐7/COSMIC‐2 (F7/C2) and ground‐based receiver network from international GNSS service, constitute a unique ionospheric database with 1‐hr cadence. Using the data spanning over half a solar cycle since the release of GIS as an operational Level 3 data product of F7/C2 mission, a comprehensive evaluation of the peak electron density (NmF 2 ) is carried out by identifying data conjunctions with Global Ionosphere Radio Observatory (GIRO) Digisonde observations. The results reveal an overall correlation of 0.93, with GIS overestimating GIRO NmF 2 by 6.4%. While the values do not show any specific dependence on geomagnetic activity, the correlation (bias) reduces (increases) when only high latitude (>35°) observations are included where no RO measurements are assimilated. A similar tendency is observed when only low solar activity years (F10.7 ≤ 100) are considered. The detailed local time dependence of the correlation and bias are examined in this study, with lowest correlation and large relative bias occurring during 0400 LT—0600 LT. The results show large local time variation of bias depending on latitude region and solar activity. Overall, the bias ranges from 10% to 30% during 1000 LT—2300 LT, whereas GIS may underestimate NmF 2 by 5%–15% during 0700 LT—1000 LT. The results also demonstrate that GIS reveals identical day‐to‐day variations as seen in GIRO measurements and compares well with TECs from Global Ionosphere Map and ground GNSS receivers.
Abstract This study presents a novel multitask deep learning framework that integrates models designed for simultaneous, multi‐horizon forecasting of Hp60 and Disturbance storm time (Dst) indices with lead times up to 3 hr. Using a Long Short‐Term Memory architecture, the models capture both mid‐latitude variability and the ring current dynamics. Input features include interplanetary magnetic field (IMF) parameters, solar wind plasma data, and historical geomagnetic indices. Model interpretability is addressed through permutation‐based feature importance analysis, providing physical context to the machine learning predictions. The models demonstrate strong predictive performance across all forecast horizons. For the Dst index, they achieve a Root Mean Square Error (RMSE) of 3.81 nT and Pearson correlation coefficient (R) of 0.98 at 1‐hr horizon (t + 1 hr), remaining robust (RMSE = 9.01 nT, R = 0.91) at 3‐hr horizon (t + 3 hr). More importantly, the models perform well under both quiet and disturbed geomagnetic conditions. However, while they accurately capture storm onset and recovery phases, they underestimate extreme storm intensities. Hp60 predictability is primarily associated with related mid‐latitude indices, such as planetary K (Kp) and ap, that are derived from similar observatory networks and respond to the same underlying geomagnetic disturbances. Dst forecasting is largely influenced by its own historical values (persistence). Southward IMF Bz is also significantly influential for forecasting both indices during disturbed periods, showing that the models learn physically consistent storm‐time behavior. The multitask approach offers a scalable and efficient pathway for future space weather models by leveraging shared representations of common solar wind drivers.
Abstract The period from 1 to 14 May 2024 was marked by some of the most intense solar activity of Solar Cycle 25, including multiple X‐class flares, fast coronal mass ejections (CMEs), and diverse solar radio emissions. We present observations from the Mexican e‐Callisto network (REC‐Mx), including long‐duration continuum emissions (CTMs), Type II bursts, and Type III bursts associated with major eruptive events. Three prolonged continuum events observed on 8–10 May 2024 were analyzed using polarization measurements from the MEXICO‐LANCE station. The degree of circular polarization showed similar behavior across 47–87 MHz, indicating stable emission conditions over the sampled coronal heights. Assuming fundamental plasma emission, the inferred coronal magnetic‐field strengths ranged from 1 to 16 G. The derived magnetic‐field decay indices cluster around n≈3–4, implying a radial dependence of B(r)∝r−n with little sensitivity to the adopted coronal density model. In addition, REC‐Mx detected eight metric Type II bursts produced by CME‐driven shocks. Quantitative analysis of the Type II burst drift rates provided shock‐speed estimates for representative events. Numerous Type III bursts tracing the escape of energetic electrons along open magnetic field lines were also observed. Several events exhibited concurrent Type II and Type III activity, indicating simultaneous shock‐related and flare‐related particle acceleration. The inferred magnetic‐field strengths and decay indices agree with previous studies of the metric corona. These observations demonstrate the scientific and operational value of REC‐Mx within the global e‐Callisto network by providing near‐real‐time diagnostics of coronal magnetic conditions, particle acceleration, and CME‐driven shocks relevant to space‐weather monitoring.
Abstract Large ionospheric disturbances were observed over Europe during the extreme geomagnetic storm on 10 May 2024. This study analyzes GNSS‐derived total electron content (TEC) and all‐sky imager data to investigate these disturbances. Pronounced TEC enhancements, irregularities, and strong gradients were confined to the auroral region above the mid‐latitudes. These anomalies evolved coherently with the equatorward expansion of the optical aurora. Large‐scale TEC disturbances were found to be moving at mid‐latitudes with an equatorward velocity of 726 m/s and an amplitude of up to 7 TECU. These characteristics were indicative of the auroral expansion rather than of large‐scale traveling ionospheric disturbances (LSTIDs), which exhibited weaker amplitudes near the auroral boundary. This finding contrasts with the interpretation of some previous studies on GNSS observations. This study demonstrates the visualization of the auroral oval at mid‐latitudes using TEC and TEC‐derived parameters.
Abstract The Mother's Day (Gannon) geomagnetic superstorm of May 2024 produced extreme thermospheric heating and expansion, resulting in enhanced atmospheric drag on low Earth orbiting (LEO) satellites. We investigate storm‐time thermospheric drag and density variability using three satellites of the Small‐scale magNetospheric and Ionospheric Plasma Experiment (SNIPE) 6U CubeSat constellation, complemented by density measurements from Swarm, GRACE‐FO, and MSIS 2.1. Despite operating at nearly identical orbits, SNIPE‐B and SNIPE‐D experienced significantly different altitude losses during the storm, attributed to differences in effective along‐track cross‐sectional area arising from different attitude control modes—demonstrating that spacecraft configuration strongly influences drag even for identical CubeSat platforms. SNIPE‐B's drag‐derived density shows strong correlation with GRACE‐FO measurements, whereas MSIS 2.1 underestimates density near storm peaks. However, the inter‐altitude drag ratio between SNIPE‐A and SNIPE‐B agrees qualitatively with MSIS 2.1 predictions, suggesting the model captures thermospheric vertical structure under extreme force. These results demonstrate that publicly available two‐line element data, when combined with attitude information, provide a cost‐effective method for monitoring storm‐time thermospheric density. Our findings highlight the importance of attitude information in drag analyses and suggest that CubeSat constellations offer significant potential for thermospheric modeling and space weather monitoring.
Abstract Geomagnetic substorms are fundamental processes of explosive geospace energy release within the Earth's magnetotail, which are usually divided into three phases: the growth phase, the expansion phase, and the recovery phase. Large‐scale electromagnetic energy transport during a geomagnetic substorm is in the form of Alfvén waves, transmitting energy from the distant magnetotail to the low‐altitude ionosphere‐thermosphere system. This explosive Alfvénic energy release process plays an important role in space weather events. However, such important process has not been investigated quantitatively in global geospace models for space weather forecasting. In this study, we use coupled global simulations to investigate the dynamic evolution of Alfvénic Poynting flux during an idealized substorm‐steady magnetospheric convection (SMC) cycle. Results show that during the expansion phase, the hemispheric Alfvénic Poynting flux is enhanced by approximately 200%, and the dawn‐dusk asymmetry of the Alfvénic oval is diminished significantly. During the recovery phase and in the SMC state, the spatial distribution of downward low‐altitude Alfvénic Poynting flux exhibits a significant dawn‐dusk asymmetry. The explosive behavior of the simulated Alfvénic power is consistent with the observed enhancement in Alfvénic power and broadband power during substorms, suggesting that the global geospace model is not only capable of reproducing the time scale of the Alfvénic variation but also the magnitude of the power enhancement during substorm‐SMC cycles.
Abstract Auroral precipitation plays a key role in coupling the Magnetosphere–Ionosphere–Thermosphere system by modifying ionospheric conductance, currents, atmospheric heating, and upper‐atmospheric dynamics. Leveraging 17 years of electron spectrum data from the Defense Meteorological Satellite Program (DMSP), we developed a Machine Learning‐based Electron Precipitation Model (ML‐EPM) that nowcasts differential energy fluxes of precipitating electrons in 19 energy channels ranging from 30 eV to 30 keV, using 6‐hr histories of solar wind (SW), interplanetary magnetic field (IMF) and geomagnetic indices as inputs. The model performance was evaluated using a 2‐year DMSP data set (2013–2014) and a geomagnetic storm event on 27–28 February 2014, both of which were not included in training. ML‐EPM achieved moderate‐to‐strong correlation coefficients (0.5–0.72) for 16 out of 19 channels while maintaining root‐mean‐square and mean‐absolute errors within the same order of magnitude. Although the model underestimated total energy flux compared to OVATION and DMSP observations, it reproduces key auroral features, including (a) auroral oval broadening, equatorward shifting of auroral boundaries, and enhanced precipitation during disturbed periods, (b) low‐energy precipitation (<1 keV) such as soft electrons in the cusp, polar rain in the polar cap, and secondary electrons generated by primary auroral electrons, and (c) various auroral spectral shapes, including diffuse, monoenergetic, and broadband auroras. Rather than relying on total energy flux and mean energy to estimate an idealized Maxwellian shape, ML‐EPM directly nowcasts electron fluxes across 19 distinct energy channels, allowing global circulation models to more accurately resolve altitude‐dependent ionization rates and predict ionosphere‐thermosphere dynamics.
Abstract The availability and accuracy of upstream solar wind conditions are critical for solar wind—magnetosphere coupling studies, whether based on observations or numerical models and simulations. We present the Merged Interplanetary Data from L1 (MIDL), which merges 1‐min magnetic field and plasma observations from ACE, DSCOVR, and Wind through a three‐stage algorithm: per‐satellite filtering and despiking, multi‐satellite quality screening with agreement‐first source selection, and using the median of the selected sources for each variable group. Special care is taken to minimize sudden changes in the data when sources change. The merged data can be propagated to a user set distance from Earth including Moon's orbit. MIDL provides two propagation methods: the traditional ballistic approach with proper handling of shocks and one‐dimensional magnetohydrodynamic simulation. Comparison with OMNI over 2005–2025 shows good agreement (Pearson r=0.99 for Vx, 0.96 for density, 0.86 for Bz) with no systematic bias. MIDL achieves near‐complete coverage (∼100%) for the magnetic field and coverage exceeding 95% for the plasma parameters for all years, compared to OMNI's ∼91% and ∼73%, respectively. The multi‐satellite consensus reduces vulnerability to data gaps, instrumental artifacts, and propagation errors. The data set is publicly available through a web interface https://csem.engin.umich.edu/MIDL and also through a public Python package CSEM‐MIDL. MIDL is well‐suited for use as simulation boundary conditions and for statistical solar wind studies.
Abstract This study uses observation data from multi‐constellation (GPS/GLONASS/Galileo/BeiDou) ground‐based Global Navigation Satellite Systems (GNSS) networks and a non‐integrated spherical harmonic function modeling methodology to develop an ionospheric Total Electron Content (TEC) model with high spatiotemporal resolution (1° × 1° in longitude and latitude, and 15 min in time) over China and adjacent regions. The response of regional ionospheric TEC during the May 2024 super geomagnetic storm was then investigated using this model. The results indicate that the model can accurately capture the spatiotemporal variation characteristics of regional ionospheric TEC during this geomagnetic storm. During the main and early recovery phases of the geomagnetic storm, positive and negative ionospheric disturbances occurred simultaneously. During the late recovery phases of the storm, long‐lasting negative ionospheric disturbances dominated, causing a substantial TEC decrease, with a maximum decrease of about 70 TECU and a relative decrease of up to 80%. The daytime ionospheric TEC on 12 May was almost comparable to the typical night‐time value. Additionally, ionospheric response exhibited east‐west differences. Furthermore, during certain periods of the superstorm's recovery phase, positive ionospheric disturbances occurred against an overall negative storm background at middle and low latitudes; this non‐locally generated disturbance spans 20°N–40°N with its center near 30°N (projected radius ∼10°), propagates westward at ∼140 ± 8 m/s, and shows no obvious spatial contraction but a gradual intensity weakening with differential total electron content decreasing from ∼30 TECU to 2∼3 TECU. These phenomena indicated that the ionosphere over China and adjacent regions experienced complex variations during this super geomagnetic storm.
Abstract Most predictions of space climate, that is, the long‐term behavior of the solar‐terrestrial environment, have focused on forecasting the 11‐year sunspot cycle. Geomagnetic activity, on the other hand, has mainly been predicted in shorter, space weather timescales of up to days to weeks. Using a 180‐year composite aa index, we aim to predict here the temporal behavior of geomagnetic activity over 16 last solar cycles. By identifying activity peaks in the ascending and declining phases of the cycle and an activity minimum between the two, we represent each aa cycle with two triangular peaks. The large‐scale features of the aa cycle depicted by the model are related to changes in the occurrence of coronal mass ejections and high‐speed solar wind streams which drive geomagnetic activity. Using past aa and sunspot observations, as well as a recent sunspot cycle prediction model, we find interesting relationships for the predictability of the aa peak amplitudes and timings, which suggest intrinsic differences between even and odd cycles and give strong support to the so‐called Gnevyshev‐Ohl rule ordering of cycles to even‐odd cycle pairs. Finally, we attempt to hindcast each past aa cycle, including the ongoing cycle 25, while also estimating the prediction uncertainty using a leave‐one‐out cross‐validation methodology. Prediction of a new cycle is made at the time of aa minimum at the start of the cycle, which occurs typically a few months after the sunspot minimum.
Abstract The European Space Agency's Swarm mission has provided over a decade of continuous in situ measurements of the topside ionospheric plasma. With the release of various official plasma density products and the development of several calibration methods, selecting the most reliable data set has become an important issue for the scientific community. This study presents a comprehensive validation of Swarm plasma density products, including the standard Langmuir Probes (LP) harmonic and sweep modes, empirically corrected LP versions, physics‐based LP calibration—developed within SLIDEM (Swarm LP Ion Drift and Effective Mass) project –, neural network LP calibration (NNcor), and Faceplate (FP) observations, against ground‐truth observations from Jicamarca, Arecibo, and Millstone Hill incoherent scatter radars. By comparing both the statistical climatology and individual measurements at conjunctions, we demonstrate that the standard LP product generally underestimates plasma density by approximately 10%–20%, but with a consistent overestimation on the nightside at low solar activity. While empirical corrections reduce the mean bias, they do not significantly improve the dispersion of errors. Conversely, the neural network‐based NNcor data set, which calibrated LP observations to Swarm FP data, exhibits the best overall performance, achieving the highest correlation and minimizing residuals across different conditions. The physics‐based SLIDEM LP product also proves to be a robust alternative. We conclude that calibrated data sets, particularly NNcor and SLIDEM, are essential for accurate ionospheric modeling and should be preferred over standard products to avoid propagating instrumental biases into empirical models like the International Reference Ionosphere.
Abstract The internal charging effects triggered by energetic (>100 keV) electrons in the Earth's radiation belts can cause anomalies and failures of electronic components aboard spacecraft. Given the critical importance of safeguarding space assets, it is essential to estimate internal charging risks, which depend highly on the distribution of radiation belt electron fluxes. In this study, we reconstruct the distribution of electron fluxes using a 3‐D data assimilative model, which integrates a radiation belt numerical model with multi‐satellite observations using the Ensemble Kalman Filter. We then estimate the spatial and temporal evolution of internal charging currents and risks across the entire outer radiation belt under specific shielding/dielectric layers and configurations. Furthermore, we analyze the impacts of shielding thickness on charging currents for different characteristic energy spectra. Our results reveal that for both power‐law and exponential energy spectra, the charging current decreases monotonically with increasing shielding thickness. However, there exists a “critical shielding thickness” for the bump‐on‐tail (BOT) energy spectrum. Below this critical value, the charging current increases with shielding thickness, whereas beyond it the current decreases. For the BOT energy spectrum, both higher electron energies corresponding to peak flux and thinner dielectric thicknesses can result in an increase in the “critical shielding thickness.” Our results shed important light on the prediction of satellite internal charging risk under the exposure to Earth's highly dynamic outer radiation belt.
Abstract Defining a standardized ionospheric storm scale is challenging given the ionosphere's inherent complexity and susceptibility to multiple driving factors. In this study, a novel ionospheric storm scale (ISS) index is developed based on statistical analysis of total electron content (TEC) data from 257 ground‐based GNSS stations across China between 2008 and 2021. Designed for broad applicability, the index is independent of season, local time, and geographic location. We first derive an ionospheric activity index in percentage by comparing observed TEC values with a quiet‐time reference. To account for seasonal, diurnal, and latitudinal variations in TEC fluctuations, each percentage deviation is normalized using the robust Z‐score method. The ISS is then established by defining thresholds for the normalized data, categorizing activity into seven levels: ISS0, ISS P 1, ISS P 2, ISS P 3, ISS N 1, ISS N 2, and ISS N 3. Results reveal that the new index exhibits no diurnal or seasonal dependence, a key requirement for a homogeneous ionospheric activity time series. Correlation analysis between the ionospheric activity index and the residuals of the ionospheric TEC model we developed previously, along with variations in the index during geomagnetic storms, demonstrates that the ISS can serve as a reliable indicator of the expected performance of ionospheric models and well represents ionospheric activity. Finally, analysis of the relationship between standard point positioning accuracy and the ionospheric activity index under different solar activity levels reveals that higher ionospheric activity corresponds to increased positioning errors during periods of high solar activity.
Abstract This paper quantitatively characterizes d B /d t peaks at four African equatorial‐ and low‐latitude stations, along latitudinal and local‐time gradients across different phases of selected geomagnetic storms. We investigate d B /d t response for intense geomagnetic storms, Dst ≤ −100 nT observed during solar cycle 24 over four stations in the African equatorial to low‐latitude region, namely, Addis Ababa (ADIS, mag lat: 0.18°) in Ethiopia, Abuja (ABJA, mag lat: 0.55°) in Nigeria, Yaounde (YAOD, mag lat: −5.30°) in Cameroon, and Medea (MEDE, mag lat: 27.98°) in Algeria. Furthermore, we determine the equatorial electrojet (EEJ) current using differential magnetometer technique over Ethiopia. Results show that the d B /d t magnitudes and timings varied significantly with latitude. The station within the EEJ zone, ADIS exhibited large sharp d B /d t spikes aligned with the EEJ enhancements. The stations outside the zone (YAOD and MEDE) recorded lower values of d B /d t during the same period, which is consistent with prior studies. Additionally, fluctuations in d B /d t recurred during the recovery phase, likely influenced by ionospheric currents, substorm activity, and local ionospheric conductivity. These findings highlight the need to re‐evaluate the temporal structure of geomagnetically induced current (GIC) risk in equatorial regions, as hazardous d B /d t . Enhanced d B /d t values are associated with rapid changes in the solar wind conditions rather than their peak amplitudes. Higher d B /d t variation does not automatically indicate higher GICs, cautioning GIC practitioners to always isolate natural enhancers of high d B /d t before GIC risk assessments. Furthermore, this research lays the foundation for the future development of necessary GIC mitigation strategies for Africa.
Abstract In this study, we employ a deep learning approach to detect auroras from near‐infrared (NIR) sequential images captured by the satellite‐based imager, the Enhanced Polar Outflow Probe (e‐POP)/Fast Auroral Imager (FAI). We also apply the Eigen‐Class Activation Map (Eigen‐CAM), an explainable AI technique in a broad sense, as a post‐hoc interpretability method to indicate the location of auroral emissions. Accurately identifying the location of auroras is crucial for understanding space weather dynamics. However, it is challenging to distinguish auroral emissions from static features such as clouds, mountains, and city lights in single‐channel NIR images. To overcome this problem, we employ a CNN‐based ResNeXt‐50 deep learning model to automatically detect auroras and discriminate them from non‐auroral features in e‐POP/FAI images. Through comparative experiments on input configurations, we find that the best‐performing model utilizes a sequence of three frames with temporal gaps (at a two‐second cadence). This temporal context allows the model to learn dynamic characteristics and effectively filter out static non‐auroral features. Eigen‐CAM is then applied to visualize the image regions that contribute most to the model's decision, illustrating its potential to highlight auroral structures without supervision. For our study, we use images from 2015 to 2017, with January and July as the test set. Our best model demonstrates high performance, achieving an accuracy of 0.84 and an F1‐score of 0.84. Ultimately, our approach facilitates large‐scale statistical studies of auroral dynamics by replacing manual classification, which also leads to better predictions about changes in Earth's upper atmosphere.
Abstract This study quantitatively assesses the drivers of the Equatorial Ionization Anomaly (EIA) morphology and latitudinal shifts over Southeast Asia during the December solstice. We use a 27‐year ionospheric data set (1998–2024) of total electron content (TEC) and F2‐layer peak electron density (NmF2) under geomagnetically quiet conditions. Crest parameters are extracted using a dual‐Gaussian fitting method, and variance contributions are decomposed from solar and geographic factors through multiple linear regression. Geographic control, arising from asymmetric solar illumination relative to the geomagnetic equator, shows a statistically significant association with EIA morphology during periods of low solar activity (F10.7 < 100). The near 10° offset between the geographic and geomagnetic equators corresponds to southward subsolar migration, which coincides with illumination asymmetry and correlates with coherent southward EIA shifts. Southern crests displace about 9.0° southward (TEC) and 7.1° southward (NmF2) from the June to December solstice, nearly 1.8–3 times the displacement of the northern crests. Under low solar activity (F10.7 < 100), the subsolar latitude (geographic control) accounts for 98%–99% of the explained variance in southern crest displacement, whereas solar EUV flux (F10.7) contributes ≤2%. Adjusted R 2 values (0.33–0.48) further indicate that 52%–67% of the total variance remains unexplained, a discrepancy likely attributable to E × B drifts, neutral winds, thermospheric influences, and electrodynamic coupling processes. These findings highlight the statistically significant role of geographic control in shaping EIA asymmetry under weak solar forcing and provide quantitative constraints for improving regional ionospheric models and space weather forecasts in Southeast Asia.
Abstract This study investigates the co‐evolution of geomagnetic field disturbances with ionospheric parameters and their characteristic latitudinal signatures by conducting a location and time‐dependent analysis of the 07–12 October 2024 events, to fill crucial gaps in understanding the coupled response of the magnetosphere‐ionosphere system. This has been carried out utilizing integrated multi‐instrument approach combining ground‐based magnetometer‐derived geomagnetic fields deviations from the mean (ΔBx, ΔBy, ΔBz), Digisonde‐derived critical frequency and maximum electron density at the F2‐ionospheric layer (foF2, hmF2), and Global Navigation Satellite System Total Electron Content (GNSS TEC) measurements. Significant disturbances in the ΔBx component are observed across all sectors in the low‐latitude region, particularly when the storm intensity increases; while in the mid‐latitude region, all magnetic field components exhibit significant disturbances across American stations. Mid‐ and low‐latitude stations displayed more stable but still disturbed patterns in both storm cases, with phase dependent and sector‐specific ionospheric responses. During the initial phase, prompt penetration electric fields increased TEC in low‐latitude regions, particularly during evening hours (19:00–23:00 UT) when the equatorial ionization anomaly was strongest. The American sector experienced the most pronounced and coincide geomagnetic field and ionospheric perturbations during both storms, showing the largest ΔBx and TEC enhancements relative to quiet‐time maxima. Notably, the main phase of the 10–12 October 2024 storm resulted in the largest depletion of TEC and foF2, measured against quiet‐time minima. Larger TEC enhancements also coincides with substantial hmF2 depression and foF2 reduction than higher hmF2 and foF2 when the intensity of the storm enhances.
Abstract As a critical parameter characterizing the vertical structure of the ionosphere, the peak height of the F2 layer (hmF2) plays an essential role in frequency selection and signal propagation in high‐frequency (HF) communication systems. Accurate prediction of hmF2 is particularly important in high‐latitude regions. Based on observations from nine ionospheric stations located above 60°N, we analyze hmF2 variations and evaluate the prediction performance of four models: three IRI‐2020 sub‐models (IRI‐BSE, IRI‐AMTB, and IRI‐SHU) and the Empirical Canadian High Arctic Ionospheric Model (E‐CHAIM). The results show that: (a) hmF2 increases during high solar activity years and decreases during low solar activity years, exhibiting distinct diurnal and seasonal variations. (b) Among the four models, E‐CHAIM performs the best. Compared with the Bilitza–Sheikh–Eyfrig model (BSE), Altadill–Magdaleno–Torta–Blanch model (AMTB), and Shubin models (SHU), E‐CHAIM reduces the mean absolute error by 10.18, 11.85, and 1.77 km, and the root mean square error by 11.82, 12.61, and 1.98 km, respectively. (c) The performance of BSE and AMTB improves with increasing solar activity intensity, and the BSE model shows a more pronounced advantage during high solar activity years. It should be noted that the evaluation is conducted using monthly median hmF2 values, meaning the results primarily reflect the long‐term climatological performance of these models, while short‐term disturbances (e.g., geomagnetic activity) are largely suppressed. These findings deepen the understanding of long‐term variations in the high‐latitude ionosphere and lay a foundation for the long‐term planning of communication systems.
Abstract Geomagnetic disturbances (GMDs) are rapid changes in the magnetic field of the Earth that may drive geomagnetically induced currents (GICs), known to cause damage to infrastructure such as power grids and pipelines. Globally, GIC measurements are often hard to obtain; therefore, GMDs are used as a proxy for GICs. GMDs have been previously shown to occur on different timescales which are related to different formation mechanisms such as magnetic storms, substorm activities and sudden commencements. 1 min cadence ground‐magnetometer data has been available for decades and are the basis of many studies showing the frequency and intensity of GMDs. In recent years one‐second cadence data have become available from a network of magnetometers through the SuperMAG portal. We take the 1‐s data, averaged 10 s data, and compare the number of GMDs identified with the number found from the 1‐minute data to test the GMDs related to the different timescales. We find that 1‐s data consistently record higher amplitude GMDs as well as recording overall more GMDs, thus showing the importance of 1‐s cadence data. This is most important for GMDs between 00 and 03 magnetic local time and at low geomagnetic latitudes (>50°) as these are the locations where difference between the number of GMD detections between the 1‐s data and the 1‐min data is the largest. However, the 1‐s data have many artefacts due to unreliable data points in them leading a 10‐s median being more reliable at correctly detecting GMDs.
