Sebastijan Mrak

The high‐latitude ionosphere exhibits rapid spatial and temporal variability during geomagnetically active periods, creating persistent challenges for accurate nowcasts. The High‐latitude Workbox for the Ionosphere (HAWK‐I) is a real‐time ionosphere nowcast and specification system designed for operational deployment to address these challenges. HAWK‐I employs the Ionospheric Data Assimilation Four‐Dimensional (IDA4D) framework, implemented as a three‐dimensional variational (3DVar) algorithm, to update a log‐electron‐density state and produce three‐dimensional electron density specifications from GNSS and radio occultation observations at 5‐min cadence. Performance is evaluated during the May 2024 geomagnetic storm using ground‐based GNSS Total Electron Content and incoherent scatter radar measurements. Results demonstrate substantial reductions in TEC error and improved representation of storm‐time ionospheric structure relative to background models, with complementary strengths of empirical and physics‐based models. Assimilation of Radio Occulation (RO) observations further improves performance in regions distant from ground‐based receivers, highlighting their importance for high‐latitude ionospheric specification.

Solar eclipses provide unique opportunities to investigate geospace dynamics driven by rapid, localized reductions in solar radiation. On December , 2021, a total solar eclipse occurred over West Antarctica during substorms. While eclipse effects on Earth's upper atmosphere are widely studied, their subsequent impact on the magnetospheric configuration and feedback loops are less explored. Abrupt changes in conductivity can also affect interhemispheric asymmetry (IHA). In this study, an eclipse EUV mask was applied in the Global Ionosphere‐Thermosphere Model (GITM), driven by different high‐latitude forcing specifications. Using the Space Weather Modeling Framework, we examined how eclipse‐induced ionospheric conductance changes impact the magnetosphere and feed back to the I‐T system. Results show a −∼30% reduction in total electron content and −∼35% drop in ionospheric conductance compared to the no‐eclipse case, consistent with observations. This leads to an enhanced cross‐polar cap potential (CPCP) in both hemispheres (north vs. south: 36 vs. 40 kV). The eclipse reduced IHA in total field‐aligned currents and CPCP, bringing asymmetry close to zero after the shadow passed over the polar region. The magnetosphere also showed dynamic responses, including changes in plasma pressure and flow that shifted the magnetotail by up to 16 in the X ‐direction. Finally, by driving GITM with eclipse‐modified magnetospheric forcing, we found that magnetospheric feedback amplified thermospheric wind perturbations over the polar cap, even exceeding the direct effect of reduced solar EUV. These results demonstrate the critical role of ionospheric conductance in the coupled M‐I‐T system during solar eclipses.

This study investigates the impact of high‐latitude ionospheric irregularities on global positioning system (GPS) position accuracy during the 10–11 May 2024 geomagnetic storm. Using data from ground‐based GPS receivers and an all‐sky camera in Antarctica, we examined evolution of GPS vertical total electron content, phase fluctuations, and precise point positioning (PPP) errors. Results reveal that auroral precipitation associated with the coronal mass ejection shock induced significant GPS phase fluctuations and positioning errors in the dayside cusp region. Tongues of ionization (TOIs) had the largest and most sustained impact on signal disruptions and degraded position accuracy, with a peak position error reaching 28 m. Furthermore, dawnside auroral activity during the substorms and poleward boundary intensifications (PBIs) caused localized disturbances in the auroral oval. Despite sufficient satellite visibility, these irregularities significantly compromised GPS positioning performance.

We analyze an episode of strong mountain wave (MW) activity over the western US from 9 to 12 January 2017 using the HIgh Altitude mechanistic General Circulation Model. We find that medium‐scale MWs were generated by strong eastward flow over the Sierra Nevada and the Rocky Mountains. During this time, part of the stratospheric polar vortex jet extended from the western US to eastern Canada such that the MWs propagated into the lower mesosphere where they dissipated from westward vertical wind shear. This resulted in secondary gravity waves (GWs) that propagated into the lower thermosphere where tertiary GWs having concentric ring structures were created. With increasing altitude in the thermosphere, certain propagation directions were highlighted as a result of the dissipation induced by the tidal winds. At 260 km, we find eastward propagation during local morning over the northeastern US, equatorward propagation around local noon over the southern US, westward propagation during local afternoon over the northwestern US, and poleward propagation over Canada after local midnight. In addition, the model shows equatorward propagating larger‐scale GWs over Canada from remote sources around local noon. The simulated regional GW‐mean flow interaction patterns are consistent with multi‐step vertical coupling triggered by the MWs. The traveling ionospheric disturbances (TIDs) during the MW event are simulated with the ionospheric model SAMI3. The simulated GWs and TIDs are consistent with the medium‐to‐large‐scale TIDs observed over the continental US in GPS TEC data.

A total solar eclipse (TSE) traversed the continental US (CONUS) from west to east on 21 August 2017. Ionosondes located under the eclipse totality at Lusk (Wyoming) and Boulder (Colorado) observed the ionospheric G‐condition 20 min after totality. The Millstone Hill mid‐latitude incoherent scatter radar recorded an anomalous low altitude F₂ peak during the recovery phase of the eclipse, which can be attributed to an ionospheric G‐condition. We perform WACCM‐X simulations to investigate the physical processes that drive the ionospheric G‐condition. Specifically, we conducted a diagnostic analysis of the simulated atomic oxygen ion continuity equation to examine the source of the G‐condition. We defined two metrics describing the G‐condition: (a) the duration and (b) the maximum difference between and . Results indicate that E‐ and F₁ plasma density reduction closely follow the eclipse obscuration, whereas the F₂‐layer density depletion lags the obscuration by 20 min. This delay increases with altitude and is caused by slower radiative recombination and transport processes in the diffusion‐dominated F₂‐region. The delay creates a period during the eclipse recovery when the F₁‐plasma density exceeds that of the F₂‐peak, which manifests as the ionospheric G‐condition. The simulation study illuminates the space‐time behavior of the G‐condition indicating that this state can last for more than 50 min during the recovery phase of the eclipse.

Solar eclipses provide unique opportunities to investigate geospace dynamics driven by rapid, localized reductions in solar radiation. On December 4th, 2021, a total solar eclipse occurred over West Antarctica during substorms. While eclipse effects on Earth’s upper atmosphere are well studied, their influence on the coupled magnetosphere-ionosphere-thermosphere (M-I-T) system remains less understood. Abrupt changes in conductivity can also affect interhemispheric asymmetry (IHA). In this study, an eclipse EUV mask was applied in the Global Ionosphere-Thermosphere Model (GITM), driven by different high-latitude forcing specifications. Using the Space Weather Modeling Framework, we examined how eclipse-induced ionospheric conductance changes impact the magnetosphere and feed back to the I-T system. Results show a ~30% reduction in total electron content (TEC) and ~35% drop in ionospheric conductance compared to the no-eclipse case, consistent with TEC observations. This leads to an enhanced cross-polar cap potential (CPCP) in both hemispheres (north vs. south: 36 kV vs. 40 kV). The eclipse reduced IHA in total FACs and CPCP, bringing asymmetry close to zero after the shadow passed over the polar region. The magnetosphere also showed dynamic responses, including changes in plasma pressure and flow that shifted the magnetotail by up to 16 RE in the X-direction. Finally, by driving GITM with eclipse-modified magnetospheric forcing, we found that magnetospheric feedback amplified thermospheric winds perturbations over the polar cap, even exceeding the direct effect of reduced solar EUV. These results demonstrate the critical role of ionospheric conductance in the coupled M-I-T system during solar eclipses.

This study characterizes the main ionospheric trough (MIT) using a newly implemented detection method applied to ground‐based Global Navigation Satellite System data. The MIT is a region of plasma depletion occurring primarily in the nighttime sub‐auroral F‐region ionosphere. Analysis is based on ground‐based ionosphere total electron content (TEC) measurements from 2012 to 2024 and is applied to both hemispheres. The data are sorted by geomagnetic condition and season. We characterize MIT dynamics and compare the results with previous studies. Detection algorithm limitations, hemispheric asymmetry, trough depth, boundary wall steepness and position are statistically quantified and visualized. Main conclusions include: (a) Automatic trough detection is highest during geomagnetically active winter in the northern hemisphere (NH). (b) This detection method creates synoptic views of the trough which we can use to demonstrate control of sub‐auroral polarization streams (SAPS) over the dusk/afternoon sector and influence of storm onset on the MIT. (c) There is a noticeable morning preference for the southern hemisphere (SH) trough. (d) The dawn‐side SH trough appears equatorward relative to the NH, potentially due to influence from polar convection patterns. The dusk‐side NH trough appears slightly equatorward of the SH trough as a response to SAPS. (e) The deepest trough occurs during dawn hours and demonstrates more consistent longitudinal patterns during quiet local winter. (f) The steepest trough boundary is at the poleward wall with a positive gradient at 12–15 local time in NH summer. Synoptic maps illustrate asymmetries in the trough structure and the influence of density plumes.

This study investigates impacts of the May 2024 superstorm on the mid‐latitude Global Positioning System (GPS) scintillation and position errors. Using 1‐Hz GPS receiver data, we identified position errors in PPP mode reaching up to 70 m in the Central United States during the storm main phase on May 10. The PPK solution becomes unstable following the arrival of storm and lasted till the recovery phase, coinciding with reported GPS outages of farming equipment. The large position errors were attributed to strong scintillation and carrier phase cycle slips around the equatorward boundary of the ionosphere trough, where large total electron content (TEC) gradients and irregularities were present. In the Southwestern United States, position errors of 10–20 m were associated with the storm‐enhanced density and equatorial ionization anomaly. Scintillation and cycle slips in this region were minor, and bending of the GPS signal paths (refractive effect) is suggested to cause the position errors. PPP outages were also associated with sudden changes in the geometric distributions of available GPS satellites used in position calculations. On May 11, energetic particle precipitation during substorms led to abrupt jumps in TEC and scintillation, resulting in rapidly evolving position errors of up to 10 m. These findings highlight the critical role of storm‐time plasma transport, precipitation, and irregularity formation in degrading GPS performance. The study underscores the need for accurate ionospheric state specification, improved signal processing technique, real‐time ionospheric corrections, and optimized satellite selection algorithms, to enhance navigation resilience during severe space weather events.

We discuss multiple observations of atmospheric dynamics during recent solar eclipses: 21 August 2017, 4 December 2021, 20 April 2023, 14 October 2023, and 8 April 2024. We use GOLD and TIMED/GUVI instruments to observe the changes in the lower thermosphere related to atmospheric oxygen O, and molecular nitrogen N2. We discuss the problematic nature of O/N2 composition estimation from brightnesses under solar obscuration conditions, and how the observations should be properly interpreted. The Ionosphere-Thermosphere (I-T) system responds to solar X-Ray and Extreme Ultra Violet (EUV) radiation, which is highly non-uniform during solar eclipses. We introduce a new model of solar eclipse obscuration and irradiance by combining the solar occultation software PyEclipse and Flare Irradiance Solar Model 2 (FISM2) -- FISM2-Eclipse. We implemented FISM2-Eclipse into the Global Ionosphere thermosphere Model (GITM) and Whole Atmosphere Model Ionosphere-Plasmasphere-Electrodynamics model (WAM-IPE) to conduct a model-data comparison. We discuss some salient features noted during the recent eclipses.

Two solar eclipse events in 2023 appeared to produce considerable enhancements in the thermospheric column density ratio of monatomic oxygen to molecular nitrogen (ΣO/N₂) as measured by TIMED GUVI. We quantify potential sources for eclipse‐induced ΣO/N₂ changes and find that the observed enhancements arise from the ionospheric O⁺ radiative recombination contribution to the OI 135.6 nm emission from which ΣO/N₂ is derived. Variations in the solar Extreme Ultra Violet (EUV) and X‐ray spectrum, due to the difference between the disk spectrum and the coronal spectrum, are also considered but shown to have negligible contributions to the ΣO/N₂ enhancements. After accounting for the radiative recombination contribution, we constrain the real thermospheric compositional change to the uncertainty level of the measurements of 5%–10%. These results are valuable for the interpretation of eclipse‐induced ΣO/N₂ changes that will further first‐principle model comparisons and lead to a better understanding of the response of the thermosphere to localized variations in solar EUV and X‐ray forcing.