A multi-instrument study is conducted at the dayside polar ionosphere to investigate the spatio-temporal evolution of scintillation in Global Navigation Satellite System (GNSS) signals during non-storm conditions. Bursts of intense amplitude and phase scintillation started to occur at similar to ${\sim} $ 9 MLT and persisted for more than 1 hour implying the simultaneous existence of Fresnel and large-scale sized irregularities of significant strength in the pre-noon sector. Measurements from the EISCAT radar in Svalbard (ESR) revealed the presence of dense plasma structures with significant gradients in regions of strong Joule heating/fast flows and soft precipitation when scintillation was enhanced. Plasma structuring down to Fresnel scales were observed both in the auroral oval as well as inside the polar cap with the associated amplitude scintillation exhibiting similar strengths regardless of whether the density structures were in regions of active auroral dynamics or not. The observations are placed within the context of different sources of free energy, providing insights into the important mechanisms that generate irregularities capable of perturbing GNSS signal properties in the dayside ionosphere. Furthermore, a strong negative excursion in the interplanetary magnetic field (IMF) By ${B}_{y}$ component during the northward turning of Bz ${B}_{z}$ led to the transport of a depleted region of plasma density into the post-noon sector that significantly weakened both amplitude and phase scintillation.
AbstractOne of the most intense geomagnetic storms of recent times occurred on 10–11 May 2024. With a peak negative excursion of Sym‐H below −500 nT, this storm is the second largest of the space era. Solar wind energy transferred through radiation and mass coupling affected the entire Geospace. Our study revealed that the dayside magnetopause was compressed below the geostationary orbit (6.6 RE) for continuously ∼6 hr due to strong Solar Wind Dynamic Pressure (SWDP). Tremendous compression pushed the bow‐shock also to below the geostationary orbit for a few minutes. Magnetohydrodynamic models suggest that the magnetopause location could be as low as 3.3RE. We show that a unique combination of high SWDP (≥15 nPa) with an intense eastward interplanetary electric field (IEFY ≥ 2.5 mV/m) within a super‐dense Interplanetary Coronal Mass Ejection lasted for 409 min–is the key factor that led to the strong ring current at much closer to the Earth causing such an intense storm. Severe electrodynamic disturbances led to a strong positive ionospheric storm with more than 100% increase in dayside ionospheric Total Electron Content (TEC), affecting GPS positioning/navigation. Further, an HF radio blackout was found to occur in the 2–12 MHz frequency band due to strong D‐ and E‐region ionization resulting from a solar flare prior to this storm.
AbstractThis study investigates the Brazilian low‐latitude ionospheric response to CIR/HSS‐driven geomagnetic storms during the declining phase of solar cycle 24, from 2016 to 2017. In this period the geomagnetic storms were mostly moderate, SymHmin ≈ −72 nT, AEmax ≈ 1580 nT, Vswmax ≈ 690 km/s and lasted, on average, for 6 days. We analyze the variations in Vertical Total Electron Content (VTEC) at three representative regions: bele, over the equatorial region; boav and cuib, at the northern and southern crests of the Equatorial Ionization Anomaly. Our findings reveal the role of High‐Speed Solar Wind Streams and Corotating Interaction Region—driven geomagnetic storms. The VTEC intensifications were up to 30 TECu, during the daytime and nighttime. Additionally, three categories of nighttime enhancements were observed and analyzed with distinct characteristics and levels of pre‐reversal strengthening; Depletions up to 20 TECu also occurred during the day and nighttime. The delay between the storm commencement and the positive and negative variations were, on average, 7 and 20 hours, respectively. We discuss the Prompt Penetration Electric Fields and Disturbance Dynamo Electric Fields following the magnetic reconnection between Earth's and interplanetary magnetic field, using observational data and modeling. Furthermore, this study presents catalogs of low‐latitude ionospheric storms, providing detailed information for space weather applications and ionospheric modeling.
Abstract During the second recovery phase of the 13–14 March 2022 storm, intense high‐latitude neutral mass density spikes are detected by satellites at ∼500 km in both hemispheres. These density spikes, accurately modeled by the Global Ionospheric Thermosphere Model (GITM), are identified as high‐latitude neutral mass density anomalies (HDAs). The GITM simulation indicates that these HDAs, which extends over the polar region, are important features in high‐latitude neutral density. Furthermore, GITM reveals that these HDAs are manifestations of transpolar traveling atmospheric disturbances triggered on the dawn side. Moreover, GITM also reveals significant interhemispheric asymmetries (IHAs) in the magnitude, propagation speed, and propagation direction of HDAs, which are linked to the IHAs in the distribution and magnitude of Joule heating deposited as well as the thermospheric background conditions. This study presents a dynamic perspective on the IHA of storm‐time high‐latitude neutral density variations that is particularly helpful to the proper interpretation of satellite observations.
Abstract In this study, we evaluate a coronal mass ejection (CME) arrival prediction tool that utilizes the wide‐angle observations made by STEREO's heliospheric imagers (HI). The unsurpassable advantage of these imagers is the possibility to observe the evolution and propagation of a CME from close to the Sun out to 1 AU and beyond. We believe that by exploiting this capability, instead of relying on coronagraph observations only, it is possible to improve today's CME arrival time predictions. The ELlipse Evolution model based on HI observations (ELEvoHI) assumes that the CME frontal shape within the ecliptic plane is an ellipse and allows the CME to adjust to the ambient solar wind speed; that is, it is drag based. ELEvoHI is used to perform ensemble simulations by varying the CME frontal shape within given boundary conditions that are consistent with the observations made by HI. In this work, we evaluate different setups of the model by performing hindcasts for 15 well‐defined isolated CMEs that occurred when STEREO was near L4/5, between the end of 2008 and the beginning of 2011. In this way, we find a mean absolute error of between 6.2 ± 7.9 and 9.9 ± 13 hr depending on the model setup used. ELEvoHI is specified for using data from future space weather missions carrying HIs located at L5 or L1. It can also be used with near‐real‐time STEREO‐A HI beacon data to provide CME arrival predictions during the next ∼7 years when STEREO‐A is observing the Sun‐Earth space.
In this paper we present several methods to identify precursors that show great promise for early predictions of solar flare events. A data preprocessing pipeline is built to extract useful data from multiple sources, Geostationary Operational Environmental Satellites and Solar Dynamics Observatory (SDO)/Helioseismic and Magnetic Imager (HMI), to prepare inputs for machine learning algorithms. Two classification models are presented: classification of flares from quiet times for active regions and classification of strong versus weak flare events. We adopt deep learning algorithms to capture both spatial and temporal information from HMI magnetogram data. Effective feature extraction and feature selection with raw magnetogram data using deep learning and statistical algorithms enable us to train classification models to achieve almost as good performance as using active region parameters provided in HMI/Space-Weather HMI-Active Region Patch (SHARP) data files. Case studies show a significant increase in the prediction score around 20 hr before strong solar flare events.
An analysis of noise attenuation during 80 solar flares between 2013 and 2017 was carried out at frequencies 8–20 MHz using 34 Super Dual Auroral Radar Network radars and the EKB ISTP SB RAS radar. The attenuation was determined on the basis of noise measurements performed by the radars during the intervals between transmitting periods. The location of the primary contributing ground sources of noise was found by consideration of the propagation paths of radar backscatter from the ground. The elevation angle for the ground echoes was determined through a new empirical model. It was used to determine the paths of the noise and the location of its source. The method was particularly well suited for daytime situations, which had to be limited for the most part to only two crossings through the D region. Knowing the radio path was used to determine an equivalent vertical propagation attenuation factor. The change in the noise during solar flares was correlated with solar radiation lines measured by GOES/XRS, GOES/EUVS, SDO/AIA, SDO/EVE, SOHO/SEM, and PROBA2/LYRA instruments. Radiation in the 1 to 8 Å and near 100 Å are shown to be primarily responsible for the increase in the radionoise absorption, and by inference, for an increase in the D and E region density. The data are also shown to be consistent with a radar frequency dependence having a power law with an exponent of −1.6. This study shows that a new data set can be made available to study D and E regions.
The Community Coordinated Modeling Center has been leading community-wide space science and space weather model validation projects for many years. These efforts have been broadened and extended via the newly launched International Forum for Space Weather Modeling Capabilities Assessment (https://ccmc.gsfc.nasa.gov/assessment/). Its objective is to track space weather models' progress and performance over time, a capability that is critically needed in space weather operations and different user communities in general. The Space Radiation and Plasma Effects Working Team of the aforementioned International Forum works on one of the many focused evaluation topics and deals with five different subtopics (https://ccmc.gsfc.nasa.gov/assessment/topics/radiation-all.php) and varieties of particle populations: Surface Charging from tens of eV to 50-keV electrons and internal charging due to energetic electrons from hundreds keV to several MeVs. Single-event effects from solar energetic particles and galactic cosmic rays (several MeV to TeV), total dose due to accumulation of doses from electrons (>100 keV) and protons (>1 MeV) in a broad energy range, and radiation effects from solar energetic particles and galactic cosmic rays at aviation altitudes. A unique aspect of the Space Radiation and Plasma Effects focus area is that it bridges the space environments, engineering, and user communities. The intent of the paper is to provide an overview of the current status and to suggest a guide for how to best validate space environment models for operational/engineering use, which includes selection of essential space environment and effect quantities and appropriate metrics.
Coupled Sun-to-Earth models represent a key part of the future development of space weather forecasting. With respect to predicting the state of the thermosphere and ionosphere, there has been a recent paradigm shift; it is now clear that any self-respecting model of this region needs to include some representation of forcing from the lower atmosphere, as well as solar and geomagnetic forcing. Here we assess existing modeling capability and set out a road map for the important next steps needed to ensure further advances. These steps include a model verification strategy, analysis of the impact of nonhydrostatic dynamical cores, and a cost-benefit analysis of model chemistry for weather and climate applications.
On 13 March 1989, the largest magnetic storm of the last century caused widespread effects on power systems including a blackout of the Hydro‐Québec system. Since then this event has become the archetypal disturbance for examining the geomagnetic hazard to power systems. However, even 30 years on from 1989, the story of exactly what happened in March 1989 is far from complete. This paper reexamines the information available about the March 1989 event and uses this to construct a timeline and description of the space weather phenomena and how they caused the power system effects. The evidence shows that the disturbance was caused by two coronal mass ejections (CMEs): the first associated with a X4.5 flare on 10 March and the second linked to a M7.3 flare on 12 March. The arrival of the interplanetary CME shock fronts caused storm sudden commencements at 01.27 and 07.43 UT on 13 March. The transit time and speed of the first (second) interplanetary CME shock are 54.5 hr (31.5 hr) and 760 km/s (1,320 km/s). Empirical relations are used to estimate solar wind speed and southward interplanetary magnetic field, Bs , and give values of v = 980 km/s, Bs = 40 to 60 nT at the peak of the storm. Key findings are that the second storm sudden commencement occurred at the same time as the substorm that impacted the Hydro‐Québec system and indicates that external triggering of the substorm may have contributed to a faster substorm onset than might otherwise have occurred. This caused the production of larger geomagnetically induced currents that caused the Hydro‐Québec blackout. The March 1989 storm had the largest recorded value of the Dst index representing the size of the magnetic storm main phase, but the Hydro‐Québec blackout occurred early in the storm when the Dst value was less disturbed. Only later in the storm did Dst reach its peak value. At this time an expansion of the auroral oval brought disturbances to lower latitudes where they caused power system problems in the United States, United Kingdom, and Sweden.
In this study, different hybridized techniques that combine an artificial neural network (ANN) with bio-inspired optimization algorithms such as particle swarm optimization (PSO), genetic algorithm (GA), and a hybridized PSO+GA were applied to update the ANN connection weights and so forecast the disturbance storm time (Dst) index. The past values of Dst index time series were used as input parameters to forecast its variation from 1 to 6 hours ahead. The database collected considers 233,760 hourly data from 01 January 1990 to 31 August 2016, containing storms and quiet period, grouped into three data sets: learning set (116,880 hourly data points), validation set (58,440 data points), and testing set (58,440 data points). Several ANN configurations were studied and optimized during the training process by evaluating the root mean square error (RMSE) and the correlation coefficient (R). An analysis of the predictive capability of each method was made year per year, and according to the levels of the geomagnetic storm. Also, an additional test was applied to the proposed method using 17 intense geomagnetic storms reported during solar cycle 24, including the St. Patrick's Day storm of 2015. Results show that the hybridized ANN+PSO method can forecast the Dst index quite accurately from 1 to 3 h in advance (with RMSE <= 5 nT and R >= 0.9), while the ANN+PSO+GA method can forecast the Dst index quite accurately from 4 to 6 h ahead (RMSE <= 7 nT and R >= 0.8)
The capability to forecast conditions in the mesosphere and lower thermosphere is investigated based on 30-day hindcast experiments that were initialized bimonthly during 2009 and 2010. The hindcasts were performed using the Whole Atmosphere Community Climate Model with thermosphere-ionosphere eXtension (WACCMX) with data assimilation provided by the Data Assimilation Research Testbed (DART) ensemble Kalman filter. Analysis of the WACCMX+DART hindcasts reveals several important features that are relevant to forecasting the middle atmosphere. The results show a clear dependence on spatial scale, with the slowest error growth occurring in the zonal mean and the fastest error growth occurring for small-scale waves. The error growth rate is also found to be significantly greater in the upper mesosphere and lower thermosphere compared to in the upper stratosphere to lower mesosphere, suggesting that the forecast skill decreases with increasing altitude. The results demonstrate that the errors in the lower thermosphere reach saturation, on average, in less than 5 days, at least with the current version of WACCMX+DART. A seasonal dependency to the error growth is found at high latitudes in the Northern and Southern Hemispheres but not in the tropics or global average. We additionally investigate the error growth rates for migrating and nonmigrating atmospheric tides and find that the errors saturate after similar to 5 days for tides in the lower thermosphere. The results provide an initial assessment of the error growth rates in the mesosphere and lower thermosphere and are relevant for understanding how whole atmosphere models can potentially improve space weather forecasting.
The Kp index is a measure of the midlatitude global geomagnetic activity and represents short-term magnetic variations driven by solar wind plasma and interplanetary magnetic field. The Kp index is one of the most widely used indicators for space weather alerts and serves as input to various models, such as for the thermosphere and the radiation belts. It is therefore crucial to predict the Kp index accurately. Previous work in this area has mostly employed artificial neural networks to nowcast Kp, based their inferences on the recent history of Kp and on solar wind measurements at L1. In this study, we systematically test how different machine learning techniques perform on the task of nowcasting and forecasting Kp for prediction horizons of up to 12 hr. Additionally, we investigate different methods of machine learning and information theory for selecting the optimal inputs to a predictive model. We illustrate how these methods can be applied to select the most important inputs to a predictive model of Kp and to significantly reduce input dimensionality. We compare our best performing models based on a reduced set of optimal inputs with the existing models of Kp, using different test intervals, and show how this selection can affect model performance.
Mass density variations can deviate from the expected behavior caused by temperature due to changes in the composition. Such deviations can be especially significant during solar minimum conditions. Model-data differences are typically resolved through temperature corrections while overlooking the role of errors in lower boundary composition. In this work, we use a data-driven methodology to simultaneously estimate thermosphere composition and temperature contributions to model-data differences. The methodology uses modal decomposition to extract high-dimensional, reduced order basis functions for the covariance of the neutral thermospheric species and temperature. The extracted basis functions are combined with CHAllenging Minisatellite Payload and Gravity Recovery And Climate Experiment mass density measurements using a nonlinear least squares solver. We demonstrate the methodology using the Naval Research Laboratory's empirical Mass Spectrometer and Incoherent Scatter (MSIS) model to derive the high-dimensional basis functions. We characterize and quantify the contribution of temperature and lower boundary effects with oxygen and helium since the two species have a direct impact on drag and orbit prediction through gas-surface interactions and mass density. We analyze the month of December in 2008, based on the work of Thayer et al. (2012), and estimate that lower boundary composition errors contribute approximately 50% of the model-data differences.
The Sun was remarkably active during the first week of September 2017 producing numerous solar flares, solar radiation storms, and coronal mass ejections. This activity caused disruption to terrestrial high‐frequency (HF, 3–30 MHz) radio communication channels including observations with the Super Dual Auroral Radar Network (SuperDARN) HF radars. In this paper, we analyze the response of SuperDARN groundscatter observations and decreases in background sky noise level in response to multiple solar flares occurring in quick succession and co‐occurring with solar energetic protons and auroral activity. We estimate the attenuation in HF signal strength using an approach similar to riometry and find that the radars exhibit a nonlinear response to compound solar flare events. Additionally, we find that the three different space weather drivers have varying degrees of influence on the HF signal properties at different latitudes. Our study demonstrates that in addition to monitoring high‐latitude convection, SuperDARN observations can be used to study the spatiotemporal evolution of disruption to HF communication during extreme space weather conditions.
We review, summarize, and comment on the 2019 National Space Weather Strategy and Action Plan.
We present evidence that variability in the STEREO‐A Heliospheric Imager (HI) data is correlated with in situ solar wind speed estimates from WIND, STEREO‐A, and STEREO‐B. For 2008–2012, we compute the variability in HI differenced images in a plane‐of‐sky shell between 20 to 22.5 solar radii and, for a range of position angles, compare daily means of HI variability and in situ solar wind speed estimates. We show that the HI variability data and in situ solar wind speeds have similar temporal autocorrelation functions. Carrington rotation periodicities are well documented for in situ solar wind speeds, but, to our knowledge, this is the first time they have been presented in statistics computed from HI images. In situ solar wind speeds from STEREO‐A, STEREO‐B, and WIND are all are correlated with the HI variability, with a lag that varies in a manner consistent with the longitudinal separation of the in situ monitor and the HI instrument. Unlike many approaches to processing HI observations, our method requires no manual feature tracking; it is automated, is quick to compute, and does not suffer the subjective biases associated with manual classifications. These results suggest we could possibly estimate solar wind speeds in the low heliosphere directly from HI observations. This motivates further investigation, as this could be a significant asset to the space weather forecasting community; it might provide an independent observational constraint on heliospheric solar wind forecasts, through, for example, data assimilation. Finally, these results are another argument for the potential utility of including a HI on an operational space weather mission.
At geosynchronous Earth orbit, the radiation belt/ring current electron fluxes with energies up to several hundred kiloelectron volts can vary widely in magnetic local time (MLT). This study aims to develop Nonlinear AutoRegressive eXogenous models using system science techniques, which account for the spatial variation in MLT. This is difficult for system science techniques, since there is sparse data availability of the electron fluxes at different MLT. To solve this problem, the data are binned from Geostationary Operational Environmental Satellites (GOES) 13, 14, and 15 by MLT, and a separate Nonlinear AutoRegressive eXogenous model is deduced for each bin using solar wind variables as the inputs to the model. These models are then conjugated into one spatiotemporal forecast. The model performance statistics for each model varies in MLT with a prediction efficiency between 47% and 75% and a correlation coefficient between 51.3% and 78.9% for the period from 1 March 2013 to 31 December 2017.
During the intense solar radio bursts on 6 September 2017, Global Navigation Satellite Systems (GNSS) signal interferences were observed at ground stations in the European longitude sector from 20°N to 70°N for all GNSS satellites in view including GPS, GLONASS, and Galileo. The solar radio noise reduced the signal‐to‐noise ratio with clear frequency dependence. The impact of the radio burst has been found at L2 and L5 frequencies, but not at L1 frequency. The ground observation of the solar radio spectrum between 1.0 and 2.0 GHz corresponds well to such frequency dependence. The maximum signal‐to‐noise ratio reduction of ‐10 dB was found when the solar radio flux was pulsating around 2,000 solar flux unit level. Precise point positioning results show that accuracy is reduced with stronger deviation for dual‐frequency solutions than for single‐frequency solutions based on L1 signal only. The positioning error refers rather to the solar extreme ultraviolet flare than to solar radio interferences. The results presented here are a clear indication of frequency‐dependent GNSS performance degradation during strong space weather events.
