We present a proof of concept for the probabilistic emulation of the Ring current-Atmosphere interactions Model with Self-Consistent magnetic field (RAM-SCB) particle flux. We extend the workflow developed by Licata and Mehta (2023) by applying it to the ring current and further developing its uncertainty quantification methodology. We introduce a novel approach for sampling over 20 years of solar and geomagnetic activity to identify 30 simulation periods, each one week long, to generate the training, validation, and test datasets. Large-scale physics-based simulation models for the ring current can be computationally expensive. This work aims at creating an emulator that is more efficient, capable of forecasting, and provides an estimate on the uncertainty of its predictions, all without requiring large computational resources. We demonstrate the emulation process on a subset of particle flux: a single energy channel of omnidirectional flux. A principal component analysis (PCA) is used for the dimensionality reduction into the reduced-space, and the dynamic modeling is performed with a recurrent neural network. A hierarchical ensemble of Long-Short Term Memory (LSTM) neural networks provides the statistics needed to produce a probabilistic output, resulting in a reduced-order probabilistic emulator (ROPE) that performs time-series forecasting of the ring current’s particle flux with an estimate on its uncertainty distribution. The resulting ROPE from this smaller subset of RAM-SCB particle flux provides dynamic predictions with errors less than 11% and calibration scores under 10%, demonstrating that this workflow can provide a probabilistic emulator with a robust and reliable uncertainty estimate when applied to the ring current.

This editorial aims to improve awareness of the current best practices in open research, and stimulate discussion on the practical implementation of AGU's data and software policy in key areas of space weather research. We also further aim to encourage authors to take additional steps to ensure clear credit to all contributors to the work, whether that is underlying data, key software, or direct contributions to the manuscript.

We have developed a new procedure for combining lists of substorm onset times from multiple sources. We apply this procedure to observational data and to magnetohydrodynamic (MHD) model output from 1-31 January, 2005. We show that this procedure is capable of rejecting false positive identifications and filling data gaps that appear in individual lists. The resulting combined onset lists produce a waiting time distribution that is comparable to previously published results, and superposed epoch analyses of the solar wind driving conditions and magnetospheric response during the resulting onset times are also comparable to previous results. Comparison of the substorm onset list from the MHD model to that obtained from observational data reveals that the MHD model reproduces many of the characteristic features of the observed substorms, in terms of solar wind driving, magnetospheric response, and waiting time distribution. Heidke skill scores show that the MHD model has statistically significant skill in predicting substorm onset times.

The radiation belts of the Earth, filled with energetic electrons, comprise complex and dynamic systems that pose a significant threat to a variety of satellite systems. While various models of the relativistic electron flux have been developed for geostationary orbit (GEO), the behaviour of the medium energy (120-600 keV) electrons below GEO remains poorly quantified. In this paper we present a Medium Energy electRon flux In Earth’s outer radiatioN belt (MERLIN) model based on the Light Gradient Boosting (LightGBM) machine learning algorithm. The MERLIN model takes as input the satellite position, a combination of geomagnetic indices and solar wind parameters including the time history of velocity, and does not use persistence. MERLIN is trained and validated on $>$15 years of the GPS electron flux data, and tested on more than $1.5$ years of measurements. 10-fold cross validation (CV) yields that the model predicts the MEO radiation environment well, both in terms of dynamics and amplitudes of flux. Evaluation on the test set yields high correlation between the predicted and observed electron flux (0.8) and low values of absolute error. The MERLIN model can have wide Space Weather applications, providing information for the scientific community in the form of radiation belts reconstructions, as well as industry for satellite mission design, nowcast of the MEO environment and surface charging analysis.

Turbulent and compressed sheath regions preceding interplanetary coronal mass ejections (ICMEs) strongly impact electron dynamics in the outer radiation belt. Changes in electron flux can occur on timescales of tens of minutes, which is difficult to capture by a two-satellite mission such as the Van Allen Probes (RBSP). The recently released Global Positioning System (GPS) data set has higher data density owing to the large number of satellites in the constellation equipped with energetic particle detectors. Investigating electron fluxes in a wide range of energies and sheaths observed from 2012 to 2018, we show that the flux response to sheaths on a timescale of 6 hours, previously reported from RBSP data, is reproduced by GPS measurements. Furthermore, GPS data enables derivation of the response on a shorter timescale of 30 minutes, which further confirms that the energy and L-shell dependent changes in electron flux are due to the impact of the sheath. Sheath-driven loss is underestimated over longer timescales as the electrons recover during the ejecta. We additionally show the response of electron phase space density (PSD), which is a key quantity in identifying true loss from the system and electron energization through wave-particle interactions. The PSD response is calculated from both RBSP and GPS data for the 6-hour timescale, as well as from GPS data for the 30-minute timescale. The response is divided based on the geoeffectiveness of the sheaths revealing that electrons are effectively accelerated only during geoeffective sheaths, while loss is commonly caused by all sheaths.

Magnetohydrodynamic (MHD) models have been used for nearly four decades to study the dynamics of magnetospheric substorms. However, until recently no demonstration has been made that MHD models can consistently reproduce substorm onset times in a statistical sense. To test whether MHD can reproduce observed substorm onset times, we developed a procedure for identifying substorm onsets that can be applied both to observational data and to MHD output. Our substorm identification procedure aims to improve upon existing methods of substorm identification by using multiple types of observations to corroborate each identified substorm. Using this procedure, we identified over 100 substorms from the period 1-31 January 2005. Using this list of substorm onset times, we show that the MHD model has weak, but statistically significant skill in predicting substorm onset times. We explore paths to improving the ability of the MHD model to predict substorm dynamics by testing different configurations of the MHD model.

The next decadal survey process for space and solar physics will start soon with white papers due during the second half of 2022 and the committees and panels working over all of 2023. Space weather science and operations will play an essential role in this survey. Therefore, the community is invited to prepare white papers and get involved in advancing space weather research and capabilities in the upcoming decades. A summary of the recommendations related to space weather from the last two decadal surveys is also provided.

Manuscripts based on machine-learning techniques have significantly increased in Space Weather over the past few years. We discuss which manuscripts are within the journal’s scope and emphasize that manuscripts fo-cusing purely on a forecasting technique (rather than on understanding and forecasting a phenomenon) must correspond to a substantial improvement over the current state-of-the-art techniques and present this comparison. All manuscripts shall include information about data preparation, including splitting of data between training, validation and testing sets. The software and/or algorithms used for to develop the machine-learning technique should be included in a repository at the time of submission. Comparison with published results using other methods must be presented, and uncertainties of the forecast results must be discussed.

This paper examines the rapid losses and acceleration of trapped relativistic and ultrarelativistic electron populations in the Van Allen radiation belt during the September 7-9, 2017, geomagnetic storm. By analyzing the dynamics of the last closed drift shell (LCDS) and the electron flux and phase space density (PSD), we show that the electron dropouts are consistent with magnetopause shadowing and outward radial diffusion to the compressed LCDS. During the recovery phase, an in-bound pass of Van Allen Probe A shows an apparent local peak in PSD. However, a fortuitous timing of a crossing of the two Van Allen Probes reveals instead how the apparent PSD peak arises from aliasing monotonic PSD profiles which are rapidly increasing due to acceleration from very fast inwards radial diffusion. In the absence of such multi-satellite conjunctions during fast acceleration events, the source might otherwise be attributed to local acceleration processes.

We present a methodology to define strong, moderate, and intense space weather events based on probability distributions. We have illustrated this methodology using a long-duration, uniform data set of 1.8-3.5 MeV electron fluxes from multiple LANL geosynchronous satellite instruments but a strength of this methodology is that it can be applied uniformly to heterogeneous data sets. It allows quantitative comparison of data sets with different energies, units, orbits, etc. The methodology identifies a range of times, “events”, using variable flux thresholds to determine average event occurrence in arbitrary 11-year intervals (“cycles”). We define strong, moderate, and intense events as those that occur 100, 10, and 1 time per cycle and identify the flux thresholds that produce those occurrence frequencies. The methodology does not depend on any ancillary data set (e.g. solar wind or geomagnetic conditions). We show event probabilities using GOES > 2 MeV fluxes and compare them against event probabilities using LANL 1.8-3.5 MeV fluxes. We present some examples of how the methodology picks out strong, moderate, and intense events and how those events are distributed in time: 1989 through 2018, which includes the declining phases of solar cycles 22, 23, and 24. We also provide an illustrative comparison of moderate and strong events identified in the geosynchronous data with Van Allen Probes observations across all L-shells. We also provide a catalog of start and stop times of strong, moderate, and intense events that can be used for future studies.

One of the prominent effects of space weather is the variation of electric currents in the magnetosphere and ionosphere, which can cause localized, high amplitude Geomagnetic Disturbances (GMDs) that disrupt ground conducting systems. Because the source of localized GMDs is unresolved, we are prompted to model these effects, identify the physical drivers through examination of the model we use, and improve our prediction of these phenomena. We run a high-resolution configuration of the Space Weather Modeling Framework (SWMF) to model the September 7, 2017 event, combining three physical models: Block Adaptive Tree Solar wind Roe Upwind Scheme (BATS-R-US), an ideal magnetohydrodynamic model of the magnetosphere; the Ridley Ionosphere Model (RIM), a shell ionosphere calculated by solving 2-D Ohm’s Law; and the Rice Convection Model (RCM), a kinetic drift model of the inner magnetosphere. The configuration mirrors that which is used in Space Weather Prediction Center (SWPC) operations; however, the higher grid resolution can reproduce mesoscale structure in the tail and ionosphere. We use two metrics to quantify the success of the model against observation. Regional Station Difference (RSD) is a metric that uses dB/dt or geoelectric field to pinpoint when a single magnetometer station records a significantly different value than others within a given radius, indicating a localized GMD. Regional Tail Difference (RTD) performs the same calculation using relevant variables in the magnetosphere at points that map down along field lines to the magnetometer station locations on the ground. We theorize two distinct causes of RSD, the first being small-scale structure in the tail and the second being station field lines mapping to spatially separated locations in the tail. We examine the differences between RSD spikes that we can reproduce in the model and those that we cannot. We categorize spikes by cause of localized GMDs to examine model capability for each theorized cause. We investigate the improvements in our model when we switch from empirical specification of ionosphere conductance to a physics-based one, MAGNetosphere-Ionosphere-Thermosphere (MAGNIT) Auroral Conductance Model. For small-scale effects we cannot reproduce, we explore the deficiencies in our model.

We present simulations of the outer radiation belt electron flux during the March 2015 and March 2013 storms using a radial diffusion model. Despite differences in Dst intensity between the two storms the response of the ultra-relativistic electrons in the outer radiation belt was remarkably similar, both showing a sudden drop in the electron flux followed by a rapid enhancement in the outer belt flux to levels over an order of magnitude higher than those observed during the pre-storm interval. Simulations of the ultra-relativistic electron flux during the March 2015 storm show that outward radial diffusion can explain the flux dropout down to L*=4. However, in order to reproduce the observed flux dropout at L*<4 requires the addition of a loss process characterised by an electron lifetime of around one hour operating below L*~3.5 during the flux dropout interval. Nonetheless, during the pre-storm and recovery phase of both storms the radial diffusion simulation reproduces the observed flux dynamics. For the March 2013 storm the flux dropout across all L-shells is reproduced by outward radial diffusion activity alone. However, during the flux enhancement interval at relativistic energies there is evidence of a growing local peak in the electron phase space density at L*~3.8, consistent with local acceleration such as by VLF chorus waves. Overall the simulation results for both storms can accurately reproduce the observed electron flux only when event specific radial diffusion coefficients are used, instead of the empirical diffusion coefficients derived from ULF wave statistics.