Coronal Mass Ejection Sheaths Research Articles

Validity of using Elsässer variables to study the interaction of compressible solar wind fluctuations with a coronal mass ejection

Context. Alfvénic fluctuations, as modelled by the non-linear interactions of Alfvén waves of various scales, are seen to dominate solar wind turbulence. However, there is also a non-negligible component of non-Alfvénic fluctuations. The Elsässer formalism, which is central to the study of Alfvénic turbulence due to its ability to differentiate between parallel and anti-parallel Alfvén waves, cannot strictly separate wavemodes in the presence of compressive magnetoacoustic waves. In this study, we analyse the deviations generated in the Elsässer formalism as density fluctuations are naturally generated through the propagation of a linearly polarised Alfvén wave. The study was performed in the context of a coronal mass ejection (CME) propagating through the solar wind, which enables the creation of two solar wind regimes, pristine wind and a shocked CME sheath, where the Elsässer formalism can be evaluated. Aims. We studied the deviations of the Elsässer formalism in separating parallel and anti-parallel components of Alfvénic solar wind perturbations generated by small-amplitude density fluctuations. Subsequently, we evaluated how the deviations cause a misinterpretation of the composition of waves through the parameters of cross helicity and reflection coefficient. Methods. We used an ideal 2.5D magnetohydrodynamic model with an adiabatic equation of state. An Alfvén pump wave was injected into the quiet solar wind by perturbing the transverse magnetic field and velocity components. This wave subsequently generates density fluctuations through the ponderomotive force. A CME was injected by inserting a flux-rope modelled as a magnetic island into the quasi-steady solar wind. Results. The presence of density perturbations creates a ≈10% deviation in the Elsässer variables and reflection coefficient for the Alfvén waves as well as a deviation of ≈0.1 in the cross helicity in regions containing both parallel and anti-parallel fluctuations.

Modelling the interaction of Alfvénic fluctuations with coronal mass ejections in the low solar corona

Context.Alfvénic fluctuations of various scales are ubiquitous in the corona; their non-linear interactions and eventual turbulent cascade result in an important heating mechanism that accelerates the solar wind. These fluctuations may be processed by large-scale, transient, and coherent heliospheric structures such as coronal mass ejections (CMEs). In this study we investigate the interactions between Alfvénic solar wind fluctuations and CMEs using magnetohydrodynamic (MHD) simulations.Aims.We study the transmission of upstream solar wind fluctuations into the CME leading to the formation of CME sheath fluctuations. Additionally, we investigate the influence of the fluctuation frequencies on the extent of the CME sheath.Methods.We used an ideal MHD model with an adiabatic equation of state. An Alfvén pump wave is injected into the quiet solar wind by perturbing the transverse magnetic field and velocity components, and a CME is injected by inserting a flux-rope modelled as a magnetic island into the quasi-steady solar wind.Results.The upstream Alfvén waves experience a decrease in wavelength and change in the wave vector direction due to the non-radial topology of the CME shock front. The CME sheath inhibits the transmission of long-wavelength fluctuations due to the presence of non-radial flows in this region. The frequency of the solar wind fluctuations also affects the steepening of MHD fast waves causing the CME shock propagation speed to vary with the solar wind fluctuation frequencies.

Comparison of Field‐Aligned Current Responses to HSS/SIR, Sheath, and Magnetic Cloud Driven Geomagnetic Storms

AbstractThe time delay from an interplanetary driver arriving at the magnetopause to the response in the ionosphere has never been quantified separately for different types of storm drivers. This study investigates the delay for storms driven by high‐speed streams and associated stream interaction regions (HSS/SIR), or by interplanetary coronal mass ejection sheaths and magnetic clouds (MC). The total field‐aligned current (FAC) and SME index lag the Newell coupling function (NCF) by 40 ± 10 min during storms driven by HSS/SIR and sheaths, and by 60 ± 10 min for MCs. The correlation coefficient between FAC and NCF reaches maximum value as NCF is averaged over the preceding 80 min for sheath, 90 min for HSS/SIR, and 140 min for MC storms.

Turbulent Cascade and Energy Transfer Rate in a Solar Coronal Mass Ejection

Abstract Turbulence properties are examined before, during, and after a coronal mass ejection (CME) detected by the Wind spacecraft in 2012 July. The power-law scaling of the structure functions, providing information on the power spectral density and flatness of the velocity, magnetic field, and density fluctuations, were examined. The third-order moment scaling law for incompressible, isotropic magnetohydrodynamic turbulence was observed in the preceding and trailing solar wind, as well as in the CME sheath and magnetic cloud. This suggests that the turbulence could develop sufficiently after the shock, or that turbulence in the sheath and cloud regions was robustly preserved even during the mixing with the solar wind plasma. The turbulent energy transfer rate was thus evaluated in each of the regions. The CME sheath shows an increase of energy transfer rate, as expected from the lower level of Alfvénic fluctuations and suggesting the role of the shock-wind interaction as an additional source of energy for the turbulent cascade.

Dynamic Mechanisms Associated With High‐Energy Electron Flux Dropout in the Earth's Outer Radiation Belt Under the Influence of a Coronal Mass Ejection Sheath Region

AbstractThe near‐Earth interplanetary environment conditions affect the dynamics of the relativistic electron population quasitrapped in the radiation belts. A complex chain of processes observed in the magnetosphere can contribute to the variability of these populations when interplanetary structures, such as the interplanetary counterpart of a solar coronal mass ejection (ICME), and high‐speed solar wind streams interact with the magnetosphere. However, as these processes can coexist, it is hard to untangle the relative contribution of each process to the loss of particles and the eventual repopulation. Here we show evidence that it is possible to distinguish the relative contribution of mechanisms related to the loss of the outer radiation belt electrons for an event observed on July 19 and 20, 2016. The interaction of an ICME's turbulent sheath with the Earth's magnetosphere resulted in a decrease in the outer radiation belt relativistic electron population. The ultralow frequency (ULF) and chorus wave activities are detected in the outer radiation belt during the time when the Earth's magnetosphere is under the influence of the ICME's sheath region, as well as the ICME's magnetic cloud region, while the electromagnetic ion cyclotron (EMIC) waves in the outer belt are observed only during the sheath region. Dynamic mechanisms such as magnetopause shadowing, outward radial diffusion driven by ULF waves, pitch‐angle scattering driven by both EMIC and chorus waves are quantitatively analyzed. Our results suggest that the structures of the ICMEs can trigger the drivers to generate the different dynamic mechanisms responsible for the radiation belt population variability.

Properties of the Sheath Regions of Coronal Mass Ejections with or without Shocks from STEREO in situ Observations near 1 au

Abstract We examine 188 coronal mass ejections (CMEs) measured by the twin Solar Terrestrial Relations Observatory spacecraft during 2007–2016 to investigate the generic features of the CME sheath and the magnetic ejecta (ME) and dependencies of average physical parameters of the sheath on the ME. We classify the CMEs into three categories, focusing on whether the ME drives both a shock and sheath, or only a sheath, or neither, near 1 au. We also re-evaluate our initial classification through an automated algorithm and visual inspection. We observe that even for leading-edge speeds greater than 500 km s−1, 1 out of 4 MEs do not drive shocks near 1 au. MEs driving both shocks and sheaths are the fastest and propagate in high magnetosonic solar wind, whereas MEs driving only sheaths are the slowest and propagate in low magnetosonic solar wind. Our statistical and superposed epoch analyses indicate that all parameters are more enhanced in the sheath regions following shocks than in sheaths without shocks. However, differences within sheaths become statistically less significant for similar driving MEs. We also find that the radial thickness of ME-driven sheaths apparently has no clear linear correlation with the speed profile and associated Mach numbers of the driver.

Magnetic field fluctuation properties of coronal mass ejection-driven sheath regions in the near-Earth solar wind

Abstract. In this work, we investigate magnetic field fluctuations in three coronal mass ejection (CME)-driven sheath regions at 1 AU, with their speeds ranging from slow to fast. The data set we use consists primarily of high-resolution (0.092 s) magnetic field measurements from the Wind spacecraft. We analyse magnetic field fluctuation amplitudes, compressibility, and spectral properties of fluctuations. We also analyse intermittency using various approaches; we apply the partial variance of increments (PVIs) method, investigate probability distribution functions of fluctuations, including their skewness and kurtosis, and perform a structure function analysis. Our analysis is conducted separately for three different subregions within the sheath and one in the solar wind ahead of it, each 1 h in duration. We find that, for all cases, the transition from the solar wind ahead to the sheath generates new fluctuations, and the intermittency and compressibility increase, while the region closest to the ejecta leading edge resembled the solar wind ahead. The spectral indices exhibit large variability in different parts of the sheath but are typically steeper than Kolmogorov's in the inertial range. The structure function analysis produced generally the best fit with the extended p model, suggesting that turbulence is not fully developed in CME sheaths near Earth's orbit. Both Kraichnan–Iroshinikov and Kolmogorov's forms yielded high intermittency but different spectral slopes, thus questioning how well these models can describe turbulence in sheaths. At the smallest timescales investigated, the spectral indices indicate shallower than expected slopes in the dissipation range (between −2 and −2.5), suggesting that, in CME-driven sheaths at 1 AU, the energy cascade from larger to smaller scales could still be ongoing through the ion scale. Many turbulent properties of sheaths (e.g. spectral indices and compressibility) resemble those of the slow wind rather than the fast. They are also partly similar to properties reported in the terrestrial magnetosheath, in particular regarding their intermittency, compressibility, and absence of Kolmogorov's type turbulence. Our study also reveals that turbulent properties can vary considerably within the sheath. This was particularly the case for the fast sheath behind the strong and quasi-parallel shock, including a small, coherent structure embedded close to its midpoint. Our results support the view of the complex formation of the sheath and different physical mechanisms playing a role in generating fluctuations in them.

Spatial Coherence of Interplanetary Coronal Mass Ejection Sheaths at 1 AU

AbstractThe longitudinal spatial coherence near 1 AU of the magnetic field in sheath regions driven by interplanetary coronal mass ejection (ICME) is studied by investigating ACE and Wind spacecraft measurements of 29 sheaths. During 2000–2002 Wind performed prograde orbits, and the non‐radial spacecraft separation varied from 0.001 to 0.012 AU between the studied events. We compare the measurements by computing the Pearson correlation coefficients for the magnetic field magnitude and components and estimate the magnetic field coherence by evaluating the scale lengths that give the extrapolated distance of zero correlation between the measurements. The correlation is also separately examined for low‐ and high‐pass filtered data. We discover magnetic fields in ICME sheaths have scale lengths that are larger than those reported in the solar wind but that, in general, are smaller than the ones of the ICME ejecta. Our results imply that magnetic fields in the sheath are more coherently structured and well correlated compared to the solar wind. The largest sheath coherence is reported in the GSE y‐direction that has the scale length of 0.149 AU while the lengths for Bx, Bz, and |B| vary between 0.024 and 0.035 AU. The same sheath magnitude ordering of scale lengths also apply for the low‐pass filtered magnetic field data. We discuss field line draping and the alignment of preexisting discontinuities by the shock passage giving reasoning for the observed results.

Cosmic noise absorption signature of particle precipitation during interplanetary coronal mass ejection sheaths and ejecta

Abstract. We study here energetic-electron (E>30 keV) precipitation using cosmic noise absorption (CNA) during the sheath and ejecta structures of 61 interplanetary coronal mass ejections (ICMEs) observed in the near-Earth solar wind between 1997 and 2012. The data come from the Finnish riometer (relative ionospheric opacity meter) chain from stations extending from auroral (IVA, 65.2∘ N geomagnetic latitude; MLAT) to subauroral (JYV, 59.0∘ N MLAT) latitudes. We find that sheaths and ejecta lead frequently to enhanced CNA (>0.5 dB) both at auroral and subauroral latitudes, although the CNA magnitudes stay relatively low (medians around 1 dB). Due to their longer duration, ejecta typically lead to more sustained enhanced CNA periods (on average 6–7 h), but the sheaths and ejecta were found to be equally effective in inducing enhanced CNA when relative-occurrence frequency and CNA magnitude were considered. Only at the lowest-MLAT station, JYV, ejecta were more effective in causing enhanced CNA. Some clear trends of magnetic local time (MLT) and differences between the ejecta and sheaths were found. The occurrence frequency and magnitude of CNA activity was lowest close to midnight, while it peaked for the sheaths in the morning and afternoon/evening sectors and for the ejecta in the morning and noon sectors. These differences may reflect differences in typical MLT distributions of wave modes that precipitate substorm-injected and trapped radiation belt electrons during the sheaths and ejecta. Our study also emphasizes the importance of substorms and magnetospheric ultra-low-frequency (ULF) waves for enhanced CNA.

Radial Evolution of Magnetic Field Fluctuations in an Interplanetary Coronal Mass Ejection Sheath

Abstract The sheaths of compressed solar wind that precede interplanetary coronal mass ejections (ICMEs) commonly display large-amplitude magnetic field fluctuations. As ICMEs propagate radially from the Sun, the properties of these fluctuations may evolve significantly. We have analyzed magnetic field fluctuations in an ICME sheath observed by MESSENGER at 0.47 au and subsequently by STEREO-B at 1.08 au while the spacecraft were close to radial alignment. Radial changes in fluctuation amplitude, compressibility, inertial-range spectral slope, permutation entropy, Jensen–Shannon complexity, and planar structuring are characterized. These changes are discussed in relation to the evolving turbulent properties of the upstream solar wind, the shock bounding the front of the sheath changing from a quasi-parallel to quasi-perpendicular geometry, and the development of complex structures in the sheath plasma.

Evolution of a Long‐Duration Coronal Mass Ejection and Its Sheath Region Between Mercury and Earth on 9–14 July 2013

AbstractUsing in situ measurements and remote‐sensing observations, we study a coronal mass ejection (CME) that left the Sun on 9 July 2013 and impacted both Mercury and Earth while the planets were in radial alignment (within ). The CME had an initial speed as measured by coronagraphs of 580 20 km/s, an inferred speed at Mercury of 580 30 km/s, and a measured maximum speed at Earth of 530 km/s, indicating that it did not decelerate substantially in the inner heliosphere. The magnetic field measurements made by MESSENGER and Wind reveal a very similar magnetic ejecta at both planets. We consider the CME expansion as measured by the ejecta duration and the decrease of the magnetic field strength between Mercury and Earth and the velocity profile measured in situ by Wind. The long‐duration magnetic ejecta (20 and 42 hr at Mercury and Earth, respectively) is found to be associated with a relatively slowly expanding ejecta at 1 AU, revealing that the large size of the ejecta is due to the CME itself or its expansion in the corona or innermost heliosphere and not due to a rapid expansion between Mercury at 0.45 AU and Earth at 1 AU. We also find evidence that the CME sheath is composed of compressed material accumulated before the shock formed, as well as more recently shocked material.

Radial Evolution of Coronal Mass Ejections Between MESSENGER, Venus Express, STEREO, and L1: Catalog and Analysis

AbstractOur knowledge of the properties of coronal mass ejections (CMEs) in the inner heliosphere is constrained by the relative lack of plasma observations between the Sun and 1 AU. In this work, we present a comprehensive catalog of 47 CMEs measured in situ measurements by two or more radially aligned spacecraft (MESSENGER, Venus Express, STEREO, Wind/ACE). We estimate the CME impact speeds at Mercury and Venus using a drag‐based model and present an average propagation profile of CMEs (speed and deceleration/acceleration) in the inner heliosphere. We find that CME deceleration continues past Mercury's orbit but most of the deceleration occurs between the Sun and Mercury. We examine the exponential decrease of the maximum magnetic field strength in the CME with heliocentric distance using two approaches: a modified statistical method and analysis from individual conjunction events. Findings from both the approaches are on average consistent with previous studies but show significant event‐to‐event variability. We also find the expansion of the CME sheath to be well fit by a linear function. However, we observe the average sheath duration and its increase to be fairly independent of the initial CME speed, contradicting commonly held knowledge that slower CMEs drive larger sheaths. We also present an analysis of the 3 November 2011 CME observed in longitudinal conjunction between MESSENGER, Venus Express, and STEREO‐B focusing on the expansion of the CME and its correlation with the exponential falloff of the maximum magnetic field strength in the ejecta.

The Response of Earth's Electron Radiation Belts to Geomagnetic Storms: Statistics From the Van Allen Probes Era Including Effects From Different Storm Drivers

AbstractA statistical study was conducted of Earth's radiation belt electron response to geomagnetic storms using NASA's Van Allen Probes mission. Data for electrons with energies ranging from 30 keV to 6.3 MeV were included and examined as a function of L‐shell, energy, and epoch time during 110 storms with SYM‐H ≤−50 nT during September 2012 to September 2017 (inclusive). The radiation belt response revealed clear energy and L‐shell dependencies, with tens of keV electrons enhanced at all L‐shells (2.5 ≤ L ≤ 6) in all storms during the storm commencement and main phase and then quickly decaying away during the early recovery phase, low hundreds of keV electrons enhanced at lower L‐shells (~3 ≤ L ≤ ~4) in upward of 90% of all storms and then decaying gradually during the recovery phase, and relativistic electrons throughout the outer belt showing main phase dropouts with subsequent and generally unpredictable levels of replenishment during the recovery phase. Compared to prestorm levels, electrons with energies >1 MeV also revealed a marked increase in likelihood of a depletion at all L‐shells through the outer belt (3.5 ≤ L ≤ 6). Additional statistics were compiled revealing the storm time morphology of the radiation belts, confirming the aforementioned qualitative behavior. Considering storm drivers in the solar wind: storms driven by coronal mass ejection (CME) shocks/sheaths and CME ejecta only are most likely to result in a depletion of >1‐MeV electrons throughout the outer belt, while storms driven by full CMEs and stream interaction regions are most likely to produce an enhancement of MeV electrons at lower (L < ~5) and higher (L > ~4.5) L‐shells, respectively. CME sheaths intriguingly result in a distinct enhancement of ~1‐MeV electrons around L~5.5, and on average, CME sheaths and stream interaction regions result in double outer belt structures.

The Interaction of Coronal Mass Ejections with Alfvénic Turbulence

We provide a first attempt to understand the interaction between Alfvén wave turbulence, kinetic instabilities and temperature anisotropies in the environment of a fast coronal mass ejection (CME) near the Sun. The impact of a fast CME on the solar corona causes turbulent energy, thermal energy and dissipative heating to increase by orders of magnitude, and produces conditions suitable for a host of kinetic instabilities. We study these CME-induced effects with the recently developed Alfvén Wave Solar Model, with which we are able to self-consistently simulate the turbulent energy transport and dissipation as well as isotropic electron heating and anisotropic proton heating. Furthermore, the model also offers the capability to address the effects of fire hose, mirror mode, and cyclotron kinetic instabilities on proton energy partitioning all in a global-scale numerical simulation. We find amplified turbulent energy in the CME sheath, along with strong wave reflection at the shock combine to cause wave dissipation rates to increase by more than a factor of 100. In contrast, wave energy is greatly diminished by adiabatic expansion in the flux rope. Finally, we find proton temperature anisotropies are limited by kinetic instabilities to a level consistent with solar wind observations.

Planar magnetic structures in coronal mass ejection-driven sheath regions

Abstract. Planar magnetic structures (PMSs) are periods in the solar wind during which interplanetary magnetic field vectors are nearly parallel to a single plane. One of the specific regions where PMSs have been reported are coronal mass ejection (CME)-driven sheaths. We use here an automated method to identify PMSs in 95 CME sheath regions observed in situ by the Wind and ACE spacecraft between 1997 and 2015. The occurrence and location of the PMSs are related to various shock, sheath, and CME properties. We find that PMSs are ubiquitous in CME sheaths; 85 % of the studied sheath regions had PMSs with the mean duration of 6 h. In about one-third of the cases the magnetic field vectors followed a single PMS plane that covered a significant part (at least 67 %) of the sheath region. Our analysis gives strong support for two suggested PMS formation mechanisms: the amplification and alignment of solar wind discontinuities near the CME-driven shock and the draping of the magnetic field lines around the CME ejecta. For example, we found that the shock and PMS plane normals generally coincided for the events where the PMSs occurred near the shock (68 % of the PMS plane normals near the shock were separated by less than 20° from the shock normal), while deviations were clearly larger when PMSs occurred close to the ejecta leading edge. In addition, PMSs near the shock were generally associated with lower upstream plasma beta than the cases where PMSs occurred near the leading edge of the CME. We also demonstrate that the planar parts of the sheath contain a higher amount of strong southward magnetic field than the non-planar parts, suggesting that planar sheaths are more likely to drive magnetospheric activity.

OBSERVATION OF HEATING BY FLARE-ACCELERATED ELECTRONS IN A SOLAR CORONAL MASS EJECTION

We report a Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) observation of flare-accelerated electrons in the core of a coronal mass ejection (CME) and examine their role in heating the CME. Previous CME observations have revealed remarkably high thermal energies that can far surpass the CME's kinetic energy. A joint observation by RHESSI and the Atmospheric Imaging Assembly of a partly occulted flare on 2010 November 3 allows us to test the hypothesis that this excess energy is collisionally deposited by flare-accelerated electrons. Extreme ultraviolet (EUV) images show an ejection forming the CME core and sheath, with isothermal multifilter analysis revealing temperatures of ~11 MK in the core. RHESSI images reveal a large (~100 × 50 arcsec2) hard X-ray (HXR) source matching the location, shape, and evolution of the EUV plasma, indicating that the emerging CME is filled with flare-accelerated electrons. The time derivative of the EUV emission matches the HXR light curve (similar to the Neupert effect observed in soft and HXR time profiles), directly linking the CME temperature increase with the nonthermal electron energy loss, while HXR spectroscopy demonstrates that the nonthermal electrons contain enough energy to heat the CME. This is the most direct observation to date of flare-accelerated electrons heating a CME, emphasizing the close relationship of the two in solar eruptive events.

High-rigidity Forbush decreases: due to CMEs or shocks?

We seek to identify the primary agents causing Forbush decreases (FDs) observed at the Earth in high rigidity cosmic rays. In particular, we ask if such FDs are caused mainly by coronal mass ejections (CMEs) from the Sun that are directed towards the Earth, or by their associated shocks. We use the muon data at cutoff rigidities ranging from 14 to 24 GV from the GRAPES-3 tracking muon telescope to identify FD events. We select those FD events that have a reasonably clean profile, and can be reasonably well associated with an Earth-directed CME and its associated shock. We employ two models: one that considers the CME as the sole cause of the FD (the CME-only model) and one that considers the shock as the only agent causing the FD (the shock-only model). We use an extensive set of observationally determined parameters for both these models. The only free parameter in these models is the level of MHD turbulence in the sheath region, which mediates cosmic ray diffusion (into the CME, for the CME-only model and across the shock sheath, for the shock-only model). We find that good fits to the GRAPES-3 multi-rigidity data using the CME-only model require turbulence levels in the CME sheath region that are only slightly higher than those estimated for the quiet solar wind. On the other hand, reasonable model fits with the shock-only model require turbulence levels in the sheath region that are an order of magnitude higher than those in the quiet solar wind. This observation naturally leads to the conclusion that the Earth-directed CMEs are the primary contributors to FDs observed in high rigidity cosmic rays.

Eddy viscosity and flow properties of the solar wind: Co-rotating interaction regions, coronal-mass-ejection sheaths, and solar-wind/magnetosphere coupling

The coefficient of magnetohydrodynamic (MHD) eddy viscosity of the turbulent solar wind is calculated to be νeddy≈1.3×1017cm2∕s: this coefficient is appropriate for velocity shears with scale thicknesses larger than the ∼106 km correlation length of the solar-wind turbulence. The coefficient of MHD eddy viscosity is calculated again accounting for the action of smaller-scale turbulent eddies on smaller scale velocity shears in the solar wind. This eddy viscosity is quantitatively tested with spacecraft observations of shear flows in co-rotating interaction regions (CIRs) and in coronal-mass-ejection (CME) sheaths and ejecta. It is found that the large-scale (∼107km) shear of the CIR fractures into intense narrow (∼105km) slip zones between slabs of differently magnetized plasma. Similarly, it is found that the large-scale shear of CME sheaths also fracture into intense narrow slip zones between parcels of differently magnetized plasma. Using the solar-wind eddy-viscosity coefficient to calculate vorticity-diffusion time scales and comparing those time scales with the ∼100-h age of the solar-wind plasma at 1AU, it is found that the slip zones are much narrower than eddy-viscosity theory says they should be. Thus, our concept of MHD eddy viscosity fails testing. For the freestream turbulence effect in solar-wind magnetosphere coupling, the eddy-viscous force of the solar wind on the Earth's magnetosphere is rederived accounting for the action of turbulent eddies smaller than the correlation length, along with other corrections. The improved derivation of the solar-wind driver function for the turbulence effect fails to yield higher correlation coefficients between measurements of the solar-wind driver and measurements of the response of the Earth's magnetosphere.