betaFeedback

WorksPhysics and Astronomy

Solar wind entry into Mercury's magnetosphere: Simulation results for the second swingby of BepiColombo

Abstract

Context. We use a global 3D hybrid plasma model to investigate the interaction between Mercury's magnetosphere and the solar wind for the second BepiColombo swingby, evaluate magnetospheric regions, and study the typical energy profile of protons. Aims. The objective of this study is to gain a better understanding of solar wind entry and analyze simulated plasma data along a trajectory using BepiColombo swingby 2 conditions, with the goal of enhancing our comprehension of measurement data and potentially providing forecasts for future swingbys. Methods. To model Mercury's plasma environment, we used the hybrid code AIKEF and developed a method to extract the particle (ion) data in order to compute the proton energy spectrum along the trajectory of BepiColombo during its second Mercury swingby on June 23, 2022. We evaluate magnetopause and bow shock stand-off distances under average upstream solar wind conditions with the Interplanetary Magnetic Field (IMF) condition derived from the BepiColombo magnetic field measurements during the second Mercury swingby. Results. We found that the magnetosheath on the quasi-perpendicular (dusk) side of the bow shock is thicker than that on the quasi-parallel (dawn) side, where a foreshock is formed. Multiple plasma populations can be extracted from our modeled energy spectra that assist in identifying magnetospheric regions. We observed protons of solar wind origin entering Mercury's magnetosphere. Their energies range from a few electron volts in the magnetosphere up to 10 keV in the magnetosheath.

Introduction

In certain aspects, Mercury’s magnetosphere is a scaled-down version of the Earth’s magnetosphere. However, the typical solar wind dynamic pressure at Mercury is about five times higher than at the Earth, and the average IMF magnitude is about four times stronger [55]. Together with the weak intrinsic magnetic field of Mercury, a small and dynamic magnetosphere, compared to the Earth, is formed that interacts with the solar wind and its embedded interplanetary magnetic field (IMF) [46, 3, 45]. In addition to a stronger dependency on the solar wind relative to the Earth, Mercury’s magnetosphere also reacts faster to solar wind changes. The Dungey cycle, which describes the plasma circulation and magnetic flux transfer in the magnetosphere, takes only about two minutes, as MESSENGER observations show [44, 47]. The timescale is significantly shorter than the Dungey cycle at the Earth (about one hour, [6]), suggesting that the Hermean system is more dynamic and can only be considered quasi-stationary for a few minutes. Compared to the other terrestrial planets, Mercury is unique in (1) the lack of a significant ionosphere and (2) the existence of a large, electrically conducting core [48]. These features significantly impact the magnetospheric current system and induction effects [26, 15, 16] (Ganushkina et al. 2015).

The pioneering work by Ness et al. [35] used the first in situ measurements of Mercury’s magnetosphere (Mariner 10, Dunne & Burgess 1978) to provide initial insights. Sc​​is​​coe & Christopher (1975) and Slavin & Holzer [46] used this data together with observational data from the Earth to evaluate the solar wind stand-off distance at Mercury to about (1.3 - 2. RM_{\mathrm{M}} (RMR_{\mathrm{M}} = 2440 km being the radius of Mercury), depending on upstream solar wind condition. More than three decades later, MESSENGER [49] observations revealed a more detailed insight and narrowed the estimate for the solar wind stand-off distance to (1.35 - 1.55) RMR_{\mathrm{M}} [55]. The average subsolar distance of the bow shock is (1.89 - 2.29) RMR_{\mathrm{M}} [44, 55]. Previous studies support these results and further examine the interaction between Mercury and the solar wind, using both measurements and (numerical) models, under average solar wind conditions and varying IMF conditions (e.g., [52, 5, 26, 17, 19]) or during extreme events, such as CMEs (e.g., Exner et al. [15]; Jia et al. (2019)).

The BepiColombo space mission (Benkhoff et al. [8]) will perform a total of six swingbys near Mercury before its final arrival at the planet in 2025. While the nominal mission does not begin before the arrival in the final science orbits, traversing Mercury’s magnetosphere at various angles during the swingbys and reaching approaches as close as about 200 km gives unique science opportunities to investigate the planet’s global magnetosphere.

While there is no significant ionosphere, Mercury possesses a tenuous exosphere consisting mainly of sodium ([10] (Milillo et al. 2005)). This gas envelope is predominantly generated by multiple surface processes of the solar wind and solar radiation interaction with the planetary surface ([57, 56, 9, 20, 16]). Hence, there are two main sources for plasma in the magnetosphere: exospheric particles and solar wind particles.

Due to limited observational coverage of previous missions, the plasma motion of planetary and solar wind ions through Mercury’s equatorial low-altitude nightside magnetosphere is largely derived from scaled-down models based on the magnetosphere of Earth. However, global numerical modeling efforts have revealed particular plasma signatures, including partial plasma rings ([12, 51, 15, 58, 59]), and increased exospheric densities ([16]) at these low altitudes. The second swingby trajectory of BepiColombo (henceforth called MSB2) is strongly equatorial, as the spacecraft passed through the southern magnetosphere and had its closest approach (CA) on the dawnward nightside at 199 km on June 23, 2022, obtaining measurements in the previously non-covered low altitudes (see Figure 1).

Image not extracted

Figure 1. Overview of the AIKEF simulation box. In the top-right corner, there is an exemplary unrefined block that is further divided into cells that define the grid resolution (size ratios: block to simulation box = cell to block). The black arrows denote the difference between the MASO and MSM coordinate system, which is shifted 0.2 RMR_\mathrm{M} in the +ZZ direction. The subsolar point is defined to be on the MSM X-axis; see modelled bow shock (green, solid line) and magnetopause (blue, dotted). The MSB2 trajectory with its closest approach is shown in red; the orange outline encloses the refined area (grid resolution doubled).

In this study, we investigate the observed particle densities and energies due to solar wind particle entry. A proton energy spectrum is computed that gives insights into regions along the trajectory where, for instance, particles are accelerated or heated. Magnetic field data from the MGF instrument (Baumjohann et al. [7]) on board the Mio spacecraft of BepiColombo are used to determine the outbound IMF parameters. The determined IMF direction and strength agrees well with commonly observed IMF conditions at Mercury resulting from the Parker spiral (James et al. [24]). We address two scientific questions in this study:

How deep do solar wind particles penetrate Mercury’s magnetosphere? What do the energy spectra of the penetrating ions look like? To tackle these questions, we employed a numerical hybrid plasma approach utilizing the AIKEF model.

Hybrid plasma approach

Using a hybrid plasma model to study the interaction between the solar wind and the planet Mercury offers several advantages. The most important one in the context of this study is the advantage of a kinetic description of the ions as it enables the resolution of non-Maxwellian distributions. In addition, such models are computationally inexpensive compared to fully kinetic (particle in cell, or PIC) simulations, as the electrons are described as a fluid and the electron kinetics that happen on much smaller length and time scales than those of ions are omitted (Winske & Omidi [54]). Therefore, physical processes down to ion kinetic time and length scales can be reasonably described using hybrid models. Hybrid modeling has been applied to investigate Mercury’s plasma environment by, for example, Trávnícek et al. [51, 52]; Paral et al. [37]; Paral & Rankin [36]; Müller et al. [33, 34]; Herčík et al. [23]; Fatemi et al. [18, 17]; Exner et al. [15, 16]; Jarvinen et al. [25]; Aizawa et al. [1]; Lu et al. (2022).

In this work, we used the Adaptive Ion-Kinetic Electron-Fluid code, or AIKEF (Müller et al. [33]). It is a global 3D hybrid plasma simulation code where ions are treated kinetically and electrons are a massless charge-neutralizing fluid. A Cartesian mesh grid is used where the electromagnetic fields are computed on the nodes of the numerical mesh. While the number of particles per unit volume gives the resolution in velocity space, the resolution of the numerical mesh gives the spatial resolution. Mercury’s surface is treated as a perfect plasma absorber, which means that particles impacting the surface are removed from the simulation box. Magnetic reconnection is handled self-consistently through anomalous resistivity (Raeder et al. (1998); Jia et al. [28]). (For a detailed explanation on how the electromagnetic fields are propagated and how the AIKEF model is implemented, see Müller et al. [33].)

Simulation setup

The size of the simulation box is 12 RMR_\mathrm{M} x 12 RMR_\mathrm{M} x 12 RMR_\mathrm{M} (see Figure 1). We set the separation between the physical nodes to 114 km, which corresponds to 2.8 di,swd_{i,\mathrm{sw}}, with di,swd_{i,\mathrm{sw}} = 40 km being the ion inertial length in the solar wind (see Table 1). However, the spatial resolution near the planet (−2.1RM<X<3.9RM-2.1R_\mathrm{M} < X < 3.9R_\mathrm{M}; −4.1RM<Y<4.1RM-4.1R_\mathrm{M} < Y < 4.1R_\mathrm{M}; −1.9RM<Z<1.9RM-1.9R_\mathrm{M} < Z < 1.9R_\mathrm{M}) is twice as high when using mesh refinement, resulting in a node distance of 57 km (or 1.4 di,swd_{i,\mathrm{sw}}). The model utilizes the right-handed Mercury Anti Solar Orbital (MASO) coordinate system where the X-axis is directed as the solar wind velocity (i.e., anti-sunward), the Z-axis is the positive normal to the orbital plane (anti-parallel to Mercury’s dipole moment), and the Y-axis is the orthogonal complement roughly pointing toward Mercury’s orbital movement direction (i.e., dawnside). To reach a quasi-steady state, the total run time of the simulation is about 760 seconds in the simulation domain, which corresponds to about 6 Dungey cycles [44]. Apart from the IMF direction derived from the BepiColombo Mio MGF measurements during the outbound trajectory phase [7], all other physical input parameters are adapted from the average values obtained from four years of MESSENGER observations [55, 4], see Table 1.

ParameterValue
IMF direction (upstream)(−0.79,0.59,0.06)(-0.79, 0.59, 0.06)
IMF magnitude (upstream)24.2 nT
Solar wind (H+^{+}) number density40 particles/cm3^{3}
Solar wind (upstream) bulk velocity400 km/s
Solar wind ion temperature17 eV
Solar wind ram pressure (Pram=2PdynP_{\mathrm{ram}} = 2P_{\mathrm{dyn}})10.6 nPa
Alfvén Mach number MAM_{\mathrm{A}}4.8
Planetary magnetic field (multipole)−190,−75,−22-190, -75, -22 nT

Table 1. Input parameters used for the AIKEF model. The solar wind parameters are adapted from Winslow et al. [55]. The planetary magnetic field is considered up to the octupole moment [4].

In this study, our simulations account only for protons (H+^{+} ions) of solar wind origin as the dominant ion species. The effects of heavier solar wind ions or ions of planetary origin, such as Na+^{+}[38], are neglected.

Results and discussion

Magnetospheric regions and penetration depth

In Figure 2 panels a and c, the quasi-steady state magnetosphere of Mercury is depicted through the ion number density nin_{\mathrm{i}} as meridional and equatorial "slices" through the simulation. Additionally, panels b and d illustrate the electric current density ∣J∣=1/μ0∣∇×B∣|\mathbf{J}| = 1/\mu_{0}|\nabla\times\mathbf{B}|, where B\mathbf{B} represents the magnetic field and μ0\mu_{0} the vacuum permeability, once again in meridional and equatorial planes. The projection of the mostly equatorial MSB2 trajectory is displayed in all subplots of Figure 2. Crossings of BepiColombo through different magnetospheric regions are highlighted in panel a: (1) Inbound magnetopause crossing, between (2) and (3) grazing of the magnetotail, (4) the outbound magnetopause crossing after the closest approach, and (5) the outbound bow shock crossing at about 2 RMR_{\mathrm{M}}.

Image not extracted

Figure 2. AIKEF MSB2 simulation overview. Panels a and b show modeled plasma density nin_i and electric current density ∣J∣|\mathbf{J}| in the XY (equatorial) plane (MASO). Panels c and d illustrate the same parameters but in the XZ (meridional) plane. The black thin lines in panels a and c represent the bulk flow velocity, whereas the black thin lines in b and d indicate magnetic field lines. The red thick lines with a white outline correspond to the projected MSB2 trajectory, with crosses indicating the points where the simulation "slice" is intersected. The models for magnetopause and bow shock are illustrated by the black dashed [55] and black solid [44, 55] lines.

We find it important to note here that Figure 2 displays only one slice through the simulation, and hence the MSB2 trajectory intersects each of these planes only at one point (marked with crosses). This also accounts for the tail current sheet that is tilted [41] and intersects the equatorial plane, for instance at YMSO=0.5RMY_{\mathrm{MSO}} = 0.5 R_{\mathrm{M}} (see Figure 2 a and b). Magnetospheric regions and boundaries become more evident when considering the particle distributions in Figure 3.

Image not extracted

Figure 3. Model results for the proton energy spectrum along BepiColombo Mercury swingby 2 trajectory. Panel a shows all particles, and panel b only shows particles moving in a sunward (vx<0v_x < 0) direction. The white horizontal dashed line represents the solar wind energy input parameter, and CA denotes the closest approach. The numbers 1-5 indicate magnetospheric region or boundary crossings: (1) inbound magnetopause crossing, (2 to 3) grazing of the magnetotail, (4) the outbound magnetopause crossing, and (5) the outbound bow shock crossing. The white bar between 09:19 UTC and 10:02 indicates the area where a higher resolution (refinement) was used.

In general, considering Figure 2 panels a and c, solar wind protons are spread throughout the magnetosphere. The lowest number densities of >0.01> 0.01 cm−3^{-3} are found at the dayside magnetopause, and a density of about (0.1−1)(0.1 - 1) cm−3^{-3} can be observed in the tail lobes (comparable to MESSENGER results [60]). The density in the dayside magnetosheath is about 130 cm−3^{-3}, and the highest density of 250 cm−3^{-3} can be seen at the cusp regions, where the particles can precipitate onto the surface, which is a driving mechanism for planetary ions that make up Mercury’s exosphere [40, 16].

Our results indicate a dawn-dusk asymmetry of the magnetosheath thickness in the equatorial plane (see Figure 2, panels a and b). As the model itself is symmetric about the X-axis, the difference might be caused by the IMF direction and also influenced by the upstream solar wind condition resulting in a foreshock on the dawnside. There are several widely used models that assess the locations of the bow shock and magnetopause, including the classic Shue model [42], a paraboloid model proposed by Alexeev et al. (2008), and the KT14 and KT16 models [30, 29], among others. These models often assume (anti-)parallel IMF conditions, leading to a rotational symmetry around the X-axis. However, studies on the Earth’s magnetosphere have shown clear signatures of asymmetries at the quasi-perpendicular (dusk) side of the bow shock, meaning that the angle between the bow shock normal and the IMF direction, θbn\theta_{\mathrm{bn}}, is greater than 45∘45^\circ, and the magnetosheath is thicker than on the quasi-parallel (dawn) side due to the greater extent of the shock [50, 13]. Models that include this dawn-dusk asymmetry have been developed (e.g., by Tsyganenko (2002)). This effect is also observable in our simulation, as can be seen in Figure 2 panels a and b. Compared to the modeled bow shock [44] using empirical parameters from [55], a difference in the general shape as well as a slight shift to the dusk side is evident in the simulation. Moreover, while in Figure 2 panels a and b the quasi-perpendicular bow shock is a clear shock front, the quasi-parallel shock shows the development of a foreshock region. We find it important to note here that in Figure 2 the ion density is shown in the geographical equatorial plane, deviating slightly from the shifted Mercury Solar Magnetospheric (MSM) XY-plane (see Figure 1), in which the modeled bow shock and magnetopause are shown. The magnetopause seems to be symmetric around ZMASO=0.2RMZ_{\mathrm{MASO}} = 0.2 R_{\mathrm{M}}, comparing well to the magnetopause model from Korth et al. [30].

Evaluating the areas of the highest electric current densities in Figure 2 panel d, the subsolar magnetopause and bow shock stand-off distances (at ZMASO=0.2RMZ_{\mathrm{MASO}} = 0.2 R_{\mathrm{M}}) are respectively RMP=1.45RMR_{\mathrm{MP}} = 1.45 R_{\mathrm{M}} and RBS=1.85RMR_{\mathrm{BS}} = 1.85 R_{\mathrm{M}}, generally reproducing the empirical models by Slavin et al. [44] and Winslow et al. [55]. The bow shock stand-off distance is about 2% smaller than the minimal expected value of 1.89 RMR_{\mathrm{M}} [55]. This might be caused by numerical inaccuracies due to the limited grid resolution in the simulation.

At the subsolar magnetopause (most) solar wind protons can penetrate down to 1100 km toward the planetary surface and are deflected around the planet and its intrinsic magnetic field. This is illustrated by the density drop toward the planet and by the thin black lines depicting the ion bulk flow velocity in Figure 2 panels a and c. The penetration depth decreases slightly toward the dawnside direction, as can be seen in Figure 2 panel a. In addition, one can see trapped particles in terms of closed streamlines at X=0RMX = 0 R_{\mathrm{M}}, Y=1.5RMY = 1.5 R_{\mathrm{M}} in Figure 2 panel a. As the MSB2 trajectory passes right through this area, these particles are also seen in the energy spectrum in Figure 3 and are discussed in the following section.

Penetrating ion energy profile

To predict the ion energy profile along the spacecraft trajectory through the magnetosphere, we used a method that extracts the

particle data from the completed simulation snapshot. This approach offers the flexibility to adjust the trajectory with minimal computational effort, regardless of the complexity of the underlying simulation.

To determine the plasma energy distribution along MSB2, we extracted all particles within a constant distance of 5 ion inertial lengths (in the solar wind) and gathered their velocity distribution. Subsequently, an omnidirectional velocity distribution and thus energy spectrum of the ions was obtained without the direct limitation of a limited mesh grid resolution. Ideally, if one aims to reconstruct measurements by a particle detector, the radius of the "particle cloud" at each trajectory point should be determined by the thermal velocity of the particles and the measurement or exposure time of the instrument. Notably, the larger the radius, the higher the number of particles within the cloud, resulting in improved statistical accuracy. We chose a radius of 5 ion inertial lengths in the solar wind (roughly 200 km) in order to obtain sufficient data points (N>1000N > 1000 per trajectory point) to compute a spectrum.

The resulting ion kinetic energy spectrum along the trajectory is presented in Figure 3. It shows the particle distribution functions as histograms at each trajectory point alongside one another. The displayed energy ranges from 1 eV to 20 keV, with 60 logarithmic energy bins per histogram. We note that while Figure 3 uses time as the X-axis, the analysis is based on adjusting the spatial position within a quasi-stationary state at the end of the simulation. To guide the eye, we indicate the solar wind initialization temperature of 1 keV with a horizontal dashed line in Figure 3.

During MSB2, the BepiColombo spacecraft entered the magnetosphere at the post-terminator duskside flank of the magnetosheath. Due to the size of the simulation box used, the spectrum in Figure 3 panel a starts at 08:45 UTC, and the spacecraft is already located in the magnetosheath. The particle energy peaks are just under 1 keV, which equates to the solar wind input energy. Due to the distance of X>+4RMX > +4 R_{\mathrm{M}} at Y∼−3RMY \sim-3 R_{\mathrm{M}}, the solar wind that was decelerated and heated at the bow shock accelerates again to solar wind speed [22].

Nevertheless, considering the width of the energy distribution at about 09:10 UTC in Figure 3 panel a, a few particles have energies that exceed 1 keV due to acceleration mechanisms at the duskside flank of the inbound magnetopause. After (1) in Figure 3, the energy decreases gradually to about 50 eV at 09:33 UTC. This lowering of the energy by one order of magnitude is seen as being a result of a passage from the magnetopause through the tail plasma lobes. Between 09:35 UTC and 09:40 UTC, marked with (2), a particle population with an energy range of about 20 to 200 eV is observed. Taking into account Figure 2 panels c and d, this is interpreted as a grazing of the tail current sheet from its southern boundary. In the area between (3) and (4), the density increases, and there is a particle population with energies of a few electron volts up to a maximum of 1 keV.

Starting from (3) in Figure 3, the energy first decreases until about 09:47 UTC and then increases again until (4). This population can be identified with the trapped particles seen in Figure 2 panel a. These particles have a predominantly negative x-velocity component (see Figure 3 panel b), meaning they move in a sunward direction, and compared to Figure 2 panel a, a convection pattern can be identified. These accelerated particles are due to magnetotail reconnection and drift planetward and toward the magnetopause. Between the closest approach and (4) there is another particle population accelerated to higher energies centered at about 6 keV and reaching up to 10 keV. These particles are accelerated near the (dawnside) flanks of the tilted tail current sheet. Due to the proximity to the dawn tail reconnection site, the energies are higher compared to the generally lower energy population between (3) and (4). Point (4) roughly marks the outbound magnetopause crossing. In the dayside magnetosheath, ranging from (4) to (5), the energy ranges from a 10 eV up to about 4 keV. At around 09:54 UTC, point (5), the spacecraft traversed through the bow shock into the solar wind. Between (5) and about 10:05 UTC, there are also ions with a sunward velocity component, despite the solar wind bulk flow (see Figure 3 panel b), that potentially describe a field-aligned ion beam in the foreshock region that was identified in Figure 2 panels a and c.

While precise times are indicated, we note that an averaging effect exists in the energy spectrum as a result of employing a radius of about 200 km for the computation of the particle spectrum. Consequently, boundary regions become "blurred" and timings may not be exact. Moreover, the indicated particle energies may deviate from the real-world scenario, especially the low-energy population in the magnetosphere. In that regard, our simulation results are more appropriately viewed as a qualitative representation of the real flyby.

Summary and conclusion

We used a global 3D hybrid plasma simulation to investigate the entry of solar wind protons into Mercury’s magnetosphere. Particle densities as well as the electric current were used to evaluate magnetospheric regions and boundaries. We utilized solar wind upstream IMF conditions derived from averaged measurements obtained by the BepiColombo MGF instrument during the MSB2 phase that are also commonly observed near Mercury [24]. The MSB2 trajectory was used as an example to evaluate proton energy distributions throughout the southern magnetosphere and to quantitatively estimate boundary crossings.

The magnetopause and bow shock stand-off distances at the subsolar point are RMP=1.45RMR_{\mathrm{MP}} = 1.45 R_{\mathrm{M}} and RBS=1.85RMR_{\mathrm{BS}} = 1.85 R_{\mathrm{M}}. The magnetopause distance falls within the range of anticipated values [55], and likewise, the bow shock distance corresponds well with values documented by Slavin et al. [44]; Winslow et al. [55]. Looking at the general shape of the magnetosphere, a clear difference between empirical models and our simulation can be seen (Figure 2). The asymmetry in the equatorial plane has been found to be a result of the oblique IMF direction during MSB2. Solar wind particles enter the magnetosphere as they pass through the bow shock and are deflected around the magnetopause. They follow magnetic field lines at the cusps toward the surface and can also enter the magnetosphere through magnetic reconnection at the magnetotail. During these reconnection processes, the particles are accelerated and convect around the planet toward the dayside magnetopause.

We initialized the simulation with solar wind protons that exhibit an average kinetic energy of 1 keV. The resulting energies of these protons influenced by Mercury’s magnetosphere vary by three orders of magnitude, ranging from a few electron volts near the dawnside terminator lower than a few 100 km surface altitude (around the closest approach of MSB2) up to about 4 keV in the magnetosheath region on the pre-terminator dawnside, which is consistent with earlier observations by MESSENGER [60]. Moreover we observed a grazing of the magnetotail plasma sheet. After the bow shock crossing, there are indications for an ion beam in the foreshock region of Mercury.

For this study, we utilized a steady state snapshot of Mercury’s magnetosphere. While the positions of boundaries and energy distributions were investigated and estimated, the dynamics of the system are not covered. Mercury’s magnetospheric system, however, is constantly changing due to IMF variability [17, 15] and the fast Dungey cycle of about 2 minutes Slavin et al. [44]. Therefore, our results should be regarded as a quantitative representation, especially if compared to particle measurements during MSB2. The flyby interval of about one to two hours might allow reconfigurations that potentially affect boundary positions and orientations.

Future work will entail comparing modeled data with ion measurements. This could include investigating dynamic effects, methods to obtain directional particle spectra, and incorporating ion species of planetary origin to investigate exospheric processes. Moreover, simulations with varying IMF and solar wind conditions can be used to provide forecast scenarios for all upcoming Mercury swingbys, contributing to mission/operation planning of BepiColombo.

Acknowledgements. This work was financially supported by the Austrian Research Promotion Agency (FFG) ASAP PICAM-4 under contract 885349. DT acknowledges the Mio MGF team (W. Baumjohann as PI, A. Matsuoka as Co-PI) for providing the mean values of interplanetary magnetic field for the BepiColombo flyby-2 outbound trajectory. Furthermore, DT acknowledges constructive discussions with colleagues from the Space Research Institute (IWF) Graz. WE has been supported by an ESA research fellowship. WE and UM acknowledge the North-German Supercomputing Alliance (HLRN) for providing HPC resources that have contributed to the research results reported in this paper. VisIt 3.2.2 has been used to generate the 2D-plots [11].

References

  1. [1]Aizawa, S., Griton, L. S., Fatemi, S., et al. 2021, Planetary and Space Science, 198, 105176, doi:10.1016/j.pss.2021.105176
  2. [3]Anderson, B. J., Acuña, M. H., Korth, H., et al. 2008, Science, 321, 82, doi:10.1126/science.1159081
  3. [4]Anderson, B. J., Johnson, C. L., Korth, H., et al. 2012, Journal of Geophysical Research: Planets, 117, doi:10.1029/2012JE004159
  4. [5]Baker, D. N., Poh, G., Odstrcil, D., et al. 2013, Journal of Geophysical Research: Space Physics, 118, 45, doi:10.1029/2012JA018064
  5. [6]Baker, D. N., Pulkkinen, T. I., Angelopoulos, V., Baumjohann, W., & McPherron, R. L. 1996, J. Geophys. Res., 101, 12975, doi:10.1029/95JA03753
  6. [7]Baumjohann, W., Matsuoka, A., Narita, Y., et al. 2020, Space Sci. Rev., 216, 125, doi:10.1007/s11214-020-00754-y
  7. [8]Benkhoff, J., van Casteren, J., Hayakawa, H., et al. 2010, Planetary and Space Science, 58, 2, doi:https://doi.org/10.1016/j.pss.2009.09.020
  8. [9]Cassidy, T. A., Merkel, A. W., Burger, M. H., et al. 2015, Icarus, 248, 547, doi:https://doi.org/10.1016/j.icarus.2014.10.037
  9. [10]Cheng, A., Johnson, R., Krimigis, S., & Lanzerotti, L. 1987, Icarus, 71, 430, doi:https://doi.org/10.1016/0019-1035(87)90038-8
  10. [11]Childs, H., Brugger, E., Whitlock, B., et al. 2012, in High Performance Visualization—Enabling Extreme-Scale Scientific Insight, 357–372DOI
  11. [12]Delcourt, D. C., Grimalt, S., Leblanc, F., et al. 2003, Annales Geophysicae, 21, 1723, doi:10.5194/angeo-21-1723-2003
  12. [13]Dimmock, A. P., Nykjæri, K., Osmane, A., & Pulkkinen, T. I. 2016, Advances in Space Research, 58, 196, doi:10.1016/j.asr.2015.09.039
  13. [15]Exner, W., Heyner, D., Liuzzo, L., et al. 2018, Planetary and Space Science, 153, 89, doi:https://doi.org/10.1016/j.pss.2017.12.016
  14. [16]Exner, W., Simon, S., Heyner, D., & Motschmann, U. 2020, Journal of Geophysical Research: Space Physics, 125, e27691, doi:10.1029/2019JA027691
  15. [17]Fatemi, S., Poirier, N., Holmström, M., et al. 2018, A&A, 614, A132, doi:10.1051/0004-6361/201832764
  16. [18]Fatemi, S., Poppe, A., Delory, G., & Farrell, W. 2017, Journal of Physics: Conference Series, 837, 012017, doi:10.1088/1742-6596/837/1/012017
  17. [19]Fatemi, S., Poppe, A. R., & Barabash, S. 2020, Journal of Geophysical Research: Space Physics, 125, doi:https://doi.org/10.1029/2019JA027706
  18. [20]Gamborino, D. & Wurz, P. 2018, Planetary and Space Science, 159, 97, doi:https://doi.org/10.1016/j.pss.2018.04.021
  19. [22]Gershman, D. J., Slavin, J. A., Raines, J. M., et al. 2013, Journal of Geophysical Research: Space Physics, 118, 7181, doi:https://doi.org/10.1002/2013JA019244
  20. [23]Hercik, K., Travníček, P. M., Johnson, J. R., Kim, E.-H., & Hellinger, P. 2013, Journal of Geophysical Research: Space Physics, 118, 405, doi:10.1029/2012JA018083
  21. [24]James, M. K., Imber, S. M., Bunce, E. J., et al. 2017, Journal of Geophysical Research: Space Physics, 122, 7907, doi:10.1002/2017JA024435
  22. [25]Jarvinen, R., Alho, M., Kallio, E., & Pulkkinen, T. I. 2020, Monthly Notices of the Royal Astronomical Society, 491, 4147, doi:10.1093/mnras/stz3257
  23. [26]Jia, X., Slavin, J. A., Gombosi, T. I., et al. 2015, Journal of Geophysical Research: Space Physics, 120, 4763, doi:10.1002/2015JA021143
  24. [28]Jia, X., Walker, R. J., Kivelson, M. G., Khurana, K. K., & Linker, J. A. 2009, Journal of Geophysical Research (Space Physics), 114, A09209, doi:10.1029/2009JA014375
  25. [29]Korth, H., Johnson, C. L., Philpott, L., Tsyganenko, N. A., & Anderson, B. J. 2017, Geophysical Research Letters, 44, 10, doi:https://doi.org/10.1002/2017GL074699
  26. [30]Korth, H., Tsyganenko, N. A., Johnson, C. L., et al. 2015, Journal of Geophysical Research: Space Physics, 120, 4503, doi:https://doi.org/10.1002/2015JA021022
  27. [33]Müller, J., Simon, S., Motschmann, U. et al. 2011, Computer Physics Communications, 182, 946, doi:10.1016/j.cpc.2010.12.033
  28. [34]Müller, J., Simon, S., Wang, Y.-C., et al. 2012, Icarus, 218, 666, doi:10.1016/j.icarus.2011.12.028
  29. [35]Ness, N., Behannon, K., Lepping, R., & Whang, Y. 1975, Journal of Geophysical Research, 80, 2708, doi:10.1029/JA080i019p02708
  30. [36]Paral, J. & Rankin, R. 2013, Nature Communications, 4, 1645, doi:10.1038/ncomms2676
  31. [37]Paral, J., Travníček, P. M., Rankin, R., & Schriver, D. 2010, Geophysical Research Letters, 37, doi:10.1029/2010GL044413
  32. [38]Potter, A. & Morgan, T. 1985, Science, 229, 651, doi:10.1126/science.229.4714.651
  33. [40]Raines, J. M., Dewey, R. M., Staudacher, N. M., et al. 2022, Journal of Geophysical Research: Space Physics, 127, doi:https://doi.org/10.1029/2022JA030397
  34. [41]Romanelli, N., DiBraccio, G. A., Slavin, J., Bowers, C., & Weber, T. 2022, Geophys. Res. Lett., 49, doi:10.1029/2022GL101643
  35. [42]Shue, J.-H., Chao, J. J., Fu, H., et al. 1997, Journal of Geophysical Research, 102, 9497, doi:10.1029/97JA00196
  36. [44]Slavin, J. A., Acuña, M. H., Anderson, B. J., et al. 2009, Science, 324, 606, doi:10.1126/science.1172011
  37. [45]Slavin, J. A., Baker, D. N., Gershman, D. J., et al. 2018, in Mercury: The View after MESSENGER, ed. S. C. Solomon, L. R. Nittler, & B. J. Anderson, 461–496, doi:10.1017/9781316560684.018
  38. [46]Slavin, J. A. & Holzer, R. E. 1979, Journal of Geophysical Research: Space Physics, 84, 2076, doi:https://doi.org/10.1029/JA084iA05p02076
  39. [47]Slavin, J. A., Imber, S. M., & Raines, J. M. 2021, Magnetospheres in the solar system, 535, doi:https://doi.org/10.1002/9781119815254.ch34
  40. [48]Smith, D. E., Zuber, M. T., Phillips, R. J., et al. 2012, Science, 336, 214, doi:10.1126/science.1218809
  41. [49]Solomon, S. C., McNutt, R. L., Gold, R. E., & Domingue, D. L. 2007, Space Sci. Rev., 131, 3, doi:10.1007/s11214-007-9247-6
  42. [50]Sundberg, T., Boardsen, S. A., Slavin, J. A., et al. 2013, Journal of Geophysical Research: Space Physics, 118, 6457, doi:10.1002/jgra.50602
  43. [51]Travníček, P., Hellinger, P., & Schriver, D. 2007, Geophysical research letters, 34, doi:10.1029/2006GL028518
  44. [52]Travníček, P. M., Hellinger, P., Schriver, D., et al. 2009, Geophysical Research Letters, 36, doi:10.1029/2008GL036630
  45. [54]Winske, D. & Omidi, N. 1996, Journal of Geophysical Research: Space Physics, 101, 17287, doi:https://doi.org/10.1029/96JA00982
  46. [55]Winslow, R. M., Anderson, B. J., Johnson, C. L., et al. 2013, Journal of Geophysical Research (Space Physics), 118, 2213, doi:10.1002/jgra.50237
  47. [56]Wurz, P., Fatemi, S., Galli, A., et al. 2022, Space Science Reviews, 218, 10, doi:10.1007/s11214-022-00875-6
  48. [57]Wurz, P., Whitby, J. A., Rohner, U., et al. 2010, Planet. Space Sci., 58, 1599, doi:10.1016/j.pss.2010.08.003
  49. [58]Yagi, M., Seki, K., Matsumoto, Y., Delcourt, D. C., & Leblanc, F. 2010, Journal of Geophysical Research: Space Physics, 115, doi:10.1029/2009JA015226
  50. [59]Yagi, M., Seki, K., Matsumoto, Y., Delcourt, D. C., & Leblanc, F. 2017, Journal of Geophysical Research: Space Physics, 122, 10,990, doi:10.1002/2017JA024082
  51. [60]Zhao, J.-T., Zong, Q.-G., Slavin, J. A., et al. 2020, Geophysical Research Letters, 47, doi:https://doi.org/10.1029/2020GL088075

Paper details

vlnc:W_7867kzdu
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents