Juno Observations of Heavy Ion Energization During Transient Dipolarizations in Jupiter Magnetotail

Abstract

Transient magnetic reconnection and associated fast plasma flows led by dipolarization fronts play a crucial role in energetic particle acceleration in planetary magnetospheres. Despite large statistical observations on this phenomenon in the Earth’s magnetotail, many important characteristics (e.g., mass or charge dependence of acceleration efficiency and acceleration scaling with the spatial scale of the system) of transient reconnection cannot be fully investigated with the limited parameter range of the Earth’s magnetotail. The much larger Jovian magnetodisk, filled by a mixture of various heavy ions and protons, provides a unique opportunity for such investigations. In this study, we use recent Juno observations in Jupiter’s magnetosphere to examine the properties of reconnection associated dipolarization fronts and charged particle acceleration. High-energy fluxes of sulfur, oxygen, and hydrogen ions show clear mass-dependent acceleration with energy ~ m^1/3. We compare Juno observations with similar observations in the Earth’s magnetotail and discuss possible mechanism for the observed ion acceleration.

1. Introduction

Magnetic field energy release and transformation to plasma heating and high-energy particle acceleration are usually associated with magnetic reconnection (Birn & Priest, 2007), which plays a key role in energizing the Earth’s magnetotail plasma populations (Birn et al., 2012; Paschmann et al., 2013). The universality of this process in space plasma systems (e.g., Aschwanden, 2002; Angelopoulos et al., 2008; Gosling, 2012; Hoshino & Lyubarsky, 2012; Louarn et al., 2015) motivates many theoretical and numerical investigations. A need to verify these theories drives many observational investigations on the magnetic reconnection, among which the most accessible space plasma system for such investigations is the Earth’s magnetotail (Gonzalez & Parker, 2016). However, the Earth’s magnetotail is rather spatially localized (especially in comparison with various stellar and astrophysical systems) and mostly filled by protons with minor contribution from heavy ions. These facts restrict the verification of many theories on ion acceleration driven by magnetic reconnection. On the other hand, in the magnetospheres of giant planets (like Jupiter and Saturn), the parameter range of magnetic reconnection is significantly expanded, which provides a unique opportunity to investigate properties of this process with in situ plasma and magnetic field measurements (e.g., Kasahara et al., 2011; Kronberg et al., 2019; Yao et al., 2017).

One important property of the magnetic reconnection in hot magnetospheric plasma is the transient nature of this process: instead of steady reconnection outflow predicted by quasi-stationary models (Petschek, 1964; Sonnerup, 1970; Vasyliunas, 1975), flows are actually bursty with strong magnetic field pulse (Longcope & Priest, 2007; Semenov et al., 1992; Syrovatskii, 1981), usually referred to as dipolarization fronts (Nakamura et al., 2002; Sitnov et al., 2009; Runov et al., 2009). Due to this transient nature of the magnetic reconnection, direct acceleration of charged particles within the X-line region (Bulanov & Sasorov, 1976; Burkhart et al., 1990; Zelenyi et al., 1990) is rather limited, but charged particles gain most energy from interaction with dipolarization fronts (Artemyev et al., 2012; Artemyev & Vasiliev, 2015, Birn et al., 2013, 2015; Gabrielse et al., 2012, 2016; Ukhorskiy et al., 2013, 2017; Zhou et al., 2012, 2018). Besides reflection from the front, which can be responsible for ambient ion acceleration (Greco et al., 2014; Zhou et al., 2010), one of the most promising mechanism is the charged particle trapping around the front (Birn et al., 2017; Gabrielse et al., 2017; Ukhorskiy et al., 2018) and quasi-adiabatic heating due to front propagation closer to the planet, across an increasing background magnetic field. In addition, ion interaction with the wave activity around the front may also contribute to ion heating, but this has been less intensively investigated (Chaston et al., 2014; Greco et al., 2017; Grigorenko et al., 2017; Lin et al., 2017).

Adiabatic acceleration does not depend on mass or charge by itself (Birn et al., 2012), but many additional factors can introduce such a dependence into the system. First, the finite ion gyroradius effect (which is even more important for heavy ions) results in violation of the adiabaticity in ion motions (Artemyev et al., 2015) and thus can provide mass/charge dependence (Greco et al., 2015). Second, broadband electromagnetic turbulence accompanying the dipolarization fronts (Chaston et al., 2012; Ergun et al., 2015; Kronberg et al., 2019) can locally demagnetize ions and violate adiabatic approximation (Catapano et al., 2017; Greco et al., 2017). Third, combination of ion drifts along the front boundary and gradient drifts in the ambient inhomogeneous field destroys ion trapping and can provide charge-dependent acceleration (Ukhorskiy et al., 2018). Observational evidence for such deviations in adiabatic heating has been reported for plasma injections in the Earth magnetoshere (Mitchell et al., 2018; Motoba et al., 2018).

The Jupiter magnetodisk, however, is an almost ideal system to examine heavy ion acceleration by dipolarization fronts that are frequently observed there (Kasahara et al., 2013; Vogt et al., 2010, 2020).

The jovian plasma and energetic charged particle environment is composed of an admixture of hydrogen, sulfur, and oxygen ions (e.g., Krupp et al., 2004; Mauk et al., 2004; Thomas et al., 2004). Associated heavy ion charge states are predominately O⁺ and S⁺⁺; however, studies have shown that this may vary with energy and/or distance from Jupiter (e.g., Allen et al., 2019; Clark et al., 2016; Selesnick & Cohen, 2009).

Large spatial scales of the system (the Jupiter radius, R_J, is ≈ 11R_E) and strong planetary magnetic field result in trapping of even high-energy heavy ions, which would be easily scattered and lost during bounce oscillations in the Earth’s magnetotail. Despite different triggering mechanisms of reconnection in rotation-dominated Jovian magnetodisk versus in solar wind controlled Earth magnetotail (Vasyliunas, 1983), transient reconnections in both systems share many common properties (Kronberg et al., 2005, 2008; Yao et al., 2019), for example, formation of fast plasma flows and dipolarization fronts (Angelopoulos et al., 2013; Kasahara et al., 2013), generation of field-aligned currents and flows (Grigorenko et al., 2011; Zhou et al., 2012; Mauk & Saur, 2007; Kronberg et al., 2012), plasma injections into the inner magentoshere (Birn et al., 1997a, 1997b; Mauk et al., 1997, 2002), and aurora signatures (Radioti et al., 2011; Sergeev et al., 2004; Lyons et al., 2012). All these motivate investigations on the Jupiter magnetodisk reconnection.

In this paper, we analyze several cases of dipolarization fronts observed by Juno (Bagenal et al., 2017) in the Jupiter magnetodisk. Using magnetic field (Connerney, Benn, et al., 2017) and high-energy particle (Mauk, Haggerty, Jaskulek, et al., 2017) measurements, we investigate typical energies of accelerated heavy ions and protons. By comparing observations with simple analytical estimates (Birn et al., 2012), we reveal possible mechanisms of ion acceleration by dipolarization fronts in Jupiter’s magnetosphere.

2. Dataset and Methodology

We use the first 16 orbits of Juno observations, when the spacecraft crossed the equatorial planet at radial distances > 40R_J from the planet. To identify strong dipolarization fronts, we use Juno magnetometer (MAG) measurements with 1s resolution (Connerney, Adriani, et al., 2017; Connerney, Benn, et al., 2017). Following Kasahara et al. (2011) and Vogt et al. (2010), we select large-amplitude peaks of magnetic field component directed along Jupiter dipole field. All selected events (except one) are within the group of the strongest fronts from statistics collected by Vogt et al. (2020). Applying minimum variance analysis (MVA; see Sonnerup & Cahill, 1968), we construct the coordinate system using one period of magnetodisk flapping just before dipolarization front observations. This coordinate system determines l, m directions as main directions of most varying magnetic field components, that is, components of the magnetodisk current sheet. Thus, n direction is nominally along the normal to the current sheet plane, and this component is expected to be strongest at dipolarization fronts with the classical geometry (Runov et al., 2011; Sitnov et al., 2009). We select events with B_n larger than (or comparable with) magnitudes of B_l, B_m; that is, dipolarization fronts should be observed around the current sheet center (B_l reversal) and current sheet should have the simplest configuration with B_m ≪ B_l (Artemyev et al., 2014; Behannon et al., 1981; Connerney et al., 1981). For each selected event, we examine the measurements from Juno Energetic Particle Detector Instrument (JEDI) that provides ion distributions (including times of flight, which establishes the mass distribution) above 10 keV (which is species dependent) and electron distributions within the energy range of 40 keV to 0.5 MeV (Mauk, Haggerty, Jaskulek, et al., 2017; Mauk, Haggerty, Paranicas, et al., 2017). We only keep events with sufficiently high (>0.05/cm²/s/sr/keV) fluxes of energetic oxygen, sulfur, and hydrogen ions (all available energy channels should have fluxes above the noise level). Our list includes six such events, and two of them have high temporal-resolution (one measurement per ~ 3 s) ion fluxes.

Standard data products provided by JEDI include energy and pitch angle resolved fluxes of hydrogen, helium, oxygen, sulfur, and combined oxygen and sulfur. Helium fluxes are rather low in the magnetodisk and are provided with low energy resolution (three energy channels); thus, we do not use this data product. For other data products, we average fluxes over all angular directions to improve the statistics. Hydrogen fluxes are grouped to spectrum at energy channels of (35, 55, 100, 200, 400, 900) keV. Oxygen fluxes are grouped to spectrum at energy channels of (500, 750, 1,250, 1,700, 3,500) keV. Sulfur fluxes are grouped to spectrum at energy channels of (600, 900, 1,400, 3,500, 7,000) keV. Combined oxygen and sulfur fluxes are grouped to spectrum at energy channels of (175, 200, 300, 370) keV. Spectra are averaged with 40-s running window to reduce the fluctuation level.

3. Ion Energization

Figure 1 shows the first event from our list. The north-south magnetic field component B_θ (B_n) exhibits a strong peak around 12:39 UT, indicative of a dipolarization front. The leading edge of the front is characterized by B_θ (B_n) drop, which is the characteristic feature of dipolarization fronts from models (Lu et al., 2016; Sitnov et al., 2009) and observations (Runov et al., 2011; Zhang et al., 2011) in the Earth’s magnetotail. The front separates two energetic particle populations: ahead of the front proton (H⁺) fluxes are mainly in the energy range of < 150 keV, and behind the front, we observe a decrease of < 150-keV proton fluxes and increase of fluxes up to 1 MeV. This distribution is very typical of energetic particle fluxes around the dipolarization front (Hwang et al., 2011; Runov et al., 2009): fronts separate the ambient dense plasma population (with low high-energy fluxes) from the accelerated ions with low-density fluxes from the reconnection region (see simulations in Birn et al., 2013, 2015; Gabrielse et al., 2016, 2017). Bottom panels of Figure 1 show the same behavior from all ion fluxes (hydrogen, oxygen, and sulfur) around dipolarization fronts in Jovian magnetodisk: a decrease of the ambient dominant population and increase of high-energy population behind the front (note that fluxes of < 400 keV from the combined oxygen and sulfur data set are shown in both O⁺ and S⁺ panels).

Figure 1.

Event #1: Juno observations at radial distance of ~ 41R_J. Top panels shows magnetic field: radial B_r, azimuthal B_ϕ, and north-south B_θ components are in Panel (a), whereas Panel (b) shows three components in the local coordinate system reconstructed for the current sheet crossing prior to the dipolarization front observation. The B_n peak around 12:39 UT marks the dipolarization front. Three bottom panels show energetic ion fluxes (see text for details); black curves are contours of the ion flux. The dashed black box shows the leading edge of the dipolarization front. Colored lines mark the times as analyzed in Figure 3.

Figure 2 shows the second event from our list. The strong B_θ (B_n) peak associated with the dipolarization front is accompanied by similar B_l variations (i.e., the local coordinate system does not separate well n and l directions) and large-amplitude (~ 5 nT) magnetic field fluctuations. Such magnetic field fluctuations resemble the mixture of small-scale dipolarization front (B_θ (B_n) fluctuations accompanied by energetic flux variations, for example, at 03:23–03:24 UT, kinetic Alfven waves, and compressional magnetosonic waves usually observed around strong dipolarization fronts in the Earth’s magnetotail (Chaston et al., 2012; Gabrielse et al., 2016; Hwang et al., 2011; Runov et al., 2014). Despite these strong magnetic field fluctuations, the dipolarization front at 03:33 UT can be well recognized from changes in energetic ion fluxes: decrease at lower energy channels and increase at higher energy channels (compare with Figure 1) (note that different from Figure 1, we do not include the combined S⁺ and O⁺ channel into the bottom two panels of Figure 2).

Figure 2.

Event #2: Juno observations at radial distance of ~ 49R_J. Top panel shows magnetic field: radial B_r, azimuthal B_ϕ, and north-south B_θ components are in Panel (a), whereas Panel (b) shows three components in the local coordinate system reconstructed for the current sheet crossing prior to the dipolarization front observation. The B_n peak around 03:33 UT marks the dipolarization front. Three bottom panels show energetic ion fluxes (see text for details); black curves are contours of the ion flux. The dashed black box shows the leading edge of the main dipolarization front. Colored lines mark the times as analyzed in Figure 4.

To characterize typical energies of accelerated heavy ions, observed behind the dipolarization front, we plot the 1D energy spectra at different moments of time (ahead of and behind the front) for both events (see top panels in Figures 3 and 4). There is a clear increase of fluxes (for all types of ions) behind the front, and this increase is energy dependent: fluxes decrease (or remain the same) at the lowest energy channels and remain the same (or slightly increase) at the highest energy channels, whereas majority of the flux increase is observed at intermediate energy channels. We then normalize the spectra measured behind the front to the spectra ahead of the front (shown in gray) and plot the ratios in the bottom panels of Figures 3 and 4. Peaks of these ratios demonstrate that the accelerated ion population is energy dependent (similar observations can be found for accelerated ions behind dipolarization fronts in the Earth’s magnetotail; see Artemyev et al., 2015; Malykhin et al., 2019). Energies at the peak of the normalized flux can be considered as typical energies of accelerated ions.

Figure 3.

Fluxes of high-energy ions for Event #1 from Figure 1. Top panels show spectra ahead of the front (gray) and spectra behind the front (colors) at different time moments. Bottom panels show spectra behind the front normalized to the spectra ahead of the front.

Figure 4.

Fluxes of high-energy ions for Event #2 from Figure 2. Top panels show spectra ahead of the front (gray) and spectra behind the front (colors) at different time moments. Bottom panels show spectra behind the front normalized to the spectra ahead of the front.

Figure 5 shows the energy dependence of the normalized flux maximum, E_max, as a function of ion mass (for hydrogen m_i/m_p = 1, oxygen m_i/m_p = 16, and sulfur m_i/m_p = 32). There is a clear increase of E_max with m_i, from ~ 300, 1,000 keV for hydrogen to ~ 1, 000, 4,500 keV for sulfur in the first and second event, respectively. This dependence is somewhere between $m_i^{1/2}$ (blue dashed curve) and $m_i^{1/3}$ (green dashed curve). Such a proportionality of $~m_i^{1/2}$ can be found in simulations of ion acceleration due to reflection from the front (Greco et al., 2015), whereas explanation of $m_i^{1/3}$ requires more complicated acceleration conditions (see details in section 4).

Figure 5.

Energies at peaks of the normalized spectra for Event #1 from Figure 1 (left) and for Event #2 from Figure 2 (right). Red error bars mark the standard deviation of E_max for all spectra behind the front (for intervals shown in Figures 1 and 2). Black error bars show width of energy channels at E_max. Blue diamond S^+* has the same energy as S⁺, but plotted against m_i/m_p = 64 (see text for details). Black dashed curve fits three red points, whereas green dashed curve fits red points of H⁺, O⁺ and blue diamond S^+*.

We should note that sulfur ions (especially high-energy ion population) can be a mixture of singly charged ions and, more dominant, doubly charged ions (Clark et al., 2016). Therefore, there may be a mixture of mass and charge dependence for the ion acceleration. To take this into account, we plot in Figure 5 the sulfur ion E_max versus 2m_i/m_p: this point does not follow the $m_i^{1/2}$ curve, but follows the $m_i^{1/3}$ curve (green dashed curve). In section 4, we explain the factor of ~ 2 in mass dependence under a mixture of singly charged ions and doubly charged ions.

Figure 6 shows the E_max(m_i) profiles for additional four events from our list (time resolution of flux measurements is low for these events, but fluxes ahead of and behind the front can be well distinguished). For these events, we observe the same mass dependence of E_max as shown in Figure 5; that is, we generally deal with mass-dependent acceleration by dipolarization fronts in the Jovian magnetodisk.

Figure 6.

Same format as shown in Figure 5, but for four additional events of dipolarization front observations: 2017 DOY 136 02:05–02:15 at R ~ 44R_J; 2017 DOY 234 17:15–17:30 at R ~ 85R_J; 2017 DOY 185 18:10–18:30 at R ~ 67R_J; 2019 DOY 37 05:00–05:30 at R ~ 68R_J.

4. Discussion and Conclusions

Figures 1–6 show properties of ambient and accelerated ion populations observed around dipolarization fronts in the Jupiter magnetodisk. Within this section, we compare these observations with THEMIS measurements in the Earth’s magnetotail and discuss theoretical models that may describe mass-dependent acceleration.

4.1. Comparison with the Earth’s Magnetotail

Figures 3 and 4 show that the accelerated ion population are observed within an energy range: The lowest energy channels and the highest energy channels show a decrease (or weak variation) in flux, whereas majority of the flux increase relative to the ambient fluxes is seen for intermediate energies (i.e., below 1 MeV for protons and below 2–3 MeV for heavy ions). Such flux change across the dipolarization fronts is typical for the Earth’s magnetotail as well. Figure 7 shows one example of dipolarization front observed by two THEMIS (Angelopoulos, 2008) spacecraft in the near-Earth magnetotail. Note that the magnetic field magnitudes are comparable for fronts from Figures 1 and 2 and Figure 7, but dipolarization fronts in the Jupiter magnetodisk are expected to be much larger due to larger kinetic scales: The thin current sheet thickness, as determined by typical ion gyroradius scale, is much larger in the Jupiter magnetodisk (Artemyev et al., 2014). This difference in spatial scales likely leads to the difference in energies of accelerated ion populations. Spectra shown in Figure 7 demonstrate energy-localized acceleration with the typical energy E_max ~ 50 keV (note that this ion population is dominated by protons), whereas E_max varies from 250 to 1,000 keV for accelerated ions observed in the Jupiter magnetodisk. This ratio of energies E_max,Earth/E_max,Jupiter ~ 5 – 20 should directly relate to the spatial scale ratio; for example, the maximum electric potential drop along the front, E_yL_y = (v_front/c)B_nL_y (y is the dawn-dusk coordinate), is determined by the front speed v_front (which is on the order of several hundreds of km/s both for near-Earth Runov et al., 2011 and Jupiter magnetodisk Kasahara et al., 2013, Vogt et al., 2020) and L_y scale that is only determined by typical spatial scales, L_y,Jupiter/L_y,Earth ~ R_J/R_E ~ 10. The similar scale relation would work for ion acceleration within the reconnection region and further trapped by dipolarization fronts (see the next subsection).

Figure 7.

An example of dipolarization front observed in the near-Earth magnetotail on 3 December 2015 by two THEMIS spacecraft (ThE and ThD): top panel shows GSM B_z magnetic field measured by fluxgate magnetometer (Auster et al., 2008) with shadow region showing the leading edge of the dipolarization front, middle panels show ion energy spectra from combined measurements of electrostatic analyzer (McFadden et al., 2008) and solid state telescope (Angelopoulos et al., 2008), bottom panels show energy fluxes ahead of (gray) and behind (colors) the dipolarization fronts (colored lines in the top panel show the corresponding time moments).

4.2. Theoretical Estimates

Ion acceleration by dipolarization fronts includes ambient ion reflection (such ions form the front precursor population; see Greco et al., 2014; Zhou et al., 2012), ion passage along the front (Birn et al., 2013, 2017), ion resonant acceleration ahead of the front (Artemyev et al., 2013; Ukhorskiy et al., 2013), and ion trapping by strong front magnetic field (Ukhorskiy et al., 2018). The former mechanism seems to be responsible for the formation of accelerated ion population observed behind the front: ions accelerated within the reconnection region drifts around the B_n (B_z) peak and move with the front across the spatial gradient of ambient B_n. Such trapping acceleration resembles adiabatic acceleration; that is, ions gain energy due to transport across the inhomogeneous magnetic field in the presence of front electric field. Such an acceleration should be mass independent, and thus, explanation of results shown in Figures 5 and 6 requires mass dependence in the energy of ions trapped within the reconnection region, E_max ~ f(B)ε_X, where f(B) is a certain function of the background magnetic field, which in the simplest case would be ratio of field magnitudes (see details in Zelenyi et al., 2013), and ε_X is the energy of ions accelerated within the reconnection X-line. Therefore, we need ${\epsilon }_X~m_l^{1/3}$, and this is the case for ion acceleration in the simplest model of X-line (Artemyev et al., 2014; Bulanov & Sasorov, 1976; Burkhart et al., 1990) that gives ${\epsilon }_X=m_i{\left(c{q}_i{L}_X{E}_y^2/B{m}_i\right)}^{2/3}~m_i^{1/3}{q}_i^{2/3}$, where q_i is the ion charge, L_X is the spatial scale of X-line, and E_y is the reconnection electric field. This model predicts that ε_X is limited due to instability of ion motion around the X-line current sheet (Birn et al., 2012; Bulanov & Sasorov, 1976). The energy gain rate is given by equation ${\dot{\epsilon }}_X\approx q_i{E}_y{v}_y\approx q_i{E}_y\sqrt{2E_x/m_i}$ (as the velocity along X-line well exceeds other velocity components during the acceleration; see Bulanov & Sasorov, 1976, and Burkhart et al., 1990), and this equations should be integrated for the time interval Δt that particles can spend within the acceleration region: ε_X ≈ m_i(q_iE_yΔt/m_i)². Time Δt is determined as the typical time scale of particle |x|-coordinate increase around the X-line. For magnetic field expansion around the X-line (i.e., B_z is given by the linear function of ≈ B · (x/L_X)), we obtain $\ddot{x}=q_i{v}_y{B}_z/c{m}_i\approx q_i^2{E}_{y}Btx/c{m}_i^2{L}_X$. The solution of this equation is exponentially growing Airy function, that is, x ~ exp((t/Δt)^3/2) with $\Delta t={\left(m_i^{2}c{L}_X/q_i^{2}B{E}_y\right)}^{1/3}$ (e.g., Birn et al., 2012). Substituting Δt into ε_x ≈ m_i(q_iE_yΔt/m_i)², we obtain ${\epsilon }_X=m_i{\left(c{q}_i{L}_X{E}_y^2/B{m}_i\right)}^{2/3}~m_i^{1/3}{q}_i^{2/3}$, and the predicted mass-dependence $~m_i^{1/3}$ is close to the result shown above.

It should be mentioned, however, that the proposed scenario of ion acceleration does not include effects of nonadiabatic ion motion (and associated acceleration; see Birmingham, 1982; Cheng, 1990; Cheng & Decker, 1992; Selesnick et al., 2001) in the vicinity of the magnetodisk equatorial plane (where magnetic field gradients are sufficiently strong; see, e.g., Connerney et al., 1981). These effects could be important for ion scattering and detrapping, but their accurate consideration requires numerical simulations of dipolarization front propagation within rotating magnetodisk. Moreover, in several events, the E_max(m_i) dependence can be approximated by $~m_i^{1/2}$ (see two bottom panels in Figure 6). Therefore, nonadiabatic effects of ion acceleration due to reflection from the dipolarization fronts (Greco et al., 2014; Zhou et al., 2011) can also contribute to the formation of high-energy ion populations. Front interaction with initially cold particles would result in an energy gain of ~ m_i (e.g., Artemyev et al., 2014; Zimbardo, 2011; Zhou et al., 2011), whereas preheated ions would gain an energy of $~m_i^{1/2}$ (see Greco et al., 2015). Although efficiency of this nonadiabatic acceleration is limited by the small front velocity, ~100–500 km/s (Kasahara et al., 2013) (note that the energy gain is proportional to the front velocity for the $~m_i^{1/2}$ regime; see Greco et al., 2015), this acceleration can still be effective for heavy ions.

It is interesting to note that if sulfur ions consist of a mixture of singly and doubly charged populations, then $E_{\text{max}}~m_i^{1/3}{q}_i^{2/3}$ for such a mixture should be larger than for singly charged ions. Taking 50% of singly charged ions and 50% of doubly charged ions (Clark et al., 2016), we obtain averaged charge q_i ~ 3/2 and E_max ~ (2m_i)^1/3. Figure 5 shows that sulfur E_max indeed can follow the $m_i^{1/3}$ fitting with a factor of 2 for mass dependence.

Electron acceleration behind the dipolarization front is also dominated by trapped acceleration (Gabrielse et al., 2016, 2017). Therefore, one can expect $E_{\text{max},e}/E_{\text{max},H^{+}}~{\left(m_e/m_i\right)}^{1/3}~1/12$. However, electron acceleration within the X-line (determining the electron initial energies E_X) is more effective than ion acceleration due to the electron trapping by electrostatic (Hall) electric fields (i.e., so-called X-line surfing acceleration; see Artemyev et al., 2013; Hoshino, 2005). Such electric field trapping increases E_X by a factor of ~2–3 (Artemyev et al., 2013), that is, $E_{\text{max},e}/E_{\text{max},H^{+}}~2.5\cdot {\left(m_e/m_i\right)}^{1/3}~1/5$. Figure 8 shows electron spectra for Event #1 from Figure 1. Typical electron energies are E_max,e ≈ 100 keV, whereas $E_{\text{max},H^{+}}$ for this event is ≈ 400 keV (see Figure 5, left panel). Therefore, the mass dependence of E_max (ranging from electrons to a mixture of singly and doubly charged sulfur ions) follows simple theoretical prediction of the charged particle acceleration in the laminar reconnection X-line, E_max ~ m^1/3. Comparison of $E_{\text{max},H^{+}}$ for Earth magnetotail and Jupiter magnetodisk gives E_max,Earth/E_max,Jupiter ~ (L_X,Earth/L_X,Jupiter)^2/3, that is, for E_max,Earth/E_max,Jupiter ~ 50 keV/400 keV, the X-line scale L_X,Jupiter ~ 20L_X,Earth. The X-line aspect ratio L_X/L_Z is determined by the reconnection rate and is usually (for a wide parameter range) L_X/L_Z ~ 10 (Liu et al., 2017), where the spatial scale across the reconnecting current sheet, L_Z, is about ion inertial length d_i or ion gyroradius ${\rho }_i=d_i\sqrt{{\beta }_i}$ (and plasma β is comparable at Earth magnetotail and Jupiter magnetodisk; see Frank et al., 2002). In multicomponent plasma, the X-line region is multiscaled, with the largest scale determined by heaviest ion species (Liang et al., 2016, 2017), that is, d_i,Jupiter = d_S,Jupiter. Typical ion density around the expected reconnection region (radial distances of ~20–40R_E) in the Earth’s magnetotail is n_Earth ~0.1–0.2 cm⁻³ (see discussion in Artemyev et al., 2017), whereas typical ion density around the expected reconnection region (radial distances of ~60R_J; see Kasahara et al., 2013; Vogt et al., 2020) in the Jovian magnetodisk is n_Jupiter ~ 0.01cm⁻³ (e.g., Artemyev et al., 2014). Therefore, for the ratio of ion inertial length, we obtain d_i,Jupiter/d_i,Earth = d_S,Jupiter/d_p,Earth ≈ (m_S/m_p)^1/2 · (n_Earth/n_Jupiter)^1/2 ≈ 5.5(n_Earth/n_Jupiter)^1/2 > 15. Therefore, the factor of L_X,Jupiter/L_X,Earth ~ 20 (and the associated ratio of E_max,Earth/E_max,Jupiter ~ 1/8) can be potentially explained by larger ion inertial scales in the Jovian magnetodisk.

Figure 8.

High-energy electron spectrum for events shown in Figure 1: top panel show spectrum versus time, bottom left panel shows several spectra ahead of (gray) and behind (colors) the front, bottom right panel shows the spectra behind the front normalized to the spectrum ahead of the front.

4.3. Conclusions

In this study, we investigate heavy ion acceleration by dipolarization fronts observed in the Jupiter magnetodisk. The main conclusions of this study are as follows:

• Typical energies of accelerated fluxes are scaled with mass as ~ m^1/3 with a correction due to multicharging of sulfur ions.

• The observed energy scaling may be explained by adiabatic ion acceleration by dipolarization fronts that trap ions after their acceleration in the X-line.

• Preliminary comparison of ion and electron acceleration, and comparison of dipolarization fronts observed in the Jupiter magnetodisk and Earth magnetotail also supports the above scenario.

Our results demonstrate the importance of reconnection-induced acceleration in the Jupiter magnetodisk. In contrast to the Earth magnetotail, the Jupiter magnetosphere is filled by heavy ions, and this provides a unique opportunity to investigate the mass dependence of the acceleration mechanism. Scaling as ~ m^1/3 was not predicted by any models of ion acceleration by dipolarization fronts but can be well described by acceleration within the X-line. This implies that acceleration on the front is almost adiabatic; that is, this acceleration keeps the same dependence as one for initial energies (scaled as ~ m^1/3). Further numerical and theoretical investigations are needed to include this interesting result into modern models of ion energization in planetary magnetospheres.

Note that in this study, we only analyzed properties of dipolarization fronts, but not their origins. Although magnetic reconnection is the most well-known mechanism of dipolarization fronts (Nakamura et al., 2002; Runov et al., 2009; Sitnov et al., 2009), the interchange instability can also produce front-like magnetic field structures (Panov et al., 2012; Panov & Pritchett, 2018; Pritchett & Coroniti, 1998, 2011). The interchange instability can indeed develop in the fast rotating Jovian magnetodisk (Achilleos et al., 2015; Hill, 1976; Southwood & Kivelson, 1989), and further analysis on the stability of the Jovian magnetodisk current sheet is needed to reveal the dipolarization front origin in Jupiter.

Key Points:

• Heavy ion and proton acceleration by transient dipolarizations in the Jupiter magnetodisk

• Ion acceleration results in flux increases in the energy range of [0.5,3] MeV

• Observations show that ion acceleration scales with mass as ~ m^1/3