Introduction

Global dipole magnetic fields commonly exist at solar system planets, with two exceptions (i.e., Venus and Mars). These planetary magnetic fields are persistently compressed by the high-speed solar wind, forming planetary magnetospheres. These magnetospheres vary significantly between planets. For instance, the energy released during substorms at Mercury is much greater than that measured at Earth1; at Jupiter, the rapid rotation rate of the planet, in addition to plasma sources embedded within the magnetosphere, results in a complex interaction with the solar wind2,3,4,5,6,7, in contrast with the less complicated Dungey-driven magnetosphere prevalent at Earth.

Magnetic reconnection often occurs when at least one component of the planetary magnetic field and the interplanetary magnetic field align in an antiparallel configuration. During this reconnection process, magnetic field lines near the poles are directly connected to the interplanetary magnetic field, allowing solar wind and magnetosheath particles to enter the magnetosphere. The region where newly reconnected field lines allow plasma to enter the magnetosphere is known as the magnetospheric cusp8,9,10,11,12. The particles present within the cusp region contribute significantly to various dynamic processes occurring in the magnetosphere, ionosphere, and upper atmosphere, including storms13, substorms14,15, and auroras16,17. The role that the solar wind plays in driving the Jovian magnetosphere has been a topic of active debate3,4,5,6,7,18. It is, therefore, crucial to improve our understanding of the characteristics of the cusp region and comprehend their implications for solar wind-magnetosphere coupling. Since the main auroras at Earth and Saturn are located close to the open-closed field line boundary, any auroras associated with the cusp are seen close to these main emissions16. Jupiter’s main auroral emissions, on the other hand, are thought to be generated far from the open-closed boundary; thus, any cusp-related auroras should be located within the polar cap5. However, in contrast to Earth’s typically dark polar regions, Jupiter exhibits bright and persistent polar auroras, which suggests that there are reconnection sites at unusual places connecting to the Jovian polar regions, hinting at differing cusp structures between the two planets4,19,20,21,22.

The exact distribution and features of the cusp regions within Jupiter’s magnetosphere remain unclear due to limited coverage by previous missions, which has hindered detailed investigation of Jupiter’s cusps. Although no direct reports of the Jovian cusp have been made, several studies23,24,25 have suggested cusp-related phenomena by utilizing observations from the Ulysses26 spacecraft’s 1992 flyby of Jupiter. During the encounter with Jupiter, evidence was found for open field lines linking to the polar region. Auroral hiss-like plasma waves were also detected during this period and deemed as potential evidence for the existence of a Jovian cusp8. However, due to the constraints of spacecraft instrumentation and data, there is a lack of conclusive evidence for the existence of the cusp regions23. The Juno spacecraft27 offered a valuable opportunity to observe the cusp region at high latitudes on the dusk side at Jupiter.

In this study, we utilize Juno’s plasma, magnetometer, and plasma wave datasets to present the comprehensive observation of the cusp region near the dusk side of Jupiter’s magnetosphere.

Results

The identification processes for Jovian cusps

Earth’s cusp has been comprehensively studied over the past few decades, and robust identification criteria have been developed12,28,29,30. Some cases of Saturn’s cusp have also been identified and reported in detail9,10,11. To avoid potential confusion, this study follows the Earth and Saturn cusp definitions, i.e., a part of the magnetosphere in the vicinity of the polar region at high magnetic latitudes/invariant latitude, where a significant quantity of magnetosheath plasma is detected inside the magnetopause position28,29,30. In practice, magnetosheath-like electron distributions well inside the magnetopause in high latitudes often serve as the key identification criteria10,31,32,33. Furthermore, the cusp is the region where magnetosheath plasma and momentum enter the magnetosphere and are closely associated with magnetopause reconnection. Therefore, the ion dispersion feature due to reconnection-associated velocity filtering effects is also a useful feature for identifying the cusp, as extensively applied to identify the Earth34,35,36,37 and Saturn’s cusp9,10,11. Additionally, whistler-mode auroral hiss waves, combined with plasma data, are frequently utilized to identify the cusps of Earth38,39,40 and Jupiter8. These waves serve as an auxiliary criterion to aid in the identification process. A detailed discussion of the different planetary cusp identification criteria is available in Supplementary Notes 1, 2 and Supplementary Tables 1, 2. Combining the above identification features, we obtained six typical Jupiter cusp events. In this paper, we focus on Case 1 and Case 2 (see Supplementary Note 3 for Jupiter’s cases 3–6, and Supplementary Note 4 for Earth’s and Saturn’s cases). And we examine in situ measurements mainly from Juno’s three instruments: the Juno Magnetic Field Investigation41, the Jovian Auroral Distributions Experiment42, and the Waves instrument43. The cusp event observations are presented in Figs. 1 and 2. Panels (a–e) in Figs. 1 and 2 show the magnetic field and plasma energy spectrum information. The electron pitch angle distribution and wave emission are shown in panel (f) and panel (g) in Figs. 1 and 2, respectively. Moreover, Juno’s footpoints, mapped to the polar region utilizing the JRM33 model44 and the Connerney 2020 current sheet model45 during the events, are displayed in Figs. 1h and 2h. Figures 1i and  2i show the trajectories of Juno around whole observations.

Fig. 1: The Jovian cusp observation (Case 1) on June 27 and June 28, 2023.
figure 1

a R-Theta-Phi magnetic field components in JSS (Jupiter-De-Spun-Sun) coordinate; b The total magnetic field strength; c The electron energy spectrogram; Ion Energy spectrogram for protons (d) and heavy ions (e), where heavy ions represent ions with m/q in the range of 5 and 6428; f Pitch angle distribution for electrons which is normalized at each time unit within energy ranges of 0.3 to 32 keV; g Plasma wave observations in the frequency range 50 to 300 Hz. The different regions that the spacecraft passes through are marked with different colors at the top and separated by dashed lines. “M” is the magnetosphere, “C” is the cusp, “BL” is the boundary layer. The red arrows and white dashed lines in panel (d) show the dispersion. The yellow arrows in panel (g) indicate the enhanced auroral hiss features. The blue dashed line demarcates the cusp into two regions, labeled as “a” and “b”, each characterized by different plasma properties. h Traced distribution of spacecraft footprints before and after cusp observation in left-handed system III coordinates. The blue regions are the main ovals, and the gray lines are the Juno footprint trajectories from 26 June to 28 June, 2023. Colored lines are the Juno footprint trajectories in cusps. i The position of the spacecraft around cusp observations in JSS coordinate, red lines representing the time interval of the cusp case.

Full size image
Fig. 2: The Jovian cusp Case 2 on 14 April, 2022.
figure 2

This figure uses the same format as Fig. 1. The black arrow in panel (b) indicates a decrease in the magnetic field occurring within the magnetosphere, not in the cusp. The red arrows point to the reversed ion dispersion.

Full size image

Pre-dusk cusp structure

For the first cusp case on 27 June and 28 June, 2023, Juno was primarily positioned on the dusk side (about 17.8 LT) at high invariable latitudes (i.e., 79°–86°) shown in Fig. 1h, i. Before 2023 27 June, 19:13 UT, the spacecraft was mainly in the magnetosphere, as clearly demonstrated by the typical electron spectrum around 1000 eV46, with the exceptions of two short periods (13:52-14:26 UT and 14:35-15:01 UT) when magnetosheath-like electrons were detected47 along with protons with energies lower than magnetospheric population. Additionally, in these two regions, a distinct increase in auroral hiss8,39 was observed, as marked by dashed lines. After 27 June, 19:13 UT, Juno rapidly entered into a region featuring with magnetosheath-like population, via a transition known as a boundary layer (BL) between magnetospheric and magnetosheath-like populations (Fig. 1c–g). The spacecraft stayed in the magnetosheath-like region for about 4 hours (Fig. 1c), connecting to the polar high-latitude region (Fig. 1h). The Juno spacecraft was about 54 RJ away from Jupiter, which is well inside the magnetosphere48. Moreover, the magnetic field with little perturbation also confirms that Juno was in the magnetosphere rather than magnetosheath47,49,50. From 27 June, 20:22 UT to 28 June, 02:12 UT, a notable step-like proton dispersion9,34 was observed (see Methods, subsection Location of Magnetic Reconnection, for details on the calculation of reconnection locations based on dispersion features) accompanied by intensified auroral hiss8,39. Based on the location (i.e., inside magnetopause and high latitude), plasma population (i.e., magnetosheath-like electrons and low-energy protons), proton dispersion, and auroral hiss waves, we thus confirm the observed structure is Jupiter’s cusp.

A quantitative comparison of electron energy distributions for the cusp, magnetosphere, and magnetosheath is depicted in Fig. 3a, showing that the cusp distribution is similar to the one in the magnetosheath while with lower energy flux, akin to observations made for Saturn9. The stepped proton dispersion suggests possible pulsed magnetopause reconnection, akin to the observations reported in Saturn’s cusp9,11. After 27 June 23:40 UT, within the cusp, there was an increase in the electron energy, accompanied by elevated fluxes of protons and heavy ions. The heavy ions exhibited enhancements at approximately 1000 eV and 10,000 eV, indicating a potential leakage from the magnetosphere. This diversity in plasma properties across different parts of the cusp region might suggest that the spacecraft traversed different flux tubes within the cusp, each carrying different particle streams51. Figure 1g highlights enhanced auroral hiss features within the three identified cusp regions, showing a strong correlation. And the auroral hiss is not confined solely to the cusp but also appears in the BL and some magnetospheric regions. This broader occurrence might result from the resonance cone angle’s propagation of whistler-mode auroral hiss52,53,54, which is tilted relative to the background magnetic field, allowing detection in adjacent or BL regions. The normalized electron pitch angle distribution in Fig. 1f displays the butterfly electron pitch angle distribution in the magnetosphere and its transition (BL) toward the cusp. But due to the lack of data coverage near 0° and 180°, the identification of field-aligned electrons associated with auroral hiss is challenging. As depicted in Fig. 1h, the field line tracing conducted in the System III coordinate system (which is fixed in planetary longitude and rotates with Jupiter) clearly shows the spacecraft’s footpoints positioned well above the main auroral emissions. This observation further substantiates the high-latitude positioning of the cusp.

Fig. 3: The electron energy distributions and magnetic configuration associated with the Jovian cusp.
figure 3

a The solid lines represent average electron energy distributions of the cusp Case 1 at 20:22 on 27 June − 02:12 on 28 June 2023 (green), magnetosphere on 27 June 2023 1501–1913 UT (blue), and magnetosheath (red) taken from the earlier study example on 2 October 2017 0900–1300 UT49. Noted that the dashed lines are inferred following the observational model67, as direct measurement of electrons with energies below 50 eV would involve large uncertainties. The comparison between the three populations at Jupiter is similar to the results at Saturn9. b Jupiter’s global magnetospheric topology and open-field configuration near dusk side based on the Zhang et al.2 simulation. We suggest that Juno is in the position indicated. The labels “1” and “2” in panel (b) indicate the spacecraft positions during Case 1 and Case 2, respectively, in this study. c The diagram of the footprint distribution corresponding to the open and closed magnetic field regions in magnetic coordinates in the southern hemisphere based on the picture proposed by Zhang et al.2. The labels “1” and “2” in panel (c) indicate the spacecraft footpoint positions during Case 1 and Case 2, respectively, in this study.

Full size image

Post-dusk cusp structure

The spacecraft was positioned on the post-dusk side throughout the observations, as indicated in Fig. 2i, with ~20.1 LT and a radial distance of ~58 RJ. Given that Jupiter’s magnetopause at 20 LT was predicted to be roughly 130 RJ to 200 RJ55, it can be confirmed that Juno’s location was well within the magnetosphere, which is also confirmed by the relatively stable magnetic field. Additionally, the spacecraft was at a very high invariable latitude, as noted in Fig. 2h. The electron energy spectrum in Fig. 2c shows that before 14 April, 2022, at 20:20 UT, the spacecraft was primarily within the BL. Here, the electron energy spectrum resembled that of the magnetosphere but with significantly lower fluxes. Besides the BL features, Juno also detected two short instances with clear magnetosheath-like distributions, and one short period with typical magnetospheric features, as marked by the dashed lines in Fig. 2 and the colored blocks at the top. From 20:20 UT to 21:30 UT on the same day when the magnetosheath-like electrons (~100 eV enhancement) were detected, Juno also observed clear reversed proton dispersion features11,36,56, as illustrated in Fig. 2d. Such ion reversed dispersion features are common in the Earth’s cusp with increasing latitudes under northward interplanetary magnetic field (IMF) conditions17,36, with opposite directions for magnetic lines convection motion and the spacecraft. It is important to note that no significant magnetic field depressions were observed in the cusp region. Instead, reductions only occur in the adjacent magnetosphere, as shown in Fig. 2b. In some observations of Saturn cusp events, similar situations were displayed where the magnetic depression is either absent or insignificant10,11,57. Field-aligned electrons (>150° and <30° to magnetic field lines) were detected both in the cusp and the regions immediately ahead and behind it. These were accompanied by strong whistler-mode auroral hiss waves, likely generated by the field-aligned electron beams53,58.

The magnetic configuration associated with cusp

Figure 3c depicts a schematic illustrating the distribution of footprints in magnetic coordinates for the open and closed magnetic field regions, as well as the footprint distribution of the spacecraft, based on the simulation results provided by Zhang et al.2 (see Methods, subsection Simulation Information). According to Zhang et al.’s work2, the yellow and orange areas (corresponding to the polar aurora) represent closed magnetic field regions. The blue region corresponds to the open magnetic lines extending towards the far magnetotail, as indicated by the blue lines in Fig. 3b. The green region represents the footprint of the open magnetic field region associated with coupling to the solar wind, as depicted by the green lines in Fig. 3b. This open magnetic field region takes the form of a thin strip, extending counterclockwise from the noon side to the dusk side. This magnetic field configuration may result from the complexity of Jupiter’s rapidly rotating magnetic field structure2,3 and its interaction with the By-dominated solar wind near 5 AU59,60. During the two events discussed in this study, the spacecraft was situated in the high latitudes of the southern hemisphere on the dusk side, as indicated in locations “1” and “2” in Fig. 3b, providing a unique opportunity to potentially detect the cusp. Furthermore, due to the rotation of Jupiter and the tilt angle of about 9.5 degrees between the rotation axis and the magnetic axis, the trajectory of Juno’s footprint through the Zhang et al.2 simulation under the magnetic axis would exhibit a circular shape in the polar region near the dusk side. The differences between the JSS coordinate system used in the observation and the magnetic coordinate system used in the sketch of Fig. 3c are detailed in Supplementary Note 1. The trajectory of the spacecraft footprint would cross the open magnetic field region near the dusk side at the particular time when the cusp is detected. The latitude of these distributions is higher than that of the main auroral oval.

Discussion

Our findings detail the characteristics of Jupiter’s cusp, contributing to the understanding of solar wind-magnetosphere interactions across celestial bodies. Although Jupiter’s cusp was observed on the dusk side in contrast with Earth’s noon-positioned cusp and Saturn’s cusp observed biased towards the dayside, the particle properties within the cusp regions are similar. These analogous cusp characteristics suggest that similar microphysical processes govern the nature of cusps across different planets. But the distinct location of these observations provides a better understanding of solar wind-magnetosphere coupling at rapidly rotating planets.

Recent studies using high-precision simulations2,3 have provided evidence that Jupiter’s magnetic field structure forms a distinctive spiral pattern that acts to inflate the polar and dawnside magnetosphere relative to dusk. This magnetic field configuration renders the location of Jupiter’s magnetopause reconnection site sensitive to the east-west (By) component of the IMF, as illustrated in Fig. 4a, b. Both the IMF azimuthal angle59 and the clock angle60,61 around Jupiter are reported to be predominantly around ±90°, which suggests that the solar wind conditions around Jupiter are By-dominated. When the IMF exhibits an eastward direction, the magnetopause reconnection sites are situated on the north and southeast sides. Conversely, when the IMF is oriented westward, these reconnection sites occur on the south and northeast sides of Jupiter’s magnetosphere3. In contrast, dayside magnetopause reconnection at Earth is dominated by the north-south component of the IMF. Depending on whether the IMF Bz component is positive or negative, terrestrial magnetopause reconnection occurs either in the high-latitude lobe region or at the sub-solar point. A more detailed discussion of the magnetic reconnection pictures related to the Jovian cusp can be found in Supplementary Note 5.

Fig. 4: The magnetopause reconnection picture at Jupiter and the predicted ion dispersion.
figure 4

The distribution of reconnection sites of Jupiter’s magnetic field configurations under a eastward and b westward solar wind conditions as revealed by high-precision simulations2,3 as well as the schematic representation of the convective motion (red dashed lines) of the reconnected magnetic field line and the motion of the spacecraft in the sun view (orange dashed line); c The expected reversed ion dispersion under eastward solar wind conditions in this study; d The expected ion dispersion under westward solar wind conditions in this study. Green lines and dark blue lines are newly connected IMF and planetary field lines, respectively, that have reconnected at the reconnection sites identified by the red crosses. Light blue lines are closed, with the dashed portions in the north and south connecting to each other in the far tail.

Full size image

Considering the spacecraft’s direction of motion, alongside the velocity filtering effect of detecting dispersed ions on magnetic field lines moving in different convective directions, we can anticipate the ion dispersion patterns under eastward or westward IMF conditions, as depicted in Fig. 4c, d. The spacecraft is predicted to detect reversed ion dispersion for eastward IMF, whereas, for westward IMF, a normal ion dispersion is expected. Comparing with our data in Figs. 1 and 2, we suggest that the IMF was directed westward during Case 1 and eastward during Case 2. A more comprehensive analysis of the velocity filtering effect is provided in Supplementary Note 6. And see Supplementary Note 7 for a more detailed discussion on the comparison between the cusp and similar boundary layers.

In summary, our findings have provided useful insights into the nature of Jupiter’s magnetospheric cusp. They demonstrate the consequences of rapid rotation on the location of reconnection sites and Dungey-cycle driving of Jupiter’s magnetosphere. The observational evidence presented here will contribute to comprehending the intricate interactions between the solar wind and Jupiter’s magnetosphere. This work will help form a more comprehensive picture of space weather, particularly its role in the physical connections between stars and planets—an asset to the exploration and knowledge of our solar system.

Methods

Location of magnetic reconnection

The gradient of the ion dispersion is dependent on the distance to the reconnection site. For the case in this study, Eq. (1) can be used to roughly estimate the distance of proton flow from the reconnection site to the observation site62

$$E\left({\alpha }_{o},t\right)=\frac{M}{2{t}^{2}}{\left[{\int }_{{S}_{i}}^{{S}_{o}}\frac{{ds}}{\sqrt{1-{\sin }^{2}{\alpha }_{o}(B(s)/{B}_{0})}}\right]}^{2}$$
(1)

where E is the ion energy; ds is arc length along a field line; \({S}_{i}\), \({S}_{o}\) are the observation and injection points; \(M\) is the particle mass; B(s) is the magnetic induction along the field line; \({B}_{0}\) is the magnetic induction at the observation point; \({\alpha }_{o}\) is the observed pitch angle; and t is the transit time. Consider the simplest and farthest case of particle flow, i.e., the ion pitch angle is 0. Then we have Eq. (2)

$${E}_{P}=\frac{M}{2{t}^{2}}{\left[{\int }_{I}^{P}{ds}\right]}^{2}$$
(2)

where I refers to the injection point and A represents the observation point. Two representative points on the dispersion lowest edge, namely points P and Q, are selected for the calculation29,63, as illustrated in Figs. 1d and 4d. As Fig. 1d shows, the time at point P is 20:35 on June 27, 2023 and the \({{{{{\rm{E}}}}}}_{{{{{\rm{P}}}}}}\) corresponds to 361.2 eV; The time recorded at point Q is 22:05, and the \({{{{{\rm{E}}}}}}_{{{{{\rm{Q}}}}}}\) corresponds to a value of 47.3 eV. Equation (2) outlines the formula for the motion of a proton from the injection point to point P, with the curve integral representing the distance the proton travels along the magnetic field line from the injection point to the observation point. Similarly, the equation showing the motion of the proton from the injection point to point Q must also be taken into account, which is

$${E}_{Q}=\frac{M}{2{(t+T)}^{2}}{\left[{\int }_{I}^{Q}{ds}\right]}^{2}$$
(3)

where T is the interval time from point P to point Q. Since the distance from point P to point Q is very close and it is uncertain whether the spacecraft is traveling along the magnetic field lines, we can consider this factor to be 0, i.e., \({\int }_{P}^{Q}{ds}\) = 0. So:

$${\int }_{I}^{Q}{ds}\, \approx {\int }_{I}^{P}{ds}$$
(4)

Using Eqs. (2), (3), and (4), which are binary quadratic equations, we can obtain the proton’s approximate travel distance from the injection point to observation point P:

$${\int }_{I}^{P}{ds}\, \approx \, 8.1{R}_{J}$$
(5)

Therefore, we can roughly surmise that Juno was approximately 8.1 RJ along the magnetic field from the reconnection site. Similarly, we employ Eqs. (2) to (4) and the other two sets of dispersion edge points in Case 1 (the other two dispersion shown in Fig. 1d) to calculate the path distance from the injection point to the observation, obtaining approximately 7.7 RJ and 5.6 RJ.

Simulation information

The simulation results presented in this study are derived from Zhang et al.‘s work2 using the Grid Agnostic MHD (magnetohydrodynamic) for Extended Research Applications (GAMERA) global model64. In the solar-magnetospheric (SM) coordinate system, where the axes are defined as X (Jupiter-Sun direction), Y (eastward/dusk), and Z (northward), GAMERA utilizes a specialized curvilinear, non-orthogonal grid to solve ideal MHD equations using a finite-volume method. The computational model employs a grid of 256 × 256 × 256 cells2, stretching 1200 RJ along the X-axis and 400 RJ perpendicular to it. Excluded from the computational domain is a spherical region with a radius of 6 RJ, centered on Jupiter and extending 100 RJ downstream from the point where the solar wind and IMF conditions are imposed. To represent ion mass loading from the Io plasma torus, the model uses a spatial function with a fixed ion mass loading rate of ~1000 kg per second, focused at 6 RJ in the equatorial plane2. Reflecting typical solar wind conditions65,66, the upstream IMF By, solar wind density, dynamic pressure, and velocity are set at 0.5 nT, 0.2 cm−3, 0.03 nPa, and 400 km s−1, respectively.