Neptune's Pole-on Magnetosphere: Dayside Reconnection Observations by Voyager 2

Understanding the magnetosheath plasma conditions at a magnetosphere is important, as it is the magnetic field and plasma in the magnetosheath near the magnetopause that interacts with the magnetosphere. As the supermagnetosonic solar wind is decelerated and heated to form the magnetosheath, the draped IMF is compressed and can "squeeze" plasma away from the subsolar magnetopause region. This results in a decrease in the plasma density near the magnetopause and forms a plasma depletion layer (PDL; Zwan & Wolf 1976). This occurs at any planetary magnetosphere and has been particularly well studied at Earth, Mercury, and Saturn (e.g., Slavin et al. 1983; Phan et al. 1994; Gershman et al. 2013; Masters et al. 2014). The thickness of the PDL is dependent on the upstream solar wind conditions and varies as ${M}_{{\rm{A}}}^{-2}$ (Zwan & Wolf 1976). When the solar wind has a high M_A, the magnetosheath plasma β is also high (where β is the plasma pressure–to–magnetic pressure ratio). This means the plasma is more fluidlike as the plasma dominates over the magnetic pressure (Gershman et al. 2013). For low-M_A solar wind upstream conditions, the magnetosheath plasma β is low, the plasma pressure decreases and the magnetic pressure increases closer to the magnetopause.

A PDL was clearly observed by V2 at Neptune's magnetopause (Figure 2). A decrease in the PLS estimated electron density was observed (Figure 2(d)) that would signify a decrease in the plasma pressure. During the electron density decrease, the magnetic field magnitude (Figure 2(h)) steadily increased, and there was no discernible jump in magnetic field magnitude to identify the magnetopause location. This is supported by the low-M_A solar wind value, reported to be ∼9 (Richardson et al. 1995; Szabo & Lepping 1995). This is lower than the average theoretical range for M_A values of 13–24 expected at Neptune (Slavin et al. 1985; Masters 2015). The resulting magnetosheath plasma β was estimated to be ∼1 (Lepping et al. 1992).

The change in plasma β across the magnetopause is important when characterizing the occurrence and rate for magnetic reconnection. Theory suggests that when there is a large difference in plasma β (Δβ) across the magnetopause current layer, a charged particle drift acts to interrupt the reconnection jets, thereby suppressing magnetic reconnection from occurring. This occurs when the particle drift has a velocity component that is nonzero to the reconnection jets, resulting in reconnection only occurring at very high magnetic shear angles (close to 180°) when Δβ is high (Swisdak et al. 2003, 2010; Phan et al. 2010). At Earth, the plasma β in the magnetosheath is relatively low, ∼1, and reconnection can, on average, take place when the magnetic shear between the interplanetary and magnetospheric magnetic fields is >90° (e.g., Trenchi et al. 2008). At Mercury, a lower M_A in the solar wind and a lower magnetosheath plasma β results in reconnection being possible for shear angles as low as ∼20° (Slavin et al. 2009, 2014; DiBraccio et al. 2013; Sun et al. 2020). As the solar wind travels radially outward into the heliosphere, the solar wind M_A increases with increasing distance from the Sun (e.g., Masters 2018). At Saturn, this results in a typical magnetosheath plasma β of ∼10; therefore, conditions at Saturn's dayside magnetopause are more likely to be suppressed (Masters et al. 2012). This is also supported by in situ measurements of reconnection at Saturn. Dayside magnetopause flux ropes that are generated by multiple X-line reconnection are more likely to be ion-scale in size (d_i ∼ 1) at Saturn, since the main flux rope growth mechanism of continuous reconnection is suppressed (Jasinski et al. 2021) in comparison to Earth, where the average flux rope size is >30 d_i (Akhavan-Tafti et al. 2018).

The conditions during this particular flyby at Neptune's magnetopause show that the plasma conditions were very conducive to magnetic reconnection. A relatively low solar wind, M_A ∼ 9, at Neptune produced conditions that are more representative of the inner heliosphere, where reconnection is more significant in driving the terrestrial magnetospheres (the average M_A at Mercury is 3.5; Sarantos & Slavin 2009). A plasma β of 1.2 and 0.7 in the magnetosheath and magnetosphere, respectively (Lepping et al. 1992), results in Δβ < 1, resulting in conditions for this particular flyby that were terrestrial-like and ripe for magnetic reconnection. An observed PDL during this flyby shows that the plasma β conditions in the magnetosheath were reduced close to the magnetopause, therefore corresponding to increased magnetosheath Alfvén speeds, which would also increase the reconnection rate (Borovsky & Hesse 2007; Cassak & Shay 2007).

When V2 crossed the magnetopause into the cusp, PLS observed a decrease in the measured current (Figure 2(c)). Such measurements are typical when crossing from the magnetosheath into a region of newly "opened" magnetospheric flux and have previously been observed by NASA's Cassini spacecraft at Saturn's cusp (Jasinski et al. 2014, 2016a) or a region of newly reconnected magnetospheric field lines in a near-equatorial magnetopause crossing (called the "open region" by Jasinski et al. 2016b during a particularly "reconnection-active" Cassini orbit).

Furthermore, PLS measured changes in the cusp plasma properties labeled by the white arrows. The white arrows show sudden modest increases in energy and the measured current. Such variation in the cusp plasma has been typically observed at Saturn's cusp (Jasinski et al. 2014) and suggests that V2 crossed through different reconnected magnetic flux tubes, which were formed from "bursts" of reconnection occurring at different locations along the magnetopause. The variability in the electron measurements is very similar to the variability in the magnetosheath plasma; therefore, this could also be explained by a single reconnection location with temporal variation. However, considering that V2 measured three separate electron events (white arrows in Figure 2(c)) separated briefly by background levels of electron current, we suggest that the spacecraft was instead crossing separate flux tubes connected to different reconnection locations. This is also supported by possible evidence of multiple X-line reconnection, discussed below in Section 3.3.

Evidence of transient and spatial variation of reconnection during cusp observations is usually also accompanied by measurements of ion energy–time dispersions (Lockwood & Smith 1994; Lockwood et al. 2001), which are common at the cusp at Earth, Mercury, and Saturn (e.g., Reiff et al. 1977; Pitout et al. 2009; Raines et al. 2014; Arridge et al. 2016; Jasinski et al. 2016a). This is caused by a velocity filter effect, where low-energy ions have a lower field-aligned velocity in comparison to more energetic ions. The ions are therefore spread in energy and latitude. This effect is typically more pronounced at the giant magnetospheres. Arridge et al. (2016) showed that the increase in scale at Saturn results in the magnetosheath ions injected into the cusp traveling a distance 10 times larger in comparison to Earth. Cusp ion dispersions are therefore more pronounced in a larger magnetosphere.

Although the ion measurements were close to the noise level, cusp ion dispersions were not observed at Neptune's cusp (see Figure 1 of Richardson et al. 1991). This, however, is not surprising during the V2 flyby at Neptune's cusp. The Neptunian magnetosphere is in the pole-on configuration. The cusp crossing for this flyby is adjacent to the magnetopause; therefore, the observed ions have traveled an insignificant distance (on the order of an R_N) with very little cross-latitudinal drift. Therefore, an ion dispersion would not be expected to be observed, since a velocity filter effect would not be able to develop like the dispersions observed at Saturn, where ions have traveled a distance approximately 2 orders of magnitude greater (Jasinski et al. 2016a). Observations made of pole-on magnetospheric cusps are more likely to be similar to the open near-equatorial region observed at Saturn during a reconnection event (Jasinski et al. 2016b). A cusp ion dispersion observation will more likely be observed at lower altitudes during the pole-on magnetosphere, when the ions have traveled a significant distance and cross-polar convection of the reconnected magnetic field occurs.

Further evidence of an open magnetopause is apparent from the LECP observations in Figures 2(a) and (b). There was a marked increase in the count rate (from ∼2 to ∼10 counts s⁻¹) of the energetic field-aligned electron population at the magnetopause boundary, as well as before it, indicating streaming energetic electrons on reconnected field lines. There were also similar variations at the 1 count level in the lowest energy proton channel of LECP (not shown here). The planetward-moving magnetopause (planetward velocity of ∼148 km s⁻¹; Lepping et al. 1992) may also explain the observation of energetic particles of magnetospheric origin over an hour before the actual magnetopause crossing, considering the spacecraft velocity was ∼17 km s⁻¹. These increases are shown by red arrows in Figure 2(b) and show regions where field-aligned energetic electrons were observed (red). Measurements of streaming electrons at or close to the magnetopause provide evidence that reconnection has occurred at the magnetopause. Similar observations of streaming electrons have occurred at Saturn (e.g., Badman et al. 2013; Fuselier et al. 2014, 2020; Sawyer et al. 2019; Jasinski et al. 2021), as well as at Earth (Gosling et al. 1990; Fuselier et al. 2012). Such observations of unidirectional streaming electrons are an indicator of open field lines. Crossing from a region where unidirectional electrons are observed to a region with bidirectional electrons is also indicative of a spacecraft crossing the OCB (labeled by the final dashed vertical line). Fuselier et al. (2014) reported similar observations when Cassini crossed from the magnetosheath into the low-latitude boundary layer. The difference here is that V2 crosses from the open cusp region, and so the bidirectional electrons are only observed upon crossing into the high-latitude closed magnetosphere. This energetic streaming population is indicative of escaping magnetospheric electrons. Like the observations at Saturn, we can learn from the directionality of the electrons about the spacecraft location with respect to the reconnection site. The electrons are field-aligned, so the spacecraft is "northward" and dawnward of the reconnection location.

The cusp crossing also shows that two magnetic field depressions were measured by the magnetometer (arrows in Figure 2(h)). The dense cusp plasma in the magnetosphere causes diamagnetic effects where a depression in the magnetic field is observed. Similar depressions in the outer magnetospheric cusp at Saturn have also been observed (Jasinski et al. 2017a).

To the right of the first black arrow is a sudden modest increase in the magnetic field magnitude (highlighted by a gray arrow in Figure 2(h) and the dashed–dotted vertical line in Figure 3). At first glance, this enhancement would appear to be a signature of a flux rope crossing, similar to the low-amplitude flux ropes observed at Saturn's dayside magnetopause (Jasinski et al. 2016b, 2021). However, a bipolar signature is not visible in Figure 3, and further minimum variance analysis (Sonnerup & Cahill 1967), not shown here, confirms this, revealing no evidence for the complex field structures (e.g., bipolar signature in the maximum varying direction and a peak at the center of the flux rope in the intermediate direction) that would be expected of a flux rope crossing. Instead, this could be a remote detection of a flux rope passing near V2, where an enhancement in the magnetic field is observed due to the draped magnetic field (called traveling compression regions, TCRs; Slavin et al. 1994).

Zoom In Zoom Out Reset image size

As a flux rope is carried away from the reconnection site, the local magnetic field compression (due to draping of the field around the "bulge" of the flux rope) travels with the flux rope. Observations at Earth's magnetotail show that flux ropes start out with diameters in the X–Y plane that are very large compared with the thickness of the plasma sheet in the Y–Z plane. However, the tension, or curvature forces, in the helical magnetic fields of the flux rope quickly evolve it toward a more cylindrical, lower-energy state as it accelerates away from the X-line. These changes cause a local swelling of the plasma sheet and, in doing so, launch a fast mode wave into both lobes that compresses and drapes the field about the bulge much like a bow wave (Slavin et al. 1994). From the magnetic field observations (Figure 3), there is an increase in the magnetic field intensity, which is mostly in the Y direction. This is also accompanied by a modest decrease in the X and an increase in the Z directions. Similar measurements of TCRs have been reported at Earth (e.g., Slavin et al. 1994). Multiple X-line reconnection is the mechanism that generates flux rope formation and has been well studied at Earth using spacecraft observations, as well as simulations (e.g., Lee & Fu 1985, 1986; Raeder 2006; Zhang et al. 2008; Zhong et al.2013; Akhavan-Tafti et al. 2019, 2020). This signature could therefore be evidence of multiple X-line reconnections taking place at Neptune's magnetopause that generated an observation of a TCR that accompanied a passing flux rope. Multiple TCRs have also been observed at Saturn's dayside magnetopause during particularly "reconnection-active" magnetopause crossings (Jasinski et al. 2016b).

As discussed above (Section 3.1), the occurrence and rate of magnetic reconnection, and therefore a magnetosphere's ability to couple and therefore be affected by the solar wind through reconnection, is dependent on the plasma Δβ across the magnetopause, as well as the magnetic shear. This results in certain IMF orientations providing more favorable high magnetic shear conditions for reconnection onset to occur across a large area of the dayside magnetopause, for example, a southward (−B_Z ) IMF orientation for Earth or northward (+B_Z ) for Saturn. At Mercury, even though magnetic reconnection can take place for low local magnetic shears (DiBraccio et al. 2013), it has been shown that magnetic reconnection is more likely to drive magnetospheric dynamics under a southward IMF orientation (Jasinski et al. 2017b; Slavin et al. 2019b). These IMF orientations therefore provide configurations where the largest possible area of the magnetopause surface is conducive to reconnection so that closed magnetospheric flux can be opened. Furthermore, at Saturn, the amount of time the IMF is oriented in a particular direction to produce very high shear angles to the magnetosphere is low due to the IMF usually being oriented in the Parker spiral direction.

However, the case of a pole-on magnetosphere is unique in the solar system. Here the magnetic shear conditions for magnetic reconnection are independent of the IMF orientation. Let us assume the worst-case scenario for reconnection: a high plasma Δβ at the magnetopause, which requires antiparallel magnetic shears (∼180°), conditions similar to those at Saturn and contrary to the conditions that were present during the V2 flyby at Neptune. For this scenario, there will always be a portion of the dayside magnetopause that will provide the required antiparallel shears required for magnetic reconnection regardless of the IMF orientation. This occurs precisely because of the pole-on configuration, where the direction of the magnetospheric field will diverge away from the subsolar point and be directed away from the subsolar point. This is shown to some extent in Figure 1, where the schematic could be rotated around the X-axis, and the magnetospheric field configuration in the illustration would stay the same. This means that the same proportion of the magnetopause will be conducive to reconnection and opened for any IMF orientation. The IMF orientation only dictates which sector of the magnetopause is opened (i.e., dawn/dusk/"north"/"south").

Masters (2015) assessed the reconnection conditions across the magnetopause at Neptune. They showed that the location of reconnection is dependent on season and phase. Masters (2015) did not discuss how, for the pole-on configuration, approximately a quarter of the magnetopause will always have a shear angle >150°, independent of the IMF orientation (their Figure 4, column 3). Figure 4 shows an adapted figure from Masters (2015); here we show the location on the magnetopause where the IMF and magnetospheric fields are antiparallel (i.e., shears of 180°) for different clock angles for the Neptune pole-on V2 case. Masters (2015) also estimated that phases other than the pole-on configuration at Neptune can produce a larger fraction of the magnetopause area where reconnection is permitted; however, these phases are dependent on the IMF orientation. Even though the pole-on magnetosphere yields a lower fraction of the magnetopause that can be reconnected, this configuration will always produce reconnection, independent of the IMF orientation. This guarantee of solar wind coupling during the pole-on configuration is therefore far more important for understanding typical dynamics at the ice giants than previously considered.

Zoom In Zoom Out Reset image size

This is an important consideration when attempting to understand the possible solar wind driving of the ice giant magnetospheres. Gershman & DiBraccio (2020) investigated the dependence of solar wind coupling with the outer planet magnetospheres for solar cycle, as well as season. The authors assessed the relative occurrence of high magnetic shears at the subsolar points of the outer planets. However, considering that the pole-on magnetospheric configuration for Neptune and Uranus is not perfectly aligned with the subsolar point—e.g., Neptune's pole is positioned slightly "southward" of the subsolar point, therefore at the subsolar point, the magnetic field would be pointing "northward"—there is an inherent bias introduced into their results, since their analysis does not take into account the special circumstance of the pole-on configuration. This means that the authors were most likely underestimating the relative occurrence with which Uranus and Neptune can couple to the solar wind through reconnection during a pole-on magnetosphere.

Eggington et al. (2020) simulated the terrestrial global magnetosphere for different dipole tilt angles, including a pole-on configuration. They found that reconnection always occurs at the magnetopause for all tilt angles and that the magnetosphere is open. They also found that flux rope generation on the dayside magnetopause increased with increasing tilt angle. Similar to Masters (2015), they found that the reconnection voltage is lower for the pole-on case; however, their simulation only includes a southern-oriented IMF, which maximizes the reconnection voltage for an Earth-like tilt (of ∼10°) and would be considerably lower for a Parker spiral IMF orientation for an Earth-like configuration.

The above-discussed IMF independence for reconnection at a pole-on magnetosphere has possible consequences for magnetospheric dynamics. The guarantee of reconnection throughout a certain phase of either Neptune's or Uranus's diurnal cycle during a season when a pole-on magnetosphere occurs may give rise to what modeling efforts have called an "on–off" or "switch-like" magnetosphere (Mejnertsen et al. 2016; Cao & Paty 2017). This was originally discussed after the Uranian flyby, when Voigt et al. (1987) proposed that the Uranian magnetosphere can go through stages of open and closed configurations. At Neptune, Mejnertsen et al. (2016) noted that reconnection "turns on as the dipole moment rotates from the Earth-like to the pole-on configuration," which suggests that the pole-on configuration contributes to forming this on–off or switch-like magnetosphere.

Furthermore, open flux transport from the dayside subsolar region to the nightside would take ∼50 minutes at Neptune, assuming an average solar wind velocity and a distance of ∼50 R_N. Assuming that reconnection at Neptune takes place twice per Neptunian day for a quarter-rotation each time, the average time in between periods of continuous magnetopause reconnection is ∼4 hr, as suggested by Mejnertsen et al. (2016). Therefore, the times required to transfer flux from the dayside to the nightside are short, compared to Neptune's rotation. This is also significantly shorter than comparable estimations at Saturn and Jupiter. Consequently, a Dungey-type cycle that is affected by asymmetric rotation of the magnetosphere could be expected to exist at Neptune, which periodically goes through phases of being more "open" or "closed" due to rotation into and out of the pole-on configuration.

Magnetohydrodynamic simulations of the Uranian magnetospheric environment have also observed such open/closed stages during both equinox and solstice (Cao & Paty 2017). However, the authors prescribed an IMF that is southward in their simulation; such an IMF would not be expected to be the typical expected IMF orientation at the ice giants. Therefore, the switch-like mechanism may have been introduced due to the selection of an unrealistic IMF. As the magnetosphere transitions through different orientations with respect to the IMF, reconnection would occur during phases preferable for reconnection with the southward IMF. Instead, we argue that this switch-like mechanism (for a more realistic set of IMF orientations closer to the expected Parker spiral, i.e., ±B_Y ) is more likely to be observed during rotation in and out of the pole-on magnetospheric configuration, where a set of antiparallel reconnection criteria are always guaranteed to occur for the pole-on phase of an ice giant diurnal cycle.

Cowley (2013) hypothesized a three-lobe pinwheel Uranian magnetospheric structure driven during equinox—when the rotation axis is directed parallel to the orbital direction—causing the magnetic dipole axis to rotate like a pinwheel into and out of the solar wind direction. This results in the Uranian magnetosphere rotating in and out of the pole-on magnetospheric configuration. Cowley (2013) prescribed a ±B_Z IMF orientation, which is also not likely to be a typical IMF orientation. However, we argue that the pole-on magnetosphere may be more critical in Cowley's pinwheel model, considering that reconnection will always take place during the pole-on portion of magnetospheric rotation.